Best Of The Overland/Wiseman brass Western Maryland Shay thread

[u18b]

Great question Pete.

No doubt there will be some wear- but I would think nothing compared to gears.

And the good part is- the brass tubing and the bearing are replaceable.... though the tube would wear faster than the bearing.  The tube is very soft.

[Chris333]

Usually bearings like that are made out of bronze. I have no idea what the Atlas part is made of.

[u18b]

No pictures today.  Catch up post.

Numbers: Measuring Progress

Having tweaked the trucks and installed an end bearing on the motor/worm shaft, it was time to do more testing.

In my testing, I have used my digital speedometer as a check.  This may not be as good as checking the amperage as Max does, but it I find it a good secondary indication of performance.  And speed is easy to remember.

I use the Accutrack II.  I bought it for speed-matching EP-2 cabs.  This thing has been a lifesaver for me. 

https://sprog.us.com/accutrack.html

I knew that previously this Shay with new this motor ran at wide open throttle at about 109 mph (yes, you read that correctly).  This was my baseline from previous tinkering.  After working on the trucks and/or in adding a bearing, if it only ran at 75 mph, then this indicated I had a bind to correct requiring more tinkering to work it out.

After all was said and done, this Shay (when warmed up) will now max out at 95-97 mph (some slight loss is expected since the bearing adds a sight bit of friction as compared to no bearing).

As for speed- I know 100 mph is stupid-fast for a Shay, but remember I re-motored this locomotive.  As it came from the OMI factory, it ran at about 300 mph!  It was absurdly comical.  I actually feared it would fly off the track on curves.

So currently, using my MRC Railpower 2400 throttle, this Shay is now at a point where it will run smoothly at about 12-15 mph.  When I first remotored this Shay, the best I could get was smoothness at ~29 mph (which was prototype max).   It could run slower, but you could see some variations.

So getting it down to 12-15 mph is a tremendous accomplishment.

[Jim Starbuck]

I really appreciate your approach Ron. It’s truly inspiring to watch your work on these little machines. You’re building jewelry that runs.
I agree wholeheartedly that getting a locomotive to run is just the starting point. Getting them to run slow and steady involves multiple aspects such as gearing, friction reduction and electrical stability. That’s where the time is spent but also the satisfaction when they run well.
I’ve “back burnered” more than a few projects to give time to think through and come up with a better working approach.

[u18b]

Thank you Jim.
This has been quite a journey.

[u18b]

This section of this large thread has been devoted to analysis and rebuilding of a version 1 Shay- which in some ways is harder than rebuilding a version 2 Shay.

As such, there is a loose end I have not covered:  the valve rocker and lifters are different from version 2.

Review:  valve lifters ride on bearings on the main crankshaft.  Those lifter blades move the rocker (which internally moves the valves on the prototype).



If you have been with me all along then you know that OMI had large holes in the piston frame on the version 2.  Also, the ver 2 rocker castings had a central pin.  Some folks inserted the long rocker pin into the frame hole and soldered the valve rockers in place.  But I wanted them loose, so I drilled holes in the rockers, and ran a .008” wire through those holes and through the big holes in the cylinder frame.



That .008 wire held everything together.  In fact, by cutting  that wire and then removing the bearing retainer plates on the crankshaft down below,  the version 2 crankshaft and valve assembly may then be removed as one unit.

This is not possible in a version 1 Shay.  First of all, the crankshaft is locked in when the mounting plate (or cover plate) is soldered to both the frame and the cylinder assembly- and as I have shown, it takes a lot of work to get that all soldered in a way that it does not break loose.   If it is good and secure, I don’t want to be messing with it.  If I ever needed to repair or replace the crankshaft parts, it would be a big job.



Second, there are no open holes up in the cylinder frame.  Instead, there is a small hole in the frame and a short brass wire sticking out as pins.  These pins are soldered in place.   Also, original ver 1 rockers already have the central hole.  The hole in the valve rockers goes on those pins.  The wire is then folder over.  (if the pins ever break, they could be drilled out and wire replacements soldered back- but repainting would be required).



The back side of the rocker has pins as well.   These pins retain the lifter blades.



While none of this assembly can be removed for sandblasting and painting, it does nicely allow the valve rockers to wiggle a bit during running.  As such it is probably a good idea to add a drop of oil on those pins once in a while.

[u18b]

Now that I’ve smoothed out the performance of this version 1 Shay, I decided it’s time to install a decoder for two reasons. 

1.  Analog top speed of this remotored Shay is 100 mph- which is crazy (the prototype tops out at something like 29 mph).   DCC will allow me to slow it down to a more reasonable number.  Even with a new motor, this Shay exceeds realistic speeds in analog at about 20% throttle- so 80% of the throttle is wasted.  DCC will give the full throttle range  by setting a Max voltage.

2.  I can also add an operating headlight- which can stay on when the loco is stopped.

Space is a lot tighter than on the version 2 Shay.  But I concluded that a Z decoder would fit.  I had a TCS Z-2 on hand (but a Digitrax DZ126T would work as well). 

The best place to put a decoder is under the motor since there is already a small gap between the firebox lead weight and my new motor housing.  We just need to make the gap larger.



Using a motor tool, I removed about one third of the weight, but remember that I previously added weight to this loco by filling up the boiler shell- so I’m still ahead.  I also added weight to the front of the tender so that it would push down on the cab.  If I still need to add some weight, I can always get a couple of grams of lead in the inside cab roof.





Now the decoder can freely sit under the motor.



I then had to update the circuit board up top.  Since it is a double sided board, I added more solder to the underside where it meets the motor housing.  It is now very secure and much better than superglue.

I then cut traces  with a fine motor tool to isolate pads for DCC.
The plan was as shown below.
Pad 1:  Frame (right rail), red DCC. The frame wire just goes to the motor housing.
Pad 2:  Red motor wire, orange DCC.
Pad 3:  Gray wires from insulated wheels (left rail), black DCC.
Pad 4:  Blue motor wire, gray DCC.
Pad 5:  Positive LED, blue DCC.
Pad 6:  Negative LED, white DCC.

Actually, a TCS 2-pin micro socket will go on pads 5 and 6.



Now, I normally like to make things modular with micro sockets and plugs for easier disassembly, but space is really tight.  So everything except the LED is soldered the old fashioned way.   When it comes time to paint the loco, I’ll have to unsolder these wires.

I mounted the motor so the decoder would sit properly in its spot in the weight, then soldered the decoder wires to their pads.   

I then added superglue where the decoder touches the motor housing.  When the glue dried, I removed the motor and added more glue. 



Putting the motor back, I then soldered the gray insulated wheel pickup wires to pad4 as well as the LED socket. 







I took it for a test drive to confirm all works.

[Jim Starbuck]

Excellent work Ron!

You may know it already but something I do on my Digitrax installs is turn the always too high BEMF default way down which smooths out performance dramatically. CV 57 default is 6. I’ve found a value of 1 or 2 makes the motor much happier.

Jim

[u18b]

I will give that a try.
Thanks.

[u18b]

Now for the headlight.
This is an afterthought since I had already filled the boiler with weight (I was not planning on DCC originally because- quite honestly- I didn’t think I would get this Shay to run this well).
I could have melted that low-temp weight out and start over, but I decided to just drill a hole in the weight.   

I followed the procedure described many pages back:
I then used resistance soldering to remove the headlight casting.
I drilled out underside of the headlight casting for a new hollow brass tube.
I drilled out the tube so the walls would be extra thin for sliding a tiny LED up through it..
Soldered the headlight casting onto the tube.
Drilled the hole on the top of the boiler for the new tubing.
Trimmed the tube and slid it into the shell.
With the revised headlight in placed, I soldered the edge of the tube to the inside of the boiler.
Trimmed the tube and cleaned everything up.

I then installed a very tiny LED with a 1K resistor in the middle and with a TCS plug on the end.

Nose view.


Boiler view.


Right now the LED sticks out because it would short inside the headlight.
After I paint the headlight casting and insulate the LED, then I’ll install the headlight properly.



The plug is color coded for polarity (matching silver mark on the socket and the plug).  Since I was running out of room, I bent the ends back a little.  It is important to add superglue to the wire, heat shrink tubing, and the plastic of the plug  (of course doing this when it is NOT plugged into the socket, and avoiding getting glue on the gold plated plug tips).  This holds everything together better.  Otherwise, one day when unplugging, the insulation will pull away exposing the solder joints.



I almost always make a plug and socket if there is some wire going to the shell (like for a headlight LED).  When you pull the shell, you want to unplug the wires, not yoink on them accidentally).



Only a few more loose ends.
Painting is getting close!

[u18b]

One more loose end.

As I do with all Shays, I added a pick-up wire to the tender.  As stock, the insulated wheels are wasted. 

I take a double-sided circuit board and solder one side to the cross brace for placement.  I then solder a 2-sided wiper to rub on the insulated wheels.  (red in photo below).

A jumper wire goes to the loco (green).  The jumper has a TCS micro plug.   I then soldered a socket which terminates on the circuit board over the motor.  The plug/socket resides under the truck boalster.

The solid/geared side also has a wiper that is pinched between the cross brace and the truck frame (light blue).  This provides better electrical flow to the drawbar (and engine frame).



Here is another view of the insulated side.  This boosts electricity to the motor by 50% (from 4 wheels to 6) and makes a HUGE difference in these locos.



And the geared/solid side.



Those last two photos are actually helpful to show that these Shays are highly susceptible to hair and fibers.  I check the loco over once in a while to remove them as they obviously degrade performance.

[u18b]

I’m almost finished (still have to add brake hoses, for example).

Performance tinkering is about as good as I can get it.

Overall- very very pleased.  Remember from where we have come with this specimen.  This is a version 1 Shay.   They are notorious for not running reliably at all.  Most are shelf queens.  If you have one that runs reliably, it is rare.  The recall was for a reason!   (but if you have one of the rare ones that runs, it runs about 300 mph wide open, as I recall- yikes!).

Most are shelf queens because:
-The twisting on the cab truck because of the  torque from the main drive gear causes derailments.
-The power reverse poses an obstruction which also aids in derailments.
-No extra power from the tender.
-I also found the angle on the main drive truck can be a problem (meaning the main drive truck does not always lie flat with all four wheels evenly touching the rails).

Even beyond the common problems above:  This particular specimen:
-was COATED in paint from a rattle can and was locked up solid.
-had a shorting assembly flaw from an improperly trimmed piece of metal that was causing a short
-had a damaged gear further preventing running.

I really wish I had a “Before” video to compare- but then I remembered:  this thing would not run at all; then it would run, but only 4” or less at a time.  By the time I figured out the problems, I had already made enough changes that it was not a stock loco anymore.  So it seemed pointless to try and make a stock/before video.

So a version 1 Shay that actually runs, and runs reliably is very satisfying.  I wish it ran a bit smoother, but this is pretty good (“a silk purse from a sow’s ear”).  That fact that it will creep at all, after all I’ve been through with this loco, almost feels like a miracle.

I find it does run rougher at first, but runs better after it is warmed up.

I programmed the decoder for a bit faster than scale speed.  Max speed or the prototype is about 29 mph as I recall.  Of course the model will run a tad slower pulling cars.

Performance Video.

[peteski]

Ron, nice job on "decoderizing" this model!  You have used ESU decoders in the past so you are likely aware of the superior motor control they have.  I'm curious why didn't you use the ESU LokPilot micro instead of a Digitrax decoder?

[u18b]

Simple answer:  I had a Z2.

 

[peteski]

Simple answer:  I had a Z2.

 

Fair enough.  If you have an older Z2, it has a bug with speed steps which causes it to go slower then  faster on each subsequent step even though you keep increasing the speed on the throttle.

TCS was aware of the bug (I spoke to their reps several years ago at Springfield) and they told me that fix was coming. But I never bought any more of them to see if they actually fixed the problem. With ESU and ZIMO having prices comparable to domestic decoders I'm not longer using TCS or Digitrax decoders, and I never used NCE decoders either.