25 Mar 09- Locomotives 6 Jun 10
- Rolling Stock
3 Apr 09 30 Nov 09 11 Sep 09 31 Oct 09 25 Mar 09 6 Jun 10 3 Apr 09 30 Nov 09 11 Sep 09 31 Oct 09
Before either large scale GIRR ever
existed in any kind of permanent form, there was an HO layout. It
was built by my son, Richard, in late 1994 and rebuilt a year or so
later but never finished. It had been sitting in my garage ever
since, gathering a thick layer of dust and "stuff."
Around 1990, Richard got an HO set (either Bachmann or Life-like) from his grandfather. In 1994, I was in the process of basement construction for the GIRR Mountain Division. He wanted a better layout than just a loop on the floor so I bought him some materials and he went to work. He selected a track plan, shown below, and after we came home with a pile of lumber, he went to work. I was astounded when I came home from work the next day and found that the benchwork was already built. We put a sheet of Homosote on it and made a cookie cutter roadbed and scenery base and then he really went to town.
Initially, he used the starter set stuff and what still worked from the HO trains that I had when I was a kid. We added to it with some new stuff later.
He worked at it quite a while. He even built an extension that lasted for awhile. But by about 1996 or so, he lost interest and moved on to other things. The HO layout just sat. I ripped out the extension to make shelf space for my large scale stuff.
I found some old home movies spanning the time from the initial construction until sometime near the end of the construction.
By 2009, the disaster that I casually referred to as a "garage" had become so clogged with the results of a lifetime of collecting "stuff" that I had to do something about it. The first thing to go was the 3rd bench seat for my old van that my younger two boys used to commute back and forth to college. I finally put the seat back in the van because it was the major blockage, besides the HO layout, that prevented me from even walking to the roll up door. If I needed something on the other side of the seat, I had to open the door and attack it from the other side. A lot more "stuff" found it's way to the trash so that I could actually walk around 3 sides of the HO layout.
At that point, I wondered if it still worked. Richard had partially rewired it, there were two power packs on it and some cab switching but the turnout controls had never been rewired. I fired it up and, after some serious track cleaning, I found that it sort of worked. The rail was all nickel silver with Atlas turnouts and controls.
There were three locos there, an Athearn F7, some kind of GP7
and an 0-4-0 dockside of unknown origin.
[ Top ]
Richard and I
selected a track plan for the layout from a book called "Track
Plans for Sectional Track" by Linn Westcott and published by
Kalmbach. This book carries a copyright of 1961 and mine is from
the 15th printing in 1993. This may be an old book, but it is a
good one for someone getting started.
We picked this particular plan because it fit in a 4x8 space, it was a twice around with a reversing cutoff, had a couple of passing sidings, a couple of stub sidings and a yard. This is a lot to put on a 4x8 table.
It was initially built with brass flex track and turnouts left
over from the 60's but the brass was too hard to keep clean so he
replaced all the track with nickel silver, most of it is flex
track.
[ Top ]
On March 15, 2009, I started to resurrect the layout, In the two days since I started working on the layout to the to the time that I started this page I made quite a bit of progress.
After an initial track cleaning, I found some dead track. One part of the main line had a broken wire. The whole yard is dead as it had not been rewired. The reversing cutoff is dead. None of the turnout motors are wired. There are derailment problems in spots. One spot was due to out of level track. The other spots are probably similar but I cannot find the NMRA HO gauge to check the track or wheel gauge. I'll have to get one. My pocket level that I use for my large scale track is too long to get into some places. I'll need a smaller level too.
None of the three locos ran, each needed to be cleaned, lubricated and run in. Only the F7 runs smoothly, the other two still have power pickup issues. There are other locos around somewhere, but I haven't found them.
The turntable didn't work. After some fiddling around, I got the thing to play. It was wired to run from an LGB power pack but I wanted to use the LGB pack as a cab so I found a 6 volt 300 mA wall wart in my junk box and installed a DPDT switch for reversing. 6 volts is just right for the Atlas turntable. I then wired the LGB pack back as Cab A on the Atlas power control switches.
Richard had used rubber roadbed. The roadbed is tacked down to the Homosote base well enough, but the tacks holding the track down are not holding well, they tend to come up and actually catch the trains. I have reattached a section of track with a dilute solution of TiteBond II wood glue every couple of inches and this seems to do much better. When I am sure that the rest of the track is aligned, in gauge and level, I will do the rest of it that way.
I have identified most of the blocks (shown as dark slashes across the track) and labeled the switches that control them. Two insulated rail joiners appear to be missing so that two of the blocks are actually wired together in two places. I haven't located any insulated joiners in the boxes of stuff that was on the layout so I just cut the rail in two places with a Dremel cutoff wheel and inserted a styrene insulator, trimmed after the CA used to hold it in place dried.
HO trains are MUCH more sensitive to track quality than are large scale trains. In large scale, one bubble error in lateral level is the starting point of a derailment problem. In HO, one bubble off is death. Small steps at the railhead of large scale rail produces a click but not much else. In HO, a perceptible step at a rail joint is a derailment in the making.
I've gone over the whole main line with a bubble level and found that almost all of it needed work. To fix it, I needed small shims, lots of them. I discovered that regular 3" x 5" filing card stock is about 8 to 10 mils thick and makes a good shim. Since white card stock would look horrible, I painted the backside dark brown with a rattle can and then cut off 1/4" wide strips in the 3" direction and then cut those strips into 3/4" lengths to make the shims. Most of the track needed a stack of 4 shims on one side or the other every 3" or so. The track was held down with small wire nails, but these tended to come loose so I started using the nails only as a temporary patch until a mixture of 50% TiteBond II and water (with a couple of drops of detergent per pint as a wetting agent) was dribbled on the track every 3" or so to act as a more permanent bonding agent AFTER the section was leveled. I applied it with a medicine measurement syringe (free for the asking at most pharmacies) to fill the gap between two ties. The glue would eventually partially flow under the adjacent ties and fill the gap (if it existed) between the rubber roadbed and the ties. The glue is strong enough to hold the track in place, but also weak enough to allow the joint to be broken in case more work needs to be done in that area.
There were several spots were two rails didn't line up properly, most often in height, but sometimes in gauge. If a step is big enough to feel as I run my finger along the rail, it's big enough to cause a problem. Sometimes, a rail can be bent slightly about a half inch back from the joint to get to to align better. Steps are better filed off, or if the problem is that the rail isn't level, then the rail needs to be bent up or down to match. Any final difference can be filed. Rail joints that refuse to stay in place should be either soldered together and then filed or glued with Zap-A-Gap CA and then filed.
The Atlas turnouts were a systematic problem. Many cars would pick the points of the curved point rail so all of them got filed to solve that problem.
After I got the railroad in
a sufficiently stable condition that I could actually run a train,
I started dragging cars around, ending up with these two test
trains. I put all the cars with Kadee couplers on the layout. Any
car that derailed got pulled from the train. Derailments that
occurred with several different cars but at a specific spot got the
track reworked at that spot. After awhile, I had triaged the cars.
There were the ones that ran properly, I took those off and put
them back on the shelf. These cars are not helpful in finding bad
track. There were cars that just derailed all over the place, those
got put aside as they were too confusing and something is probably
wrong with the car. Then there was a set of five identical 40' box
cars lettered for ICS. I don't know the manufacturer, but it is
probably Athearn. All five of these tended to derail at specific
spots. These were the ones that I was looking for. Since all five
were essentially identical, there is something about their design
or configuration that makes them especially susceptible to less
than perfect track. Where ever these cars have problems needs track
work. When I can pull and push this set over the whole layout, I
will declare victory on the trackwork for awhile and move on to
other things. I don't intend to try to fix these cars for now. The
way that they are, they are too valuable as test cases for track
quality.
As of March 26, 2009, the layout is pretty reliable if I don't use the four really bad actor cars that obviously need some kind of work. Trains run smoothly with only an occasional derailment as there are now only a few spots with difficulties. These are being identified and fixed as I locate them. The track tweaking is an iterative process. Sometimes a fix in one spot disturbs an adjacent spot. However, the overall improvement in operation since I checked, leveled and glued the whole main line is striking. The glue method is working well too. I have had to lift some sections that were previously glued and it is no problem to lift the track and then glue it again.
All the turnouts are
Atlas. Some of the turnouts seemed to have a systematic derailment
problem when running on the diverging route facing point. On the
really bad ones, the curved point rail is lower than the stock rail
by 10 to 20 mils. This allows a wheel to literally drive over the
point rail. The result looks the same as a "picked" point, but the
actual point wasn't involved. I found that I could slip one or two
slivers of card stock between the throwbar and the point rail to
raise the point rail up a little to match the stock rail and the
problem would simply go away. After inspecting all of them, I found
that several of them had the same problem. The straight point rail
doesn't seem to be an issue, just the curved one.
The resurrection of the layout to it's previous state is pretty much done. There is still lots to do to make it run and look better, but I'm declaring victory for this phase.
[ Top ]
I went rummaging thorough boxes in my garage and found quite a bit of stuff. Besides the hulk of a Revell F7 and a couple of dead Athearn rubber band linkage F7's, a set of streamliners and a bunch of freight cars, I found the powered stuff in the following photos. All of this stuff dates from either the early 90's or the early 60's.
I think that this is either a
Life-Like or Bachmann GP7. It says made in Yugoslavia on the
bottom, I found no other ID. I think that this was the engine in
the set that Richard got from his grandfather in about 1990. He was
just 7 years old at the time.
This engine actually runs reasonably well. It makes about an average amount of noise, tracks well and has average power pickup. It had an NMRA hook and horn on one end and a Kadee on the other but the NMRA coupler has been changed out to a Kadee body mount.
This loco eventually got a DH163 during the eventual DCC conversion. It doesn't run well on the LAMRS club layout because the flanges are too large and it bumps along on the spike heads that hold the code 83 club rail. On the GIRR HO code 100 rail, it does fine.
This
Athearn F7 was sitting on the layout when I cleaned it off. It
didn't run at first as the motor was bound up with dried
lubrication. Some new Aerocar lubricants and a little gentle work
brought the current down from 2 amps to a half amp. The loco
doesn't track very well, but I've been using it in that state to
find bad spots on the layout.
There is a set of 5 streamliners for this loco to pull. The baggage car was missing a coupler for awhile until another Athearn car surrendered it's metal coupler box cover. These streamliners were my first trackwork test train.
This one got a DH163 too. However, at the LAMRS club track voltage, it doesn't run well, it is too slow. I run the GIRR HO at about 16 volts instead of just under 14 volts.
This dockside is
unidentified but it might be a Riverossi. There is no marking on it
to indicate a manufacturer. It doesn't run very well, the power
pickup is marginal at best so that to run at all, it has to be
going pretty fast. It never got Kadee couplers before, probably
because it never ran well enough to deserve them.
However, as a test, this loco was the first to be converted to DCC, as described below, and it tended to run better. The coupler mounts were completely non-standard so I basically knifed them out and rebuilt coupler boxes from Kadee #5 parts. I then installed a pair of Bachmann EZ-Mate couplers and secured the covers on with hot glue. The trip pins needed a little adjustment, but they fit fine and the coupler knuckles are at the right height. The front and rear mounts were a little different so I sort of improvised until I got it right.
The dockside was the first loco I converted to DCC, see below. It has a DH123 jammed into the cab. The loco actually runs MUCH better under DCC than it ever ran on DC.
The Athearn
Santa Fe freight F7 is the same mechanism as the Warbonnet version.
I don't recall ever seeing a Warbonnet like paint scheme in blue
and yellow on any real F7 but I found a photo of a yellowbonnet F9. I don't much like that scheme on an
F unit so I swapped the shell from an older F7 (a dead Revell
model) with a more traditional blue and yellow scheme that also
happened to match my other two ATSF F7's.
It was also all gummed up but it runs ok after some new lubrication. It had a Kadee in the front and a NMRA coupler in the rear but the rear one eventually got changed to a Bachmann EZ-Mate because it fit easily in the snap on plastic coupler box.
After the shell swap, this one also got a DH163.
This SW1500
has a similar mechanism to the F7's so I assume that it is an
Athearn also. It was also gummed up and but runs ok now. It has
Kadee couplers at both ends. This loco is also somewhat sensitive
to less than perfect trackwork so it has been set aside for
now.
The SW1500 also got a DH163 DCC decoder and a new LED headlight. This one was a challenge as there wasn't quite enough room for the wires.
This trolly set is
from Mehano, made in Slovenia. It has only 4 wheel pickup, but it
runs MUCH better than the dockside.
During the initial DCC conversion, I put a DH123 into the trolley, but during testing on the LAMRS trolly line, it didn't have the steady slow speed performance capability that it needed so I swapped the decoder for a DH163 and it ran better.
Last time I was at the LAMRS, the trolly was running for an extended period but somebody saw it starting to smoke. I tested it and I could see arcing through the windows, it had a brush problem. While test running it on the bench, it issued a big puff of smoke and the booster shut down. The motor completely shorted and the decoder died. RIP.
The best runner of the
bunch however is this Mehano
Pacific. It's quiet, smooth and tracks very reliably. These
engines are still made.
This one has a dummy coupler in the front and a Kadee on the tender.
It picks up power on only four of the drivers. The tender also has power pickup to support the rear headlight, 2 wheels on one truck on one rail and 2 wheels on the other truck on the other rail. Eventually I cross jumpered power between the loco and tender so that there was 8 wheel pickup and it ran over turnouts much better.
Over a period of time, the Mehano Pacific got a DH163 installed inside the loco, a Digitrax SFX0416 (for sound and the rear headlight) installed in the tender and a Kadee #6 coupler installed on the pilot. It has been certified to run on the LAMRS club layout. I used it last week at an operation session and it worked well and, except that it is a small loco and doesn't have exceptional pulling power, it ran quite well.
This is one of
the rubber band drive Athearn F7's. It actually runs. I thought
that it had burned up. It took parts from two of them to make one
running one. I still need to locate a brush and brush spring to get
the other one going. These locos can run like a bat out of hell but
they also could run slowly and typically more quietly than gear
drive locos because there are no gears.
This loco seems to be in good condition but it's missing couplers. I pulled a brush and spring out of this loco and put them in the other one, it runs too. Further Athearn, which is 8 miles from my home, says that they have parts for these things.
I also found that I had 3 shells for these two locos, two F7A shells and an F7B shell, all painted to match. The extra F7A shell fits the much newer loco with the funny paint job so it got swapped so that the unit looks similar to the older ones. I figured that the two different types would be badly speed mismatched and they were. The rubber locos run faster.
I did manage to fix the other one too. Athearn no longer has the parts that I needed but a brush for a newer loco is similar to the older brush and I found that I could wind my own spring from 0.020" music wire. The replacement brush needed some "adjustment" to fit but fit it did. Both of these locos speed match fairly well, but they are inconsistent. Their speeds change depending on direction of motion and sometimes one or the other will start to run a little faster or slower. However, they are close enough considering the built in compliance of the rubber drive mechanism.
Installing Kadee couplers on these is not straightforward. A Kadee #5 (or Bachmann EZ-Mate or Intermountain) is too short on the pilot end where there is a dual mounting hole. The long shank Kadee #6 kit barely fits. The mounting hole on the rear end is set far enough inward so that the back of the long shank #6 box will barely fit (with some minor trimming) and still not interfere with the rear truck. On both ends of both units, the mounting pad needs to be filed to make it parallel with the frame so that the coupler comes out to the right height.
For those of
you in the younger set, this is a 50 year old HO locomotive. There
are two shafts leading from the central motor. Over each are looped
two rubber bands which also loop over a spindle on each axle. This
spindle is almost as large as the wheel. Of course, the remains of
the rubber bands that were in there were just hardened globs. I
picked all those out and found some new rubber bands at the local
Dollar Tree store. They were in the hair styling section and cost
me a buck for a bag of 250 of them. Technically, I've actually
spent something on the layout, 1.6 cents.
After 6 months or so, the cheap rubber bands lost strength and tended to stick to themselves and to the spindle and drive shaft. They would get all wound up and the loco would stall. I replaced them with heavier rubber bands that were intended for these locos. We'll see how long those last. They were about $0.30 each.
Both Athearn rubber drive F7's got DCC as well. The A unit got a DH163 and the B unit got a DH123. However, due to the variable brush resistance in the older motors, the DH163 turned out to be a better deal and the B unit got one also.
While rummaging around in
darker corners, I found two Bachmann F7's in Warbonnet. Neither of
them ran. On one, the lights came on but there was nothing from the
motor. The other appears to have a gearing problem. If I remember
correctly, Richard initially used these to pull the streamliners
but they crapped. This is why I bought him the Athearn Warbonnet
F7.
One of the F7's was eventually converted to a dummy and wired to provide 8 more wheels of power pickup for the Athearn Warbonnet F7. They weren't good for anything else.
I also found this road
unit which I think is probably a GP40 because of the three large
fans in the back. It didn't run well either. The front truck was
not being driven.
Well, I found out
why the front truck wasn't running in the GP40. It isn't part of
the drive train. The loco has a small pancake motor on the rear
truck and a high reduction ratio gear train. The drag of the power
pickups on the front truck is so high, even with lubrication, that
the loco will hardly drag itself around. It slips going up a
straight upgrade and even slips going DOWN a curved grade. This is
a true piece of junk and was scrapped.
The F7's are the
same deal, 4 wheel drive on an 8 wheel loco. At least they have a
LITTLE bit more weight on the rear end and will drive themselves up
a grade. The one where the motor wouldn't run had a stuck gear
train. The poor little motor didn't have enough torque to break
itself free. Even at full stall, it only drew a half an amp. These
locos are very nearly worthless. They don't pull and they make an
incredible amount of noise.
At the Nov 2009 train show at
the Los Angeles County South Coast Botanic Garden, there were a
couple of tables set up up selling HO rolling stock really cheap. I
got this Bachmann USRA 0-6-0 for $5, condition unknown. It actually
ran but needed some bits and pieces reattached and some
lubrication. After a little attention, it ran fine.
Eventually, I added power pickups to the tender and cross jumpered power between the engine and tender. I replaced the Bachmann EZ-Mate coupler on the tender with a Kadee #5 which fit well.
The front coupler
was a dummy which is kind of useless for a switcher so it got
changed to a Kadee as well. However, to do it I had to rebuild the
whole structure of the front end as cutting off enough material to
make room for a Kadee box would also detach the whole pilot
assembly from the loco. I built another structure under the
existing one from 0.060" styrene sheet. It was attached to the
bottom of the loco with a #2-56 cap screw. A couple of pieces of
0.100" x 0.125" styrene strip allowed this piece to attach back
to the pilot beam which was still attached to the loco by the original structure. I then hacked and ground out the existing
coupler and cut the orignal pilot assembly away from the loco. This left
just enough room to slide in a Kadee #5 box which is attached to
the underside of the pilot beam with another #2-56 cap screw. The
coupler came out at just the right height.
The O-6-0 got a Digitrax DZ125 decoder stuffed into the little pocket where the smoke unit used to be. This loco was a problem due to motor issues, but I worked around that.
This HO Doodlebug was part of
an estate sale. I got it for $5 too. It ran but there is a gearing
problem resulting in a repeating clunk and the power pickups on the
rear truck are flakey. I sent it back to Bachmann for repair, $20
under the lifetime warranty because it was a "Spectrum" product.
The drawings indicate that it does have a DCC socket inside, I
didn't open it to look. In any event, it'll get DCC once it is
running properly. I'll also letter it for the GIRR.
Bachmann called and said that they didn't have the parts to fix it nor an undecorated Doodlebug so they were going to send me another one painted in some other roadname.
However, when it came back it
didn't look much like a DoodleBug at all. They decided to replace
it with a DCC equipped undecorated SD45. They didn't have ANY
Doodlebugs to replace it with so I got a "better" loco. This one
came mostly as a kit with all the detail parts in little bags to
make painting it easier. I'll probably just paint it weathered
black and letter it for the GIRR, then rust it up some. This is
going to be quite a project as the detail parts are all very small
and fragile. There are no assembly instructions, just an exploded
diagram.
The decoder is a POS, it makes some PWM of noise and is not very programmable. I plugged in a DH123 and it ran much more quietly but I put the Bachmann decoder back in until later.
The locomotive runs well enough and handles my track just fine. It seemed a little "sticky" at low speed so I added some LGB gear lube to the worms on the gear towers and it immediately improved. Slow speed performance was much better and what little gear noise there was virtually vanished.
It didn't take long to
simply paint it weathered black, apply some decals and clear coat
it with a clear matte from a rattle can. All the detail stuff,
steps, grills, grabs and handrails, still need to be added but that
will happen later.
The front headlight came as a loose item. This is because it is assembled into the shell and cannot easily be released. When the shell comes off (remove both coupler pockets to release the outer frame and shell), the wires would normally have to be pulled from the little push on connectors on the PWB. This is a pain so I installed some connectors in the front headlight wires to allow the shell to be more easily released from the frame.
Compare
this picture with the one above. The detail parts have been added.
The bad thing about HO is that detail parts are really tiny and
very hard to deal with. The good part about HO is that the detail
parts are really tiny and hard to see so that if the detail isn't
really good, it doesn't matter much anyway.
I also found a "weathering kit" that had never been opened under the layout. It consists of finely ground pigment and a heat activated adhesive. It is simply brushed on and the heat of rubbing causes the adhesive to bind the powder to a matte surface.
I had already determined that airbrushing over decals to dull them was beyond my skill level by experimenting on test pieces. Brushing a very little of the soot pigment over the decals and brushing lightly faded them just right. A little rust spread around enhanced the effect. I haven't done anything lower than the walkways yet.
I have no idea where I got
this kit. It had a sticker on it saying that it cost $22. The
company, Bragdon
Enterprises is still on line and this kit, FF-65, is still
listed at $22. This stuff is really easy to use a very little of
the stuff goes a long way. The "Weather System" is also available
with other pigments, and in smaller quantities. From the amount I
used on this loco and an entire large scale gondola, I'll never go
through 10% of this material.
I made some
numberboards by drawing them on the computer and then printing them
on plain white paper. A laser printer produced a sharper image but
an ink jet produced a more opaque black. I went with the deeper
black even though it was still not quite deep enough. The paper
tabs were cut out of the sheet and then attached with diluted tacky
glue.
It took many days of work, on and off, to finally install all of the detail parts on the SD45. The only part I did not install was the speedometer cable from the front truck to the chassis. I deemed that that that part could not be easily seen and would get broken off in a hurry during handling while simply putting the loco on the track. I stared at a bag of 6 small slivers of silver plastic before I finally figured out that they were windshield wipers. However, since they would install in locations not easy to touch, I put them on. The last parts to go in were the "bridges" that allow a crewman to cross from one loco to another. I had dropped one of the smaller pieces and lost it on the floor for awhile, but I eventually saw and recovered it.
The SD45 still has the original Bachmann decoder in it but I've ordered a LokSound Select sound decoder for it. It's been on backorder for months though.
I liked the Doodlebug so
much that I started watching them on evilBay. Three of them were
bid up to the mid $30 range before I started bidding. There were
three more due to close within a couple of days. The first was a
Warbonnet version, the other two were the earlier "moustache"
version with a red and yellow front end and a pullman green body. I
wanted one of the green ones but, as a test, I made a $25 bid on
the Warbonnet version and won it for $21.38. Just for grins, I bid
the same $25 on the other two in sequence, they both went above
$35. One more came up, I was outbid in the last 3 seconds. It went
for $26.50.
This one makes a fair amount of gear noise but I eventually figured out how to get the truck apart to lubricate it. After a few minutes of run time it calmed down a little bit. I suspect that this one will take some extended run time to quiet down. The rear truck wasn't picking up power so some adjustments were needed there. I stuck a DH123 in the unit and that worked so it runs on DCC. I suspect that this was purchased by somebody and it ran like crap so it sat on a shelf or in it's box for years. It had never had it's pilot or rear coupler even installed. They were in sealed bags in the box.
After a few days of trying to get the thing to quiet down, I ran it without the shell to see what was making the noise. It didn't make nearly as much noise with the shell off, it sounded pretty normal. Something is interacting with the shell so that the shell is acting like a sounding board. I glued in the window strips better and the noise got less harsh but it still makes more noise with the shell on than off. This will take a little work to sort out.
With the shell on but
the roof off, it got a little quieter yet, so the roof is involved.
There is a plastic strip that presses against the decoder. This
could be transmitting vibration from the motor, up through the
assembly above through the decoder to this strip and then to the
shell and roof. I cut it off with a cutoff wheel and things
improved a little more, but it is still louder with the shell on
than off.
After more messing with the Doodlebug, I found that a window insert on the left side was buzzing against the shell. I used CA to glue the window insert to the shell every inch or so all around and that source of noise went away. I did all the other window inserts as well. Now the noise level with the shell on is only slightly higher than with the shell off. Overall, its a major improvement from where it started. I am assuming that that the noise is what caused the former owner to put this loco back in the box and not run it. It still makes too much noise to qualify for a sound decoder so I'll run it this way for a while to see if it gets better once it is broken in.
I messed with the Doodlebug some more and at least I've found the source(s) of the remaining buzz. Hard squeezing the shell right around the baggage compartment doors makes a difference. If I run the Doodlebug on the bench with the shell off and then use the shell to touch various parts of the loco, I can get the same buzz if ANYTHING in the motor area touches the shell. The strongest buzz came from the motor brush contacts. They MAY have been touching the shell. Bending them out of the way helped but didn't totally solve the problem. It appears that the PWB that is attached to the top of the motor is transmitting motor vibration to the shell as well. Even partially installing the shell causes the problem to return so it isn't a result of the decoder pressing upward as the shell isn't on far enough to reach the decoder. There is very little clearance between the PWB and the inside of the shell so I'll either have to find a way to center it better or to damp that possible interference.
Some of the
wiring on the top of the PWB was interfering with a brace on the
shell that crossed the PWB. I cut much of it away and the noise
level went down again, not all the way, but nearly to the level
with the shell off. At some point, I'm going to reach a practical
minimum.
I worked on this one some more today. The cabin light was wired directly across the track. This is a bad thing as it can mess up programming, although Service Mode programming still worked. Since the Doodlebug doesn't have a reverse headlight, I moved the cabin light to the yellow decoder lead (F0R) and reprogrammed the decoder to have the cabin light on F3 and made it non directional.
In the process of rewiring, I determined that some of the motor noise was conducting through the plastic clip that mounts the back end of the PWB to the motor housing. I removed the clip, which was actually broken anyway. I also removed the stiff cardboard insulator between the motor housing and the bottom of the PWB and insulated the motor with electrical tape. No leads rest on it anyway. The noise level went down another notch. There appears to be just one noise source left as the character of the remaining noise got much "simpler" in nature and intermittent. I still need to find that one last source of conduction from the motor to the shell.
I finally found the remaining noise source. By removing the decoder, I was able to install the roof without an increase in noise. The decoder was touching the roof and conducting noise to it. I used some double back foam tape to hold the decoder to the PWB, which itself is floating supported by it's wires alone, and the thing now runs as quietly with the roof and shell on as with it off. It is still not as quiet as it should be but at least it is as good as it is going to get.
After doing some extended test runs that were going well, I noticed that the headlight was brighter than it had been in the past. I figured that I had better tone it down before it burned out, but I was too late. It blinked out. Apart it came again and a 3 mm white LED with an 820 ohm current limiting resistor went it. Then when I put it back together again, it was noisy. It took a little fiddling with the position of the PWB, but I got it to quiet down again. Hopefully, I won't have to remove the roof again for awhile.
I had such good experience with two
of these Bachmann 2-8-0 locos that I found one on the internet and
bought myself one. Not counting the included crummy Bachmann DCC
decoder, the loco runs quite well. There is a derailment problem in
one spot with a questionable track geometry (vertical gradient) but
the track is fixable. When ESU finally releases a US steam sound
for the LokSound Select decoder, I'll replace the included decoder
with a good sound decoder. The tender is set up to take a speaker
facing downward and an 8 pin NMRA medium connector for a decoder.
Unlike the tenders on the other two locos, this one has some weight
in it.
Instead of fixing the track, I adjusted the drawbar on the loco. When the loco was crossing the gradient, it was lifting the front of the tender. I bent the drawbar a little so that the loco doesn't apply upward force on the tender at that spot yet still provides a secure connection. Problem "fixed."
Both couplers got changed out to Kadee's. This thing is ready to certify at the club.
The loco is waiting for a better decoder. I have another LokSound Select on backorder for it whenever ESU sees fit to ship the steam sound sets.
I rolled the dice one
more time on a Doodlebug I found on evilBay for $19, rated LNIB,
and as expected, it has problems. The motor runs but the doodlebug
does not. It would appear that there is a stripped gear in the
power truck. The seller is going to get some bad feedback.
It may have been "like new in box" if it was never assembled correctly in the first place. It didn't take long to determine that the drive shaft socket was not attached to the gear tower. I pulled out the front truck and pressed the socket back on the gear tower shaft until it clicked in place and then reassembled the Doodlebug. It ran. It also made some noise, but not nearly as much as the Warbonnet Doodlebug did. I properly attached the windows and bent the motor contact tabs so that they did not rest against the shell and the noise level came down to just a little louder than the Warbonnet one makes after working on it for weeks. The rear truck power pickups were bad on this one too. I bent them so that they actually touch the wheels and then the rear truck worked.
I put a Bachmann DCC decoder that came out of an On30 Davenport in the Doodlebug and it ran without a lot more noise than it had without it. There is still some vibration reaching the shell, but it should be possible to fix. The rear coupler has been changed out to a Kadee.
At the LAMRS, I found
one of these Berlyn Inspection Cars in a cabinet. The LAMRS car ran
so well and was so cute that I found one on the internet and bought
it for $89. It arrived on March 30, 2010. I also bought a Digitrax
DZ125 to install in it. This is actually a photo of the LAMRS car,
mine is identical.
This is the inside of
the car. It is a pretty simple mechanism but there is really no
room inside for a decoder. This car was a little stiff out of the
box and drew much more current than the LAMRS car. As I was
breaking it in, it had a problem, the motor speed increased and the
car slowed down. This wasn't good. I took the shell off and didn't
see anything wrong so I ran it that way for a lap or so and then
something went plink and the car stopped dead with the motor
running fast. The brass drive gear on the motor shaft just came
off. Fortunately, I was able to find it and I pressed it back on
the motor shaft with a little CA applied to hold it in place. Then
I ran the car on the bench for awhile and watched the current. It
started off at 250 mA and slowly worked it's way down to 150 mA as
the mechanism slowly increased in speed at a constant 10 volts on
the car. The gearing was tight and was breaking in. After the
current stabilized, I lubricated the gears with gear grease and the
bearings with a light AeroCar oil.
Then the decoder
went under the car. It barely fits, even as small as it is. The
resistor is there to protect the decoder from issues that the motor
may have. It is about 3.2Ω even though the banding indicates
that it is 2.7Ω. The DZ125 is rated at 1 amp but it uses very
small parts with little thermal reserve. The resistor is there to
help protect it some from any motor mis-behavior even though my testing didn't
indicate a problem. The car will run plenty fast enough with the
resistor and it only drops a half a volt or so anyway at normal
current levels.
The power pickup on this car is not quite as steady as the one that the club has, the problem is in the wipers. It'll take some more breaking and perhaps some adjustments to get it running as reliably as the club car.
After running it at the club and doing tests on my own layout, there is a problem. The decoder seems to upset due to short power interruptions and become catatonic. Most often, it will stop and fail to restart. This looks like it lost power but I installed a small bulb across the power pickups and it is getting power. I need to power cycle the track or lift one side of the car off the track to cause the decoder to reset properly and then it picks up and scampers off. Occasionally, it will upset while running and just keep going, out of control. Same solution, power cycle and it gets well. This is a decoder problem so I'm probably have to send the decoder back to Digitrax for replacement. The USRA 0-6-0 has a similar Z scale decoder in it, a DZ143, and it has none of these issues.
Last night I determined that the decoder was not right, either because that is the way that this kind of decoder works or because this one is defective. It got pulled and went out in the mail to Digitrax this morning for replacement. I put the car back together wired for DC and tested it on the layout again. It has indeed broken in a little and now it runs as smoothly across my insulated frogs as the club car did. If the replaced decoder still has problems, I'll try and NCE or TCS Z scale decoder and use this one some less demanding application in the future.
Digitrax replaced the decoder and I test installed it and it seemed to be much better. However, the brass drive gear fell off again and when I tried to reattach it with a dab of LocTite 222, I apparently got some in the drive shaft bearing as well and the thing locked up. I am working on getting the stuff out of there but it may take some time. I had it running again, but the shaft still isn't completely free. The drive gear has also fallen off again and I have to determine how to get it to stay in place before I can proceed to clean out the degraded bearing.
After working with the inspection car for quite a while, I got the bearing free enough to use. The current is still a little high, but the motor can handle it without overheating. I assume that it will continue to improve as it wears in. I did fix the gear by epoxying it in place and it seems to be holding. I temporarily wired the replacement DZ125 to it again and it behaves better, but it is still having some of the same problem. I may have to try another kind of decoder and use the DZ125 in some less stressful application. After removing the decoder, it actually ran better, although it is still having power pickup issues on some, but not all, of my insulated frogs. I probably should add power pickups to the trailer. It would not be hard to do.
At the swap meet and
Open House at LAMRS today, this model of the Blue Goose followed me
home because it didn't sell to the public. The Blue Goose (ATSF
3460) was the first in this class and the only one to get
streamlined. The real Blue Goose was an oil burner, this one has a
coal tender. The model has a DH121 in it and runs well enough but
the club didn't want it because of the prototype and that the power
pickup seemed a little flakey on the club layout. The members that
tested it gave it the thumbs down and the club elected to sell it.
I bought it from the club. It does have some derailment problems,
especially with the tender, but I think that it because the tender
is WAY too light. I'll drop some weight in it later to see what
happens.
I had some derailment problem with the tender of the Blue Goose so I added some weight and that seems to have fixed it. It runs pretty well now.
After adjusting the coupler heights on all the streamliner cars (they were all over the place, usually high) and on the Blue Goose (low), I was able to pull the five streamliners on my 5% climbing 18" radius curves. The Blue Goose struggled a little but it didn't loose traction as it has traction tires.
The Blue Goose had a DH121 decoder in it the way that it came. This is not a great decoder so it will eventually get changed out, probably to a DH163 or maybe even a LokSound Select sound decoder if it runs well enough in the long run to deserve one.
The headlight on the Blue Goose wasn't right either, it was on all the time and the decoder would not respond properly to the programming track. I opened up and found that the bulb was in a socket that was attached to a bracket that was bolted to the frame, at least when the shell was on. The white decoder wire was electrically connected to the frame which is a big no-no. It shorted out the decoder's F0f function output and was freaking out the decoder. I elected to chuck the incandescent bulb and replace it with a 3 mm warm white LED. I drilled out the metal headlight bezel to 1/8", just right for a 3 mm LED. Now the LED lens sticks right through the bezel and it at the right spot to represent the headlight. It is run though a 510Ω resistor so that it runs at just about 18 mA at the club track voltage and about 22 mA at my track voltage. It is quite bright.
The DH121 was physically damaged so it got replaced with a DH163. The loco now responds to the programming track and the headlight is also controllable. However, the power pickup needs some work. I'll add two wheels to the tender as the center axle doesn't contribute right now and two more wheels to the loco as the rear truck wheels don't contribute either.
It turned
out that adding functional pickups was a trivial modification, it
took about 10 minutes to do all three trucks.
This is one of the tender trucks. There is a contact strap that brushes on two axles and connects to the tender body at the truck pivot. One wheel on each axle is insulated. The center axle wheelset was identical so all I had to do was bend a piece of 12 mil phosphor bronze wire over the truck frame so that it rode against the center axle and one of the end axles. It stays in place by itself, but I put a drop of CA on it anyway. The spring pressure is fairly light so that the drag of the wheels isn't a lot higher than it was before. The second tender truck was done identically.
The rear
loco truck was done similarly. A single piece of phosphor bronze
wire clips over the truck frame and presses against both axles.
These axles were also set up with one wheel (the correct one)
insulated. I then soldered a short piece of flexible wire to the
contact wire and terminated it to the loco frame under the screw
that holds the truck assembly on (to the right in the photo). The
frame is connected to one rail via the pilot truck and two of the
drivers. The center driver still does not collect power.
The difference was remarkable. Before, the headlight flickered quite a bit and the loco stalled in places. After, not even a flicker.
However, the increase in drag was enough to cause the loco to stall with drivers slipping on the worst part of my layout with five streamliners in tow. It now needs a little more weight, maybe as much as four ounces.
I added all the weight I could fit, 2 oz, and it helped but the loco still slips at the worst spot. I guess that this is the way that it is. I turned the train so that it doesn't hit the worst grade going uphill.
The front and rear drivers on one side have traction tires. The rear one was loose enough so that, under load, it would spin off the wheel. I tacked it back on with a couple of very small dabs of CA, but that just created hard spots and the thing got a little bumpy.
At the Big Train Show in Ontario yesterday, I picked up a jar of Bullfrog Snot to try to replace the balky traction tire. This is a rubber like material that is applied to a wheel wet and allowed to cure while the wheel is slowly turning such that it forms a sticky tire surface on a wheel that would not normally have a traction tire. Since I had determined that Riverossi replacement parts are simply not available, I am trying to use this stuff to make a new tire. Running without the tire is not a good plan because then the driver diameter is wrong. To replace a missing tire, several layers of the stuff have to be applied to build up a thickness that replicates the original tire. I am partly done with the first attempt. If it doesn't work, the stuff can be peeled off and reapplied.
The Bullfrog Snot appears to work. It leaves a tough, translucent green film on a treated wheel. With enough coats, about four, it can fill the groove left behind by a failed traction tire and replace the function of the tire. It does improve traction, but it also needs weight to keep the wheels on the track or the loco will start to hop. It is clear that it helped, but the loco still needs a little more weight to get by the very worst spot. All I need to do is lightly touch a sand dome and it will grip and go.
[ Top ]
The rolling stock dates from about 1960 at the earliest to about 1995. Some of it was part of a train set, some was clearly purchased as individual cars. Some runs well, some does not, some are not even complete with broken or missing parts. Some have been converted to Kadee (or Intermountain) knuckle couplers, some still have the original NMRA hook and horn couplers.
There are
several cars with sprung trucks. Digging back in dusty recesses of
my memory, I think that some of these are Revell cars that came in
my very first train set. Three of these cars track very poorly,
they seem to pick every point on the layout. I think that the
springs allow the trucks to flex so much that a wheel can actually
walk up on a point more easily than others. I cannot feel the
points as these cars roll through, but a wheel still comes up.
Eventually, these trucks will probably have to be replaced or
superglued into rigid trucks.
It's is generally not a good idea to install a turnout on a grade but it can work if the grade through the turnout is constant and the turnout is level. I have one turnout on a grade and this one is the primary troublemaker.
Eventually, I tried the Superglue treatment on the car in the photo and it didn't help. However, replacing the trucks with Athearn ATH90400 Bettendorff trucks with metal wheels completely fixed the B&M boxcar. It would hardly roll at all before and derailed on most turnouts. With the replacement truck, it rolls VERY well and passes all turnouts without difficulty. Conclusion: the old trucks are AFU.
Many of the older cars also roll poorly even though they track well enough. There really isn't much to a train car except for the trucks. When there are tracking or friction problems, the problem is virtually always due to the truck design or construction. Some of these old trucks actually have friction bearings, not needle bearings. No amount of work or lubrication is going to help these. It looks like some wholesale replacement of trucks is in order at some point.
As it eventually worked out, EVERY car was recoverable. They all roll and track well although some still have plastic wheels. They all have knuckle couplers, most are Kadee but there are some Intermountain and Bachmann EZ-Mate couplers in the mix. All the freight cars have body mounted couplers. The passenger cars are too long to tolerate body mounts on my trackwork (S curves in a few places) so they still have truck mounts.
The cars are more or less grouped together by either type or status in the following table. I'm mostly bothering to do this for my own documentation purposes. If the status indicates that it got new Athearn trucks, it has metal wheels.
| Car | Roadname | Qty | Mfgr | Photo | Status |
|---|---|---|---|---|---|
| Streamliners | |||||
| Baggage | Santa Fe | 1 | Athearn |  |
1990 vintage, plastic truck frames truck mounted Kadee couplers runs well |
| RPO | Santa Fe | 1 | Athearn |  |
1960 vintage, metal truck frames, lights work truck mounted Kadee couplers runs well |
| Coach | Santa Fe | 1 | Athearn |  |
1960 vintage, metal truck frames, lights work truck mounted Kadee couplers runs well |
| Diner | Santa Fe | 1 | Athearn |  |
1990 vintage, plastic truck frames truck mounted Kadee couplers runs well |
| Observation | Santa Fe | 1 | Athearn |  |
1990 vintage, plastic truck frames truck mounted Kadee couplers runs well |
| RPO | Southern Pacific | 1 | Athearn |  |
Blue box vintage, plastic truck frames, no lights. Purchased at
the Great Train Show, Long Beach, 13 Feb 10 for $8. Had NRMA couplers, got Bachmann EZ-Mate couplers immediately runs well |
| Good Runners with Body Mounted Kadee Couplers | |||||
| Steel Caboose | ATSF | 1 | Athearn |  |
I bought this car for $5 at a train swap meet and assembled it the same day. It got a set of Intermountain couplers. Then less than an hour later as I was rummaging through Richard's boxes, I found ANOTHER IDENTICAL one, unbuilt, still in it's shrink wrapped box. |
| Wood Reefer | Armour private car | 1 | Mantua | 
|
The trucks have replaced with modified Athearn trucks and it runs fine. The trucks required the mounting hole opened to 0.150" to fit on the car. It had 2 oz of extra weight, this has been removed. |
| 3 Bay Hopper | unlettered | 1 | unknown | 
|
This car has a die cast frame, but it is very light. It may just need some weight. However, with the trucks replaced with Athearn trucks, it runs fine, low weight and all |
| 40' Flatcar | Union Pacific | 1 | Athearn |  |
This car also had sprung trucks and tended to derail. It got new Athearn trucks and some Kadee couplers, now it is fine. |
| 40' Boxcar | Pennsylvania | 1 | unknown | 
|
This car had sprung trucks and derailed on nearly every turnout when on the diverging route. The trucks got replaced with new Athearn trucks, now is is fine.... but still with missing doors. |
| 40' Boxcar | Boston and Maine | 1 | Athearn | 
|
This car had sprung trucks and derailed on nearly every turnout when on the diverging route. This is a nice looking car. It deserved recovery. A set of new trucks, Athearn ATH90400, completely fixed this car. |
| 50' Boxcar | ATSF | 4 | Athearn |  |
One in the "map" style, one in the "Chief" style, one wide door version in the large herald version and one plug door "Shock Control" version. These cars ran well when the track was in very bad condition. They all run on stock trucks with plastic wheels. |
| 40' Boxcar | ICS | 5 | Athearn |  |
These cars are identically painted except for the road numbers. They are sensitive to imperfect track but run well when the track is in good condition. They currently run on stock trucks and plastic wheels. |
| 4 Bay Hopper | Western Maryland | 1 | Athearn |  |
This car ran well on bad track, but it had a problem. The brake rigging at one end would drag on the wheel flanges in a turn. Some file work took care of that. This car runs on plastic wheels. |
| 3 Bay Hopper | Virginian | 1 | unknown |  |
This car has the same diecast frame as the unlettered hopper above, but it runs a lot better. It handled bad track well while the unlettered one can't handle good track. It runs on plastic wheels. |
| 30' Flatcar | Baltimore and Ohio | 1 | Accurail? | 
|
The trucks say Accurail but they may be replacement trucks. The car runs on plastic wheels. |
| Short Gondola | Union Pacific | 1 | unknown |  |
This car is REALLY light, but it runs as long as it isn't near the front of a train. It runs on plastic wheels. It's trucks also are tight by the frame and the flanges of the end axles were dragging on some mold marks on the frame in some vertical gradients. These got ground off and the wheels no longer drag. |
| Gondola | Southern Pacific | 1 | Crown |  |
This car also has sprung trucks but of a different and clearly newer variety than the Revell ones. It got new Athearn trucks and Kadee couplers. |
| Caboose | ATSF | 1 | Revell |  |
This car had the same sprung trucks and very high drag. The trucks got replaced with Athearn reefer trucks (all that the carshop crew had handy) and body mounted Kadee #5 couplers. Now it rolls and tracks well. |
| Wood Coach | D&RGW | 1 | Athearn |  |
This is the only coach of this kind on the railroad. I don't know where it came from. It runs on the original plastic wheels. |
| 3 Dome Tank Car | Mobilgas private car | 1 | Crown | 
|
This car had sprung trucks similar to the SP gon and ran poorly. The trucks were changed out to Athearn trucks and the couplers changed out to Kadee. This was the last one with stock body mounted couplers. All the rest were truck mounted. |
| Note: All of the cars below, to the next section have been converted from truck mounted NMRA couplers to body mounted Kadee couplers and have received new wheels or trucks as appropriate to solve their drag and/or derailment problems. | |||||
| 2 Bay Operating Hopper | Pennsylvania | 1 | unknown | 
|
This car has a metal frame and operating hopper doors. I remember these cars, I had an electric actuator that opened the hopper doors and allowed the load to be dumped. This one got new Athearn trucks. |
| 2 Bay Operating Hopper | Virginian | 1 | unknown |  |
This is basically the same car as the one above in different paint. It also got new Athearn trucks. |
| 4 Bay Hopper | Rock Island | 2 | Bachmann |  |
These are pretty ordinary cars. Both still run on the original Bachmann plastic trucks and wheels. |
| 4 Bay Hopper | Burlington Northern | 1 | Bachmann | 
|
This is the same car as the Rock Island ones, just in different paint. It also runs on the stock wheels and trucks. |
| Gondola | Burlington | 1 | Bachmann |  |
Another ordinary car with truck mounted couplers. It runs on the Bachmann trucks, but with Proto 2000 metal wheels. |
| Gondola | New York Central System, P&LE | 1 | Bachmann |  |
It's pretty light but it runs ok. The trucks look very Bachmann like but with odd couplers. It runs on the original trucks and Proto 2000 metal wheels. |
| Caboose | ATSF | 1 | Bachmann |  |
This car runs on the original plastic trucks and wheels. |
| Stock Car | Katy | 1 | Life-Like |  |
This car carries 2 oz of lead weight inside and had very high drag, it would not even roll downhill. It got new Athearn trucks but the extra weight is still there. |
| 50' Plug Door Boxcar | Milwaukee | 2 | Bachmann | 
|
Both have the original Bachmann wheels and trucks. One had 2 oz of extra weight inside, that has been removed. |
| Gondola | Soo | 1 | Bachmann |  |
Pretty much a standard Bachmann gondola. It still runs on the original Bachmann wheels and trucks. |
| Gondola | Union Pacific | 1 | Bachmann |  |
It still runs on the original Bachmann wheels and trucks. |
| Reefer | Pacific Fruit Growers private car | 1 | Revell |  |
This one had sprung trucks too and had very high drag. It then got old Athearn trucks with an old set of metal wheels, there is an obvious insulator on one side of each wheelset. However, it still derailed so it got the set of Proto 2000 wheels that the Shell single dome tank had because it is going to get new trucks. Now it seems to run reliably. |
| 50' Plug Door Boxcar | Burlington Northern | 1 | Bachmann | 
|
New Athearn trucks fixed the dragging. |
| Single Dome Tank Car | Shell private car | 1 | Bachmann | 
|
It got new Athearn trucks but already had Kadee couplers. |
| 3 Dome Tank Car | Cyanamid private car | 1 | Bachmann | 
|
New Athearn trucks and Kadee couplers. |
| Gondola | Burlington Northern | 1 | Bachmann |  |
New Athearn trucks and Kadee couplers. The frame was also broken and wouldn't stay in by itself, Zap-A-Gap fixed that. |
| Track Cleaner | M of W | 1 | Revell |  |
This is also a 50 year old car. There are two abrasive disks that drag on the track and make LOTS of noise. The mechanism that lifts the track cleaning blocks was broken and was fixed. Modified Athearn trucks and body mounted Kadee couplers were installed. Now this car roll further than any of them by several feet, perhaps due to it's weight. |
| 4 bay hopper | ATSF 183091 | 1 | Athearn |  |
purchased from LAMRS, plastic wheels and couplers |
| 4 bay hopper | ATSF 183080 | 1 | Athearn |  |
purchased from LAMRS, plastic wheels and couplers |
| 2 bay hopper | ATSF 180517 | 1 | Athearn |  |
purchased from LAMRS, plastic wheels and couplers |
Richard also has some unbuilt kits, still in their original boxes.
There were lots of cars that ran poorly, some were mine from 50 years ago, some were Richard's from 15 years ago. There were a variety of manufacturers and just as wide of a variety of truck mounts. Some already had wheels with an NMRA looking bearing, some (mostly Bachmann) had a friction bearing with a little taper at the end and some were straight friction bearings. Some were sprung, some metal, some with metal bolsters and plastic sideframes and some were all plastic. Many had truck mounted couplers and if the truck was retained, the truck mount was hacked off.
If a car had drag or derailed, I preferred just to change out the trucks with Athearn Bettendorff trucks which are inexpensive ($3/car) and seem to work quite well. However, these wouldn't fit all the cars so some cars got new wheels (Proto 2000) which were actually more expensive ($1/axle) than an Athearn truck set. Where new wheels didn't fix it and I could not modify an Athearn truck to fit, I simply hacked off the old trucks and parts of the truck mount, filled in what was left with Alumilite resin and drilled and tapped a new screw hole in the resin for a new screw and an Athearn truck.
After a car got new trucks and/or wheels and then tested good in a roll test down a curved grade with a facing point curved turnout (my worst derailment spot), then it got Kadee #5 body mount couplers. I built up a pad using strip styrene and CA to the right height (different on nearly every car) and attached the Kadee coupler box with a single #2-56 screw in a hole that was drilled and tapped to fit. After a check for coupler height and adjustments to the trip pin or mount as necessary, the car got a final roll test and was declared good. The very best roller was a Bachmann gondola with the original plastic wheels and trucks, go figure.
After the first pass at rebuilding, all the modified cars were strung into a train and run. Anything that derailed at all got pulled from the train, even if it's derailment was by association with the actual bad car. After all I had left was a train full of cars that did not derail, I took them all off the layout and built a train of all the suspect cars. Repeating this process left me with two cars, the Shell single dome tank car and the PFE reefer that still derailed and need more work. The PFE reefer got the Proto 2000 wheels from the Shell tank which fixed it. The Shell tank got moved to the RIP track.
Eventually, the Shell single dome tank got new trucks too and it rejoined the active roster. Then all the cars got weighed and if they were underweight per the NMRA standard, they got some kind of weight to bring them up to spec. Open hopper or gondola cars got loads of DG or black #10 crushed rock to bring them up to weight then the load was soaked in a glue/water mixture to bond it in place. If a "full" load was too heavy, I rolled up some tissue paper and used it as a filler and filled around it before gluing the load. The flat cars got some kind of load attached, usually roughly HO scale vehicles. Closed cars got lead fishing weights or 1/4 oz stick on weights installed inside to bring them up to spec.
The layout was basically
functional but all the turnouts were manually controlled and some
didn't even have motors attached. What motors were there weren't
wired. As part of the resurrection I had identified all the track
blocking and added two new blocks. Now all but one of the Atlas
Selector switches are wired. The control panel had a row of Atlas
Switch Control panels, but I didn't intend to use these so I took
them off the panel.
In a blank section of
the panel, I drew a cheap and dirty track diagram with space for
the 8 turnout control switches that drive the interlocking turnout
control system that I built for this layout. There are actually 14
routes. Six of the switches select two routes depending on which
direction they are thrown. Two of them are wired for operation in
only one direction. These are DPDT (on)-off-(on) switches,
momentary center off.
The diode matrix itself
routes power from the switches to the 15 turnout motors. I prefer
hard wired routing controls because, once designed properly, the
logic doesn't need changing and a hard wired matrix can never
succumb to memory corruption of more modern routing systems. These
things work, they are cheap and very reliable.
This is the diagram of the diode matrix that was
built following the design process at my Interlocking Turnout Control
Tips page. The Atlas motors are twin coil machines so that they
are wired with a pair of diodes each to make them act like single
coil bipolar machines.
Atlas motors have much lower resistance than LGB or Aristo turnout motors. An Atlas coil is about 5Ω. An LGB coil is about 22Ω so an LGB coil draws quite a bit less current than an Atlas motor. The Aristo motor draws even less current. I initially used the LGB power pack as a source of 15 VAC and it worked as I was building up and testing the matrix but as I added more motors, things stopped working. The matrix powers up to 5 motors at one time. The LGB pack could barely handle 2, sometimes 3 motors. By test, the AC output was maybe good for 1 amp and even then it sagged to less than 10 volts under that load. I tried running the matrix from the AC output of an MRC 9500, but it was obviously hurting although things started to work. The MRC is rated at 30 VA total, I assume that I can't get all 30 VA on one output even if the others are unloaded. The meter lights dimmed significantly when it was trying to power the motors.
I ordered a 2.5 amp, 16 VAC (40 VA) wall wart for $5 from Hosfelt which turned out to work well enough. It gets dragged down from 18 VAC at no load to 12 VAC at the worst load, but it throws the turnouts.
Because this system runs from half wave rectified AC, it will cause the motors to buzz when it is active. This is normal and provides a fine indictor that something is happening. If there is a problem in the matrix, a short or miswired diode for example, it won't work properly but it also won't burn up. What will burn up the diodes is a short at a turnout motor.
I used this MRC 9500 power pack
on the GIRR Mountain Division layout for a couple of years until I
put in the Train Engineer walk around throttle. Initially, I ran
the TE from the fixed DC output of the MRC but I eventually
upgraded the power supply to a Crest Elite 13 amp power supply.
This power supply also powers the DCC system that is there now. The
MRC became surplus and sat for years before I dragged it home to
put on the HO layout. The meters are handy to determine what is
going on and the large, nice feeling knob is easier to use than the
LGB. The MRC is wired to Cab B. Cab A is currently being run from a
Lionel DC power pack that came in a James the Red Engine large
scale set. This makes it easy to switch the whole layout from DC to
DCC. All of the locos are set so that they do not analog convert.
When DC is on the rails, they just sit there, sometimes with lights
burning. I turn off the blocks that contain locos with lights on so
that the metering on the power pack can be used to measure the
current consumption of DC locos. The meters are fairly accurate, I
checked them with a DC voltmeter and a 10Ω load resistor on
the rails. At high rail current, the voltmeter on the power pack
reads a little high due to voltage drop leading to the rails, but
at low loads, it is virtually dead on.
Once I get all the wiring going properly I'll need to start thinking about some scenery. Richard started on it, that much I'll leave in place. There isn't a lot of room for scenery and structures but I think that everything that I need is already there. I have heavy weight paper towels, Hydrocal plaster, texturing materials and some HO scale trees. There are lots of buildings stored in boxes, mostly plastic kits, some dating from 1960 or so which are, amazingly, in fair to good condition.
Track ballasting won't happen for some time. I want to make sure that I've found and fixed all the trouble spots. Richard left a small baggie of HO scale gray ballast but it won't be enough. I'll probably used some colored fine chicken grit. I have about 40 pounds of it in dark gray sitting in a bag under the layout. I used it for making an asphalt road on the outdoor GIRR. It might be a little large for HO scale, but not by a lot and there is plenty of it there. This stuff costs $20 for an 80 lb bag.
I use DCC as the primary means of running trains on the large scale outdoor GIRR and the large scale indoor GIRR Mountain Division. DCC actually was developed in the HO world (in a variety of forms) dating from the 1960's. One form (Lenz's) was standardized as "NMRA Compliant DCC" in the early 1990's. It migrated downward in scale to N and Z as the decoders got smaller, and upward to O and G (or F or whatever it is) as the decoders got bigger. I had never used DCC in the smaller scales but I have converted almost 40 locos in large scale.
DCC really isn't necessary
on a 4x8 HO layout, especially one with only one operator, but I
wired the DCC signal from my garage test track over to the HO
layout, just a few feet away, and installed a DPDT switch to switch
Cab B between DCC and the Lionel power pack. I did this so that I
could run analog locos on address 00 so that I could walk around
the other side of the layout and use my radio throttles to control
locos and trains as I was testing for derailments. It is a real
pain in the backside to have to work on one side of the layout
while the cabs are on the other side. This worked out well but the
locos make the typical DCC singing sound and I don't like to run
them that way for very long.
However, conversion of the layout to DCC (which required just the addition of the switch and about 10 feet of two conductor wire) was the first step down a VERY slippery slope. I got to wondering how the locos would work relative to their DC performance under DCC. These are dangerous thoughts....
I had an HO sized
Digitrax DH123 decoder in my parts box. It had been in the Bachmann Railtruck
for a time before the ESU sound decoder was
installed in it's place. I decided to install it in one of the HO
locos to see how it worked. I selected a loco to do the conversion
based on the ease of the conversion and the ease of reversing the
conversion later if necessary. The 0-4-0 Dockside was the likely
candidate but it was also the least reliable runner of the bunch. I
figured that it would point out any problems most clearly. The
conversion was indeed quite easy as the loco had a can motor with
easily isolatable motor leads. There was just enough room in the
cab behind the can motor for the decoder to fit.
As it turns out, the loco runs BETTER with the decoder in it than it ran from DC. I was not sure why this was but it was clearly more controllable especially at low speed and seemed to run by power pickup problems better than with DC. It also analog converted well. This was NOT good news. I could find myself buying a bunch of HO decoders. This is what I ended up doing. DCC did indeed make the layout, and every loco, run better and it made the layout easier to use.
After I had converted all the locos and run them for awhile, I did conclude why the poorest runners tended to work better on DCC than DC. My issue was the motor brushes, and to a lesser extent, the power pickup. These are all old locos and many of them use open frame motors with lots of miles on them. Newer locos with better motors usually don't have brush problems.
In locos that have poor brushes or contacts, the problem is usually not an actual open circuit, but variable resistance in the contact or brush or both. DCC puts high voltage on the track and also places high voltage pulses on the motors. This makes the impact of variable resistance less significant at slow speeds where, on DC, the track and motor voltage would be low. Further, in the case of motor brush resistance, the back EMF (BEMF) capability of some better decoders can sense the motor speed and compensate the pulse width of the motor drive to partially overcome an increase in brush resistance. The rubber drive Athearn locos presented a striking difference in overall performance when converted from a non-BEMF DH123 decoder (which in itself was much better than DC) to a similar but BEMF equipped DH163 decoder. With the BEMF decoder, they either run or they don't run at all. When they don't run, a drop of "conductive" oil on the commutator of the motor will bring them immediately back to life.
Another complication is that ALL of the locos that were originally on the layout were built before DCC became very popular. NONE of them are built with DCC in mind. Most of them have one side of the motor fairly well attached to the frame. There may be some careful mechanical work needed to isolate these motors.
My programming
track for large scale was a 3' piece of track wired back to the
command station. The command station and booster were installed in
a cabinet on the back of the house near the large scale layout. I
used to keep the programming track in the control cabinet when I
was not using it. When I needed it, I pulled it out and set it on
the short brick wall near the cabinet. To program the HO locos, I
simply set a short piece of HO track, which I picked up off the
garage floor, onto the large scale track. Eventually, I glued it in
place but I left the clip leads as I can then move them over to the
large scale running track to run quick tests on the results of
programming.
I normally run my large scale track at 22 volts. This is pretty hot by HO standards but this is not a big problem. There is a switch on my large scale booster that allows me to switch my track voltage between the "normal" setting of 16 volts and the "adjust" setting which is turned all the way up. The one remaining issue is that the booster will source up to 20 amps before it shuts down. It is true that there is a lot of wire between the booster and the end the of the large scale track in the garage, but I've managed the voltage drops such that a short at the end of the line will STILL trip the booster. The additional wiring to the HO layout will probably present sufficient resistance to prevent the booster from tripping so I should probably add a circuit breaker module with a much lower trip point in the feed to the HO layout. This is to prevent a derailment from resulting in smoke in a hurry.
Since it was basically
unsafe to run the layout with so much current available I bought
and installed some electronic circuit breakers. The smaller board
is a Digital Specialities PSX circuit breaker. It is installed as
delivered with about a 3 amp trip point. The larger board is a
PSX-AR. This is the same circuit breaker circuit but with an
autoreversing module added. This drives the Hereford reverse
cutoff. Both run from the DCC coming into the layout. The PSX is
wired in before the Cab B DC/DCC switch.
Both circuit breaker boards are set up to drive a Sonalert annunciator to indicate a fault. I found some generic annunciators with the right lead spacing at All Electronics for $1.25 each so I bought and installed some. The PSX boards are not marked for annunciator polarity, but the square pad is positive.
The autoreversing
circuit breaker is also driven from the DCC signal that enters the
layout. However, it's output is wired after the Atlas Controller
switch assembly so that it ignores which cab is selected on the
Controller.
Once I moved the "8 amp" DCS100 command station to the garage to drive the HO layout directly, this became less of an issue, but the circuit breakers are still there, set to 3 amps, to protect the equipment.
The large scale layout uses a larger booster so that there were no other changes required when I moved the command station. As before, the booster gets it's input signal from the RailSync pair on the Digitrax LocoNet. Now the signal just comes from further away. This allowed me to run the HO off the command station's internal booster and the large scale layout off the external booster so that each could have their own unique track voltage.
I found some nice panel meters at
an on line store and added them to the panel. These are straight up
DC meters, but they have been wired to properly read out the RMS
value of the DCC track power entering the layout. They won't
function when the layout is running on DC, but the MRC power pack
has meters already.
This is the configuration of the
metering. Two bridge rectifiers are used so that the current
flowing in the meters is DC and represents the true RMS value of
the DCC signal. This trick only works with flat top waveforms like
DCC but it does work.
One bridge rectifier converts the DCC signal to DC to drive a voltmeter. The rectifier drops about 1.5 volts so that it reads 1.5 volts lower than the DCC power coming in. The other bridge rectifier, a high current one good for at least the maximum expected track current, routes the DCC signal to the track without modifying it, BUT the DCC signal is converted to DC as it runs through the DC meter. The bridge rectifier has the same 1.5 volt drop as the bridge on the voltmeter so that the voltmeter reads something similar to the actual track voltage. The readings taken with these meters have been checked against a true RMS 4 quadrant multiplier type instrument and they read close enough.
After all the
infrastructure changes were made to the layout, the second loco to
get converted was an Athearn "classic" loco. These were not
specifically designed to accept DCC, but Digitrax has made a
"plug-n-play" kit for this very common loco. In the front is an F7
warbonnet loco after the kit was installed. In back is an
unconverted freight F7. The kit isn't really plug and play because
it takes a fair amount of disassembly and the instructions are not
really clear, but at least it doesn't require any soldering.
The motor is normally connected via clips that go over the motor frame, top and bottom. The top one is extended via another longer clip to touch some contacts that come up from the trucks. This longer clip is discarded and the shorter motor clips are discarded and replaced with ones that are delivered soldered to the decoder motor leads. The whole motor must be pulled from the frame to access the lower clip.
The motor is mounted on two soft plastic insulators and can simply be pried out. Digitrax supplies two additional plastic insulators in case that the original ones are damaged during installation. I managed to get the motor in and out without requiring the spares. When the motor clips are taken off, a brush spring is free to fly away. It must be carefully captured. The original lower motor clip has a couple of fingers that contact the frame, once it is removed and replaced, the motor is isolated. The decoder is provided with push on clips to connect to the truck contacts, the headlight and the headlight post. The headlight is not powered from the blue wire, but from one rail instead. This is acceptable. The light runs at 50% duty factor but it is still bright enough.
It is not really necessary to spend the extra $4 for this version of the DH163 as all the parts are there to make the conversion. The long clip is still discarded. The upper clip can be reused by soldering the orange wire to it (after removal from the motor to prevent melting the motor insulators). The lower clip can be modified to remove or bend the contact tabs and the gray wire soldered to it. The black wire can be soldered to the headlight bracket (frame or one side power pickup) at the front and the red wire can be soldered to one truck contact tab and then extended and soldered to the other truck contact tab. This connects to the other rail. If the stock headlight is used, then the white wire can be soldered to it's tab. The only thing handy to have from the kit is a replacement plastic motor mount in case one gets damaged in the installation. However, not all the "classic" Athearn locos use the same plastic motor mount. The SW1500 has a different design.
The decoder
itself is mounted to the rear of the weight with some foam tape.
The wiring has been tucked away under the weight. Later, a cable
tie was added to hold the decoder up against the weight.
I do have some problems with the loco lugging where it did not used to lug. This appears to be due to voltage drop in DCC wiring. There is just too much stuff between the booster which is in the backyard to the garage. If I start at 16 volts at the booster, I lose up to 3 to 4 volts in all the wiring. Most of the loss is within the garage itself as the end of the line just outside the door only drops 0.1 volts.
If I turn the DCC system voltage back to 22 volts, the Warbonnet loco runs fine, there is about 18 volts on the track which is enough. But the higher voltage when nothing is running will burn out lights. I worked over the headlights in the F7 to replace the incandescent bulb with two LEDs because I'd like to have a Mars light too but I'm still fitting everything in.
I moved my DCC command station from the control cabinet in the backyard to the garage and connected it back to the LocoNet and all runs as it did before. Then I put an adequate power supply (Loy's Toys Fuel Tank) on the DCS100 command station and wired it's booster (which was unused) to the HO layout. Now I can set the large scale track voltage separately from the HO track voltage and also avoid some of the voltage drops. Now I get a pretty steady 16 volts on the track which is plenty to run the F7 with it's fairly heavy load of five streamliners on my sharp curves and steep grades. Most newer HO equipment will run fine on 12 to 14 volts, but the older stuff that I have won't break 20 smph at those low voltages. I need a full 16 VRMS on the track to get reasonable speeds, at least 40 smph anyway.
After all the
DCC stuff was working again, I attacked the headlights in the F7. I
used two 3 mm white LEDs. The upper one is a warm white LED, I
would prefer a hard white one but I didn't have one in 3 mm. The
lower one is a hard blue LED, but it is an old one and it is not
very bright. Both will eventually be changed out with the newer and
brighter 3 mm white LEDs.
I cut the metal headlight bracket short and reformed it so that it was lower so that it didn't block the lower headlight position. The lower one is is simply tied to the bracket. The upper one is supported by it's leads and sits close enough to the right position to fit into place. I took the lenses out so that the LEDs could be supported by the holes in the shell and also so that they would shine brighter. The lower one is wired to the purple wire from the decoder which would normally be F2 but I remapped it to F6 and gave it a Mars light attribute.
The road number for this loco was somewhat arbitrarily picked from the F7A units listed at The ATSF All Time Diesel Roster. It was simply printed on a P-touch label maker at 9 points. Later I found P-Touch tapes with white lettering on a black background so I changed out the number boards. Note that this tape is opaque to backlighting so that it doesn't work any better on lighted numberboards.
The F7A had problems with
dirty wheels after extended running and it would spit and sputter.
I ran it at the Los Angeles Model
Railroad Society layout and it got so unreliable that I had to
take it off the layout and that was AFTER it got a good wheel
cleaning. I pondered getting an F7B in matching paint for it but I
didn't want to spend the money. Then it occurred to me that one of
the old Bachmann F7A's could be depowered and server as a dummy
unit with power pickup. That is what I did.
The motor was electrically and mechanically disconnected and the pickups were wired to a set of my standard "D" connector pins to the rear. Another matching set of pins was wired to the rear of the Athearn F7A. A set of Kadee #5's was added at both ends of the dummy F7A.
The problem is fixed. I can lift the Athearn F7A completely off the track and it still runs. It still has plenty of power to drag the dummy and the consist up my worst grades while hardly slowing down (BEMF helps some here). We'll see how stable the power pickup is over time.
Then I got to the Athearn rubber band drive locos. I searched the Internet for information posted by some other misguided soul that might provide some ideas for installation of a DCC decoder in an Athearn rubber band drive loco. What I found is that people had decided to put them aside and NEVER convert them to DCC. I suppose that they are being reasonable, but I am not always a reasonable person so I went ahead and did it with a couple of cheap decoders. The B unit has no lights and therefore didn't need lighting functions. I didn't think it need BEMF either. If BEMF is used at all it would be in the A unit. Consisting two locos both with BEMF enabled is usually not a wise thing to do. The two locos will continually fight each other. However, one can use BEMF to set the tone with the other locos simply speed matched to provide power.
The F7B unit initially got a Digitrax DH123D decoder, about $16. This is a basic decoder with silent motor drive and a fairly hefty 1.5 amp motor driver. This decoder is well suited to this loco. The F7A got a slightly better decoder, a DH163D. This one is similar except that it has 6 functions and BEMF motor control. It is also "protected" such that it is not supposed to burn up if there is a short between a motor lead and a track pickup. I wanted the extra function for a Mars light. I could have assigned the reverse headlight function to the Mars light but I used a regular function instead. The DH163D is also slightly more expensive at $24.
As with any DCC installation, the motor must be isolated from the
power pickups. HO locos often use the frame as one conductor and in
these locos the motor is screwed right to the frame with a single
#2-56 screw. There is also a rotation stop tab to prevent the motor
from rotating on it's mount even if the mounting screw comes loose.
The tab is the hard part to deal with.
I attached three pieces of 0.005" styrene to the frame with
superglue. One narrow strip sits in the bottom of the groove, it is
glued only at one end. The other two pieces overlap the gap just a
little so that when the anti-rotation tab is pressed into it's
slot, it will bend the pieces over the edge and insulate the edges.
The narrow strip is then pressed further down by the tab. Because
it is glued at only one end, it can slide to form to the tab. The
motor was remounted with a plastic #2-56 screw but the original
screw could be reused if a sleeve is formed around the screw and an
insulating washer (like a Kadee red fiber washer) is used under the
screw head. The bottom of the motor was filed to clean off mold
flash and protrusions that could puncture the thin insulator.
Proper isolation was achieved with this method.
The decoder itself is simply jammed into a spot next to the motor.
The red wire (engineer's side power pickup) was spliced into the
original wire which was cut off it's lug. The orange motor wire
reconnected back to the stub left behind. The black power pickup
wire was screwed down to an existing #2-56 hole on the top side of
the front coupler pad. The gray motor wire was screwed to an
existing #2-56 hole on the motor frame. As it turned out, I got
both sets of wires reversed but that was no big deal as I as able
to turn the motion of the loco around by setting the NDOT (Normal
Direction of Travel) in CV29. That was it for the wiring. After
setting a new address, 325 (also the number of an existing ATSF F7B
of some kind), and setting start and Vmax CV's, the installation
was complete.
Eventually, I ended up with a spare DH163 decoder with BEMF so I swapped it in for the DH123. Normally one would not want to run two BEMF decoders in the same consist as the locos will tend to buck each other. However, these rubber drive locos have so much compliance in the drive train that it works out. The DH163 is better at dealing with the variable brush resistance of these very old locos and the loco tended to run much more smoothly, especially at low speeds.
I also discovered, after a period of time, that the position I selected for the decoder wasn't the best. The decoder would transmit vibration from the motor to the shell. The shell then acted as a sounding board and made lots of noise. The decoder was eventually attached to the top of the motor with foam mounting tape and all the wiring was routed so that it did not touch the shell either. The noise went away.
The F7A was a
little more involved because there were new headlights to be
installed in a loco that never had them. The DH163D decoder was
initially mounted next to the motor and then later moved to the top
of the motor. The motor was isolated in a similar fashion. The 3 mm
warm white LED headlights were mounted on a 0.030" x 0.250" strip
of styrene. The strip was first taped to the front of the frame and
then the shell was installed. The positions for the LEDs were
marked through the shell with a pencil and then the strip was
removed and and clearance holes drilled for the leads. It was then
reattached with Zap-A-Gap CA and fit checked. The rest of the
wiring was straightforward except this loco didn't have the nicely
placed tapped hole on the top of the motor frame. The orange motor
wire was captured under another screw that also holds on the brush
insulator on the front of the motor frame. On this one, I got the
wires on correctly and I didn't have to reverse the direction of
the loco.
Speed matching these two locos was a bit of a challenge as neither one is a real good runner. I used JMRI to do OPS mode programming to set up the speed matching. Since the locos are also somewhat inconsistent in their speed performance, I figured that using a 28 step table would be a waste of time, I used the basic 3 step method instead. I first set the start voltage, CV2, on both locos to make them move slowly on the first throttle step. Then I enabled BEMF on the F7A and cranked it's gain all the way up to make the decoder try really hard to speed lock the loco. I then set maximum voltage, CV5, to limit the top speed to something that looked reasonable (instead of mach 3). I consisted the locos so that they would respond to the same throttle commands and set them on the layout half way apart. On this layout, a twice around, one loco would be above the other on the front straight. The back straight is not good for speed matching as one loco is climbing and one descending. I then cranked them up and let them run, delaying the faster one as it entered the front straight until the other caught up. Then I could let it go and watch them run more or less side by side. I set the F7B top speed to match the F7A. I then did the same thing at 50% throttle and used Vmid, CV6, on the F7B so that it again matched the F7A. At this point, I coupled them and let them run at various speeds. The locos would push or drag a little in different places and at different speeds. Since the tendency to push and drag was fairly well balanced versus time and wasn't real severe at any time, I declared victory. When I get a decoder in the newer F7A freight unit, I'll do this again, using the newer flywheel drive F7A as a standard and matching the other two locos to it. This is fairly painless to do with JMRI managing the programming. I just drag a slider back and forth for the performance feature that I want and push a button. This can be done while the engines are running and the change takes place immediately.
Note that for the DH163, Vmax will not do anything with BEMF enabled UNLESS, Vmid is set to something other than zero. During the speed matching work, I would set Vmax and then set Vmid to half of Vmax until I got the maximum speed right. Then I would set the throttles at 50% and tweak Vmid as necessary.
Conversion of the Union
Pacific GP7 (of unknown manufacture) was a piece of cake. It was
all a wire cut and splice job. I retained the original headlight.
The only thing I should do is paint the inside of the short hood
black so that the whole thing doesn't glow.
The decoder is a Digitrax DH163D installed free floating at the end of the long hood. I used Brother P-touch labels to number the loco 129 after I found that this was an original number in the UP roster for a GP7 of an unknown paint scheme.
The Athearn SW1500 was a
little more difficult. I used another DH163D. I was able to pull
out the motor easily and crush the fingers on the lower motor
contact so that they did not touch the frame anymore, thereby
isolating the motor. I soldered the orange and gray motor wires
directly to the appropriate contact (when they were not attached to
the motor so as to not melt the plastic they clip over). I changed
the headlight to an LED as the headlight bracket was not firmly
mounted to the frame and it's contact was flakey. I had to ream out
the hole in the bracket a little to make a 5 mm LED fit. To attach
the black wire to the frame, I tapped the hole in the pad of the
rear coupler mount with a #2-56 tap and bolted the black wire down
with a very short screw. That same hole on the other side is used
by coupler mount so I had to be careful that the top screw left
room for the coupler mounting screw.
The real problem was getting it all back together. There isn't a lot of room inside the hood. The decoder is jammed up into the cab, but the wiring continually wanted to rub on a flywheel causing lots of noise and drag. I must have had the shell on and off 30 times before I got it right once. Then I realized that I could not get the couplers back on without removing the shell again. Then there were several more shell installation cycles before I managed to assemble it without a wire rubbing on a flywheel. However, to make that work, I had to build a "wire cage" by gluing a styrene strip to the top of the motor extending past the flywheel. I the glued another strip crosswise at the end to make a "T" shape such that the top of the "T" was just as wide as the shell. This holds the wires up and out of the way so that the didn't rub on the flywheel.
The Athearn
"Classic" freight F7A was a straightforward install using a DH163D.
The extra stuff in a DH163AT kit was not required. The motor came
out easily. I routed all the wires on the outside of the weight to
keep them away from the flywheels. The two motor wires are tied up
to the top contact strip with a small piece of insulated solid
copper wire and routed out the front of the weight and then around
to the back.
I checked the speed matching between this converted loco to the older rubber band drive units and I found that it was close enough as is and I didn't have to mess with it. I set up the rubber band F7A and F7B as a consist, then I added the newer F7A to it as the rear loco. This was because I used the original headlight in the newer F7A as the shell didn't have a hole for a Mars light anyway. The bulb runs at half intensity but it is usually off anyway. I turned the BEMF intensity up on the newer loco and left the BEMF enabled on the older F7A. Usually running two locos in a consist, both running BEMF, is not a good plan as the locos will tend to buck each other. However the rubber band drive units have so much compliance in the drive train that it doesn't seem to matter here. The set runs well together.
The trolly was a very
straightforward cut and splice job to install a Digitrax DH123D
decoder although it eventually got a DH163D for better slow speed
performance. All the places that needed work were easily
accessible. I left the stock headlight in place and arranged the
decoder so that it fit up near the roof at the front, above the
headlight diffuser. The trailer had already been wired across to
the trolly to aid in power pickup, that wiring was simply left in
place.
The Pacific was also pretty
straightforward. The DH163D decoder sits in the cab/firebox, stuck
to the top of the motor with double back foam mounting tape. The
power pickup and headlight wiring is held off the flywheel by
routing then under the flywheel and attaching them to the frame
with a bit of CA.
This pretty much completes the DCC conversion of the HO layout and the existing locos. There is one thing left to do. I have an SFX0416 sound decoder that came out of a large scale Lehmann Porter because it wasn't loud enough. However, it will probably be acceptable for this loco. When it comes back from Digitrax (after being repaired due to a sound programming problem), it'll go into the Pacific's tender.
Then, the next
day, the SFX0416 was returned from Digitrax. The old one was
apparently dead because they replaced it with a new one. The
installation in the Pacific's tender was straightforward. The
tender picks up from the left rail on the front truck and from the
right rail on the rear truck. I tested it both with the Digitrax
8Ω speaker and a "high bass" 8Ω speaker. The high bass
speaker did sound a lot better and was slightly louder as well so I
used that one. It is mounted under the coal load with modeler's
epoxy putty.
The decoder is wired to the rear headlight and provides the reverse headlight function. It also needs the load on a function output so that it will respond properly on a service mode programming track.
Since both the engine and tender picked up power, I cross wired them to allow them to share their contacts. This made the loco more reliable over turnouts and materially reduced the tendency of the sound system from dropping out.
The coal load
was drilled with an array of small holes to let out the sound. The
whole tender shell is the speaker enclosure. The holes are visible,
but you have to look for them. As expected, the sound level
produced by this decoder was weak, but it is marginally loud enough
for HO scale where there isn't a lot of wheel on track noise. This
decoder did not cut it in a large scale installation in a Lehmann Porter.
I
use radio throttles and there is a UR91 receiver in the garage so I
can run the layout without plugging in the throttles, but with the
older Digitrax simplex radio, I need to plug in my DT400R to select
or deselect a loco. I have a UP5, like this one, across the room
but eventually I stretched a long LocoNet cable to the HO layout so
I could leave the throttle plugged in at the HO layout. However,
after the day that I got tangled up in the cable across the floor
and my expensive throttle hit the concrete floor (no damage that
time), I realized that the long cable was not a good deal. I
installed another $15 UP5 remote panel on the HO layout's control
panel so that the throttle can be left there, plugged in and
safe.
The initial DCC installation is all done and it works just fine. I find that I don't even bother with the DC power packs anymore, they are not needed, although all the engines will analog convert and they can run from DC. I use the Cab B/DCC switch to turn the DCC power off when the layout is not in use. The AC power to the rest of it is run off a switched outlet that also powers in the large scale storage yard panel in the garage so that all that stuff goes off with just one switch.
Later a couple of "new" locos showed up. These were the UP USRA 0-6-0 switcher and a gas-electric Doodlebug. Both are Bachmann products that I picked up from bargain tables at a train show for $5 each. As could be expected, both had problems. The Doodlebug had gearing issues and went back to Bachmann to get fixed under the lifetime warranty. It has a DCC socket so that it will take a 8 pin decoder, what type I use has not yet been determined but initially, it will be a spare DH123. The UP switcher appeared to run well, but it had other problems as I found out later the hard way.
This is the USRA 0-6-0 apart.
It will get DCC as soon as I locate a decoder for it. I have a
spare DH123 but it's simply too big. The entire boiler is filled
with weight and motor, the only place is inside the cab and the
DH123 is marginally too big to fit there. I'll have to look around
for much smaller decoders. The stall current of this loco pushes an
amp so that it'll take a hefty one to do the job.
Wiring will be an issue too because the main structure of the loco are the weight halves which also provide power distribution. The motor, headlight and smoke unit (dead) are connected to this structure. Isolating the motor tabs will require complete disassembly of the loco. The smoke unit is going to get tossed and the headlight will probably be replaced.
I did find
a spot inside the boiler for a decoder by removing the smoke unit
which just pulled out. The cavity left behind is just big enough
for a DZ143 decoder. This is a full function decoder with BEMF but
with limited capability for handling track voltage. It will handle
1.25 amps which is enough but it is limited to 18 volts. Since I
run 16 volts on the track, this is good enough.
The headlight is clamped between the weight halves, I will be able to pull it, replace it with a 3 mm LED and then clamp the track inputs to the decoder between the case halves just above the LED. The headlight itself is above the boiler so that the LED will shine upwards from the clamp and into the headlight diffuser.
The motor
tabs, however, will be a problem. They are buried between the
weight halves so I'll have to pull the weights apart to separate
and insulate them.
The DZ143 decoder
came today and it is pretty small but it still just barely fit. The
motor contacts were a little easier to deal with than I thought. It
was easy to remove the rear pinch screw from the case halves and
spread them just far enough to expose and remove the contact
springs that were in the halves. I then soldered the motor wires
from the decoder to them and insulated them with shrink tube. I
filed a little bit on the weights at the front to see what they
were made of and it appears to be some from of copper based casting
alloy. This means that I could solder the track leads to the
weights, which I did.
The decoder responded on the programming track and indicated that it was happy. However, it went downhill from there. But when put on the running track, it ran for about 2' before stopping dead. It never ran again. The decoder still responds to the programming track but it won't drive the motor. It probably died. Now I get to evaluate Digitrax's "No Worries" warranty.
The motor in this loco tests at a full stall current of 2 amps at 18 volts, a loaded current of 500 mA and a free running current of 150 mA. If I have to, I can add some resistance in series with the motor to limit the current, but then I have to find room for the resistor. I'll need about 4 or 5 ohms at about 1 watt. I may have to make it from several smaller resistors so I can form them to the available space.
After some investigation, I find that this particular motor has an intermittent commutator short. Every few seconds, it'll draw a spike of as much current as is available. It was spiking over 20 amps for a microsecond or so when run from my bench supply. This is probably what did in the decoder. Motors do not normally do that. The solution would be to replace the motor or to add some current limiting resistance in the motor circuit so that the decoder doesn't fold up.
After a few days of messing with the motor, the spikes started to get smaller and fewer. I temporarily installed a DH123 with a 10 ohm resistor. The loco runs but the shell won't go back on. Then I let it settle down and run slowly for several hours and the spikes essentially went away. The worm will run at about 1/3 rev/sec with some slight hesitation 5 times per turn. This hesitation is clearly due to either commutator friction or torque changes as the various segments are connected and disconnected. When a new DZ143 arrives, I'll use a 4.7 ohm half watt resistor in each motor lead. The two resistors on two wires will be able to snuggle down into the cavity with the decoder and then the shell will fit back on.
The
0-6-0 has pretty good power pickup by itself, but even 6 good
wheels are marginal so I added power pickups to the tender. The
stock Bachmann trucks had plastic wheels so they were junked and
replaced with Athearn Bettendorff freight trucks with metal wheels.
I fabricated some wipers from 0.015" phosphor bronze wire by
drilling a 0.015" hole through the truck frame, inserting a wire
and forming it. A 15 mil drill bit has roughly the same stiffness
as a wet noodle. The holes were drilled by hand with the aid of a
pin vice.
The wire wanted to rotate in the hole so I tried securing it with CA but the CA didn't adhere well to the truck material. I eventually epoxied the wires in position. The epoxy doesn't stick to the truck material very well either, but it is wrapped around the frame and provides a mechanical clamp to the truck. These trucks no longer free roll. Some drag is expected but it isn't a lot.
Digitrax was good to
their warranty and sent me a new decoder. The thing just barely fit
with the LED headlight and the two resistors. I had to grind part
of the weight away to allow the decoder to settle a little deeper
into the pocket. This time, it ran properly.
Bachmann also called a few days ago and said that they didn't have the parts for the Doodlebug so they are going to send me a new one instead under their warranty. I'm not sure what roadname it will be because they didn't have any undecorated models.
After I got the trackwork together and started in the scenery, it was clear that I was going to have to see what kind of structures that I had and what I was going to need to be able to plan where to place them and make sure that the hardshell allowed that placement. As it turned out, after digging through more boxes, I think I've found all the existing structures and most of them won't interfere with the scenery work. Most are railroad related buildings that there will fill up the yard area. What I need is some non-railroad related small town structures. Much of the stuff in the following photos dates from 1960 or so. Most of the stuff is a little beat up and needs some cleanup and paint work, but it will do.
The engine house
has been sitting out on the layout the whole time. It really
shouldn't be sitting at the end of a couple of yard tracks, but
that's where it will go because that's the only place that it will
fit. I glued the back doors shut, the front doors will remain
open.
This sand house
will sit next to the engine house servicing one of the tracks
leading to the engine house.
This water tank,
less spout, will also sit on that same yard track. I guess that the
last yard track is now an engine service track.
The turntable is in
place and I'm not moving it. It would normally service a
roundhouse, but the Atlas turntable has far too great an angle
between tracks to work that way. It will simply serve the engine
storage tracks.
This is a small
yard, so it gets a small yard office. There is no room for it near
the engine house so it will sit on the other side of the yard.
There are
actually four switch towers or shacks. The largest one will protect
the yard spine, but from the wrong side. There is no place to put
it on the "right" side.
There is also no room
for this oil tank near the other engine service stuff so it will
sit next to the yard office.
This little
lineside shack can go just about anywhere. I'll find a spot for
it.
I don't know what
I am going to do with this, probably not use it. This thing is set
up to actually dump a load by pulling on the knob in front. It also
assumes a raised track below it so that loads could be dumped from
the cars as well. I have two cars with working dump doors that were
intended to go with this but it doesn't belong in a yard and I
don't have a coal mine so it will probably go back in the box.
I have a variety
of switch shacks. I found the PlasticVille tower in pieces, but I
found ALL the pieces. Anyway, a little CA and it's more or less
back together again. I'll find places around the layout to place
these structures to protect turnouts.
The section
house was also sitting out on the layout. It is amazing that it
didn't get crushed by all the stuff stacked on the layout over the
years. It will probably go right were it is.
I have both
a passenger and a freight station. However, the spur that services
the town area isn't long enough to hold the 5 car passenger train.
I think I'll leave it up on the ledge overlooking the town, right
on the main line and maybe build a stairway, ramp or even a road
from the town to the station. I need to find a way out of town for
a road anyway, maybe this is it.
The freight
station will go in the town area and service the spur that goes
there. The Doodlebug can service this track for both passengers and
freight.
The only non-railroad
related building that I have is this small factory building. I
don't remember this one so it is probably one that Richard built.
I'll need another five or six representative structures to fill out
the town are which right now is a flat chunk of Homosote.
I bought this kit at
an estate sale at the BTS last June. I think I paid $5 for it. I
priced HO building kits at retail and kits run $25 up now. It looks
like I'll be attending model railroad swap meets looks for
buildings of opportunity in the future as I would like to find
another half dozen small town buildings of various kinds.
It took three months but I finally got around to building the kit. HO is just too small, the paint work is a little sloppy in spots.
Last weekend I went to a model railroad swap meet in Lakewood CA. There weren't many vendors there, but there was one who had what I was looking for, buildings. Most important, he was selling them cheap. I got the following five buildings for a total of $20. These were clearly used and taken off somebody's layout, but that was fine with me. They were adequate to fill out my "town."
This nondescript office
type building could be a bank, a store or an office building. Since
I have a bank, see below, it'll be something else.
The church is a large and imposing
paper and cardboard kit. However, it looks pretty good.
This little office
supply store could be converted into any kind of store with just a
sign change.
The Wells Fargo
express office will go right next to the railroad's freight
station. It is the right kind of town building that would sit next
to the town siding.
Every town needs a bank. This is a
perfect 50's vintage bank branch, complete with an interior.
After the trackwork was reliable and the electronics and DCC upgrades were complete, it was time to start on improving the scenery which was in pretty poor condition. Richard and built some hardshell and textured some surfaces, but it was nowhere near done. There are still gapping holes where hillsides should be and the two "tunnels" were just passages under the raised portions of the layout. I decided to work the tunnels first.
There are four tunnel portals. Two are fairly conventional and and opening the of the shape on the NMRA gauge would be fine. One is somewhat oblique and would need to be wider and one was just plain weird with the track exiting at about 30° to the face of the tunnel.
I was pretty sure that I could not get a prefab foam portal that would work in the weird location so that I would have to fabricate something. Since I was going to fabricate one, I elected to fabricate them all. A normal HO modeler would probably use foam, it's light, cheap and easy. I elected to use concrete, real concrete. I had a lot of scraps of 1/4" WonderBoard from previous large scale building constructions project. Two of the scraps were exactly the right width needed to make the portals and there was nearly 4' of it. WonderBoard coated with a stucco of RapidSet cement mix makes a fair looking concrete surface.
This is the odd
tunnel portal with a cardboard template taped in place. It needs to
be this wide to allow the streamliners to pass through it at a
shallow angle. I used the templates to transfer the shape onto the
WonderBoard strips and cut them out. Actually, WonderBoard doesn't
cut too well, it needs to be scribed and broken. The inside shapes
are made by scribing and mashing on it with a hammer to shear out
the shape. The scribed edges are then coated wit RapidSet to seal
and form them to shape.
The two odd portals can be
seen in this view with a large section of "retaining wall" in
between them. The edge nearest the camera was scribed and broken
along the scribe line a little to conform to the curve. The break
was filled in with RapidSet. The pieces are held in place with some
Gorilla Glue. This stuff is good at filling gaps but is not super
strong so that mistakes can be broken out and redone. Due to some
kind of measurement error, there was a 3/4" gap between a couple of
pieces. I just made a form with blue painter's tape and filled it
from the top with RapidSet. I think that I'll continue the wall
past the Lionel power pack to the end of the layout and also around
the curve on the left some distance.
In the upper part of this photo, the backside of the third portal can be seen. The hardshell will be extended to make that a real tunnel. Both tunnels have access from underneath so that the track in the tunnels can be cleaned and wrecks cleared without a terrible hassle.
I hardshelled the
areas that still needed hardshell using the time honored cardboard strip and plaster
soaked paper towel method. Then I painted the plaster with an
earth tone latex house paint and sprinkled on a couple of grades of
sifted decomposed granite for texture. I still need to add grass
textures and some trees and bushes.
Besides the area
round the town, there was another gap at the other end of the
layout that got hardshell as well.
The flat parts of the town need surface texturing, a road and a bridge to the upper corner where the passenger station is. There are a few other flat areas that need surface texture too. Once this is done, the scenery, basic as it is, will be complete.
This page has been accessed times since March 17, 2009.
© 2009-2010 George Schreyer
Created 17 Mar 09
Last Updated July 13, 2010