Bachman motor control settings using ESU 5XL decoder

P

pmindigo

Registered
Country flag
I just installed an ESU 5XL decoder in my Bachman 2 Truck Shay (upgraded trucks) but it is jerky at low speeds. ESU has a table of motor control settings but I need to know what motors Bachman uses. Options include Marklin , Fleischmann, Hag, Trix, Faulhaber, Piko, and Buhler.
Does anyone know which motor Bachman uses?
Has anyone installed an ESU 5XL in a Bachman Shay?
Thanks!
 
Last edited:
pmindigo said:
I just installed an ESU 5XL decoder in my Bachman 2 Truck Shay (upgraded trucks) but it is jerky at low speeds. ESU has a table of motor control settings but I need to know what motors Bachman uses. Options include Marklin , Fleischmann, Hag, Trix, Faulhaber, Piko, and Buhler.
Does anyone know which motor Bachman uses?
Has anyone installed an ESU 5XL in a Bachman Shay?
Thanks!
Click to expand...
I believe there is an option to let the decoder decide the best option. It from memory involves invoking a cv and letting the loco power itself to and from on a few yards of track. Notes on how to do it in the manual. I did it with my G1 Terrier with a similar decoder and whilst not perfect made the running so much better.
 
I've heard that option is about 50/50 as long as you don't run the engine off the truck. Still looking for specific CV settings.
Thanks.
 
The automatic motor programming worked and the engine is now running smoothly.
The following settings were automatically updated in the controller:
CV Value
2 2
9 40
51 2
52 5
53 120
54 5
55 21
56 255
116 50
117 150
118 6*
119 6*

* I set 118 and 119 to 6 to reduce the motor sampling window. These were not automatically adjusted.
 
Have you removed all the electronics from the motor terminals? I can't remember which Bachmann loco it was, but I couldn't get smooth operation until i had bare motor terminals. There is the auto-tune function as described above using cv54 set to value 0. Then press F1 and the loco will fire off at full speed for 2 secs. Make sure you have the loco direction correct and there is enough track! If you read the cv settings back to the LokProg software, it'll be easier for making finer adjustments.
 
I recently finished a rebuild of a wreck of a Bachmann 3-truck Shay which was sold as C6 with radio control - turned out there were no track connections and the bits were loose in the coal bunker. The shop in the US did have the decency to refund US$150 towards a decoder.

It is now fitted with a LokSound 5XL decoder, a CVP Airwire CONVRTR-60X and a 14.8 LiIon 5 S/Hr battery - had its first run at the AMRA Glen Iris club track a fortnight ago and even though the antenna is tucked inside the water tank I was pleased with both the reception and the smooth running of the locomotive (it had also run on the rolling road at Sandown over the weekend of the show).

Program and soundfiles are the LokSound v5 Shay file with some minor modifications (the wiring includes an RCA plug that allows direct connection to a LokProrammer for changes to the DCC settings).

It is not quite finished - still has some detail parts to be fitted, custom decals and some more weathering but I'm reasonably happy with it as my first Fn3 project after working in 1:48 since my return to the hobby:

20240318_125302.jpg
 
I seem to have a significant jerky drive in reverse, forward is much more smooth but I have a feeling the locomotive is working too hard.I can't see all the CV's above as I set up the decoder in MFX mode but CV 53 to 56 are set like above. I changes the minimum speed.but that did not help, it's jerky in reverse even at faster speed. I had these this engine prior to decoder install in analog and it seemed to drive very smooth
 
My locomotive shown above continues to run smoothly in both directions

Current CV settings are set as follows:

CV53 - 120
CV54 - 5
CV55 - 21
CV56 - 255

The jerky drive in reverse suggests a mechanical problem to me - not a decoder issue.

If you have a LokSound programmer I would download the file from ESU and reload it.
 
I'm with David..
I doubt jerking is due to the decoder, but believe there is a procedure to 'tune' the decoder to the motor in the model?

PhilP.
 
I didn't have to do any modifications to the ESU Shay file from an operational point of view - just changed the whistles and a couple of auxiliary functions like the Westinghouse pump sound - my three-trucker has the last series motor blocks.
 
did some research.... I think it has to do with the MFX mode, I am using marklin CS3 that starts the loco in MFX mode, it does not show all the motor control CV's. the BEMF is asymmetric for the 2-truck shay in reverse and forward because of the gearing and transfer rod angle that is a bit different (like the real thing) so being jerky in one direction means the PID feedback is too twitchy... CS+ writes its own speed table as well which is not meant for geared loco so I changed that from default to linear to make it closer to the decoder's exception..... it got a bit better but still "steps" in reverse ....I will report back if I find a solution

speed_control.jpg
 
I run essentially the same ESU Shay file in both On30 Shays like these two modified Bachmann 2 cylinder two-truckers (under NCE control):

Double_Shays_in_the_Bush[1].jpg

As well as in the Bachmann Fn3 three-trucker under CVP Airwire control - this is a shot of it after weathering and decals (can barely cope with the LGB radius curve in this siding at the AMRA club track):

Shay & Mallet at AMRA.jpg

(The Mallet also runs an ESU 5XL decoder under Airwire with appropriate sounds from the units at the Black Hills Central Railroad).

I say essentially because I tend to tweak the decoder settings and change sound files such as whistles and pumps to suit my liking.

All of the Shays (including a Precision Scale brass 26 tonner regauged to On30) run reliably at exhibitions.......
 
Ok , so I did quite bit of digging and I think I found the root cause of the problem

the motor optimization was just a distraction, since the decoder was already loaded with shay files the settings were readily optimal and had nothing to do with CS3 either.

I ran the locomotive indoor on a simple layout with no issues. My outdoor layout has a large number of isolated contact tracks sections attached to the Marklin S88 DC. the S88 DC senses motor return current that is connected to ground of the system internally via a diode.

The ESU decoders reliers on motor's BEMF sensing to compensate the motor torque for the load, the BEMF is sampled at very high frequency like 40KHz, this creates a very sensitive feedback loop by its nature.... now when the locomotive crosses from one contact section to another the return path has a delay to establish feedback in the circuit via the S88. the delay is in the order of msec... until the feedback is established there is some ground bounce from the motor return current (also evident in flickering of layout lights that feed directly from that section of track)..... the ground bounce screws up the BEMF measurement (it can even reverse polarity) so the feedback starts to oscillate causing stutter in the motor.... I had actually seen the same thing with Piko with Soundtrax DCC decoders in the Piko moguls but never quite nailed down the root cause

so why LGB locomotives don't have this problem the answer is that LGB decoders use a carefully designed low pass filter in the feedback loop to reject this momentary noise, and when feedback is not reliable they fall back on the factory calibrated load curve that is provided by the MFX controller, so they are much less sensitive to "noise on the track" especially from this ground bounce when passing between the isolated sections....

I measured the track waveform by a scope, you can see the difference when LGB 28442 is running vs. when shay is running , the current oscillation from the motor PWM modulation is distorting the track voltage in case of Shay with ESU Loksound whereas in case of LGB 28442 the waveform is clean. so safe to say Marklin/LGB designed the decoder with this detail in mind as they also make the S88

Next to be 100% sure I will install an LGB MFX decoder for the shay and confirm smooth operation, unfortunately this decoder does not have Shay sound and some other features found on the LokSound so it is not a solution

I have a few ideas how to solve this problem, I will report back if I am succesful.

I thought this information might be useful for others. it applies to folks who have a layout with many isolated contact track sections (for automation) fed back to an S88DC.... otherwise ESU should work find with marklin cs3 , in fact I also confirmed it works fine with MS2.....




LGB 28442.

fig1.png

ESU Loksound 5XL

fig3.jpeg
 
David Price said:
I run essentially the same ESU Shay file in both On30 Shays like these two modified Bachmann 2 cylinder two-truckers (under NCE control):

View attachment 352486

As well as in the Bachmann Fn3 three-trucker under CVP Airwire control - this is a shot of it after weathering and decals (can barely cope with the LGB radius curve in this siding at the AMRA club track):

View attachment 352488

(The Mallet also runs an ESU 5XL decoder under Airwire with appropriate sounds from the units at the Black Hills Central Railroad).

I say essentially because I tend to tweak the decoder settings and change sound files such as whistles and pumps to suit my liking.

All of the Shays (including a Precision Scale brass 26 tonner regauged to On30) run reliably at exhibitions.......
Click to expand...

excellent envy your layout
 
Thank you.

The 1:48 scale work is a section of a two part exhibition layout - my colleague Dan Pickard's "Kinmont" & my "Winch No.3" - pictures can be found in the Layout Galleries section of the Australian section of the 7mm Society here: Layouts

I should add that all of the locomotives and most of the rolling stock in the "Kinmont" photos are my work - my "Winch No.3" diorama is the first I have ever done and I am currently building another diorama to replace "Kinmont" for exhibition at the Australia Narrow Gauge Convention in Melbourne this coming Easter.

The shot of the Fn3 locomotives are at the Australian Model Railway Association (Victorian Branch) track at Glen Iris - I don't have a track of my own.

I also run live steam there as well - two 7/8" scale units of mine:

20260105_135207[1].jpg

I run both live steam and battery power with radio control at the club layout, despite it having track power as I can run both at the same time without having to worry about insulated wheels (which none of my live steam have).
 
Last edited:
And well done with diagnosing your issue with the decoder

And partly for this reason there is somewhat of a movement among my colleagues (even in 1:48 scale) to move to battery power/radio control like the work of Giles Favell's (look for gilesengineer on YouTube).
 
@LGB_Chur_Arosa Excellent piece of diagnostic work. Trying to get to the bottom of the strange faults, can be a frustrating, and time consuming, process.

PhilP.
 
You must log in or register to reply here.
Back
Top Bottom