DCC conversion and lights switching update of the DRS Bachmann Class 57/0 locomotive plus a comparison of motor control performance between an ESU LokPilot V4 decoder and a TCS EU621 decoder

 

Introduction:

This page provides a summary of the process adopted to incorporate a DCC decoder plus independently switchable rear and forward lighting in the DRS liveried Bachmann Class 57/0 Locomotive (OO gauge).

A comparison of the motor control performance will then be carried out using a) an ESU LokPilot V4 decoder and b) a TCS EU621 decoder.

Note that the smaller pictures can be enlarged by clicking on the image.

Class 57/0 at York station in 2008.

 

Purpose of the exercise:

Two DRS Class 57s are required to power a nuclear flask train. The plan is to equip the first loco with an ESU LokPilot V4 decoder and to incorporate a pair of simple transistor switches to enable the front and the rear lights to be independently operated under DCC function button control.  The ESU decoder motor control performance will then be compared with that of an otherwise identical second Class 57 unit equipped with a TCS EU 621 decoder. The best decoder will then be chosen and duplicated on the loosing loco, to provide identical control systems for optimum consisting.

 

The lighting fit on the real world DRS Class 57/0 locomotives:

At the forward end of the locomotive, just above the buffer beam, is a single high intensity headlight, used for both day and night running. Additionally,  there are two warm white marker lights near the centre of the front body panel. At the rear of the locomotive are a pair of independent red rear lights. The lighting arrangements are identical at both ends of the locomotive.

When the locomotive is the lead vehicle of a train, its rear lights are turned off. (In this situation, the rear warning indication is provided by a flashing rear light on the rear wagon, instead.) If however, the train also has a locomotive positioned at the rear (in "top & tail" mode) then the rear lights of that locomotive provide rear warning for the train and the front lights of the rear locomotive are switched off.

DRS Class 57/0 locos can be found operating at both front and rear of nuclear flask trains. There is thus a need to make both front and rear lights switchable in a representative model.

 

The Original Bachmann Lighting Arrangement:

The Bachmann model includes directional headlight, twin front marker lights and twin rear lights, using LED technology. A switch is provided on the underside of the locomotive to enable or disable the rear lights.

Changes to be incorporated:

Provision to independently switch on and off both forward lighting and rear lighting by means of the DCC controller function buttons. Subdued (directional) cab lighting is also added.

 

Dismantling the locomotive:

Six screws must first be removed as indicated in the photos below. The upper body shell can then be carefully separated from the chassis assembly.  

Twist the bogie to find the screws

Six screws in total

 

A brief moment of panic as I discover the service sheet advises we have an 8 pin DCC socket (I have 21 pin decoders!)

Then:

The Class 57/0 Chassis Revealed

 

Exploring the Bachmann Lighting Arrangement:

The Bachmann model uses leaded hemispherical ended lighthouse style LEDs, fitted to small lighting PCBs, which fit into black plastic light housings at each end of the locomotive. Marker LEDs are yellow, Headlight LEDs are white and rear light LEDs are red. The LED series resistors are located on the small lighting PCBs. These are all tiny SMD 1k2 types. Every LED has an SMD radio frequency decoupling capacitor fitted in parallel. The small SMD resistors must have a limited dissipation. This may be the reason for Bachmann to place a parallel pair of large 470 ohm SMD resistors on the main PCB, in series with the positive feed from the decoder. Several volts are dropped across these resistors when the LEDs are active, somewhat reducing the power dissipated in the small SMD 1k2s.

 

The light fittings

 

Although screw holes are in place, the housings are glued

 

Headlights and rear lights underneath

The rear of the lighting PCB

 

Tracing  the track pattern with a meter, revealed the circuit diagram

Add a couple of simple transistor switches and that takes care of the hardware.....

This works fine BUT...... If you plan to do something similar, check out the "simpler hardware solution" near the end of the page below first!

 
Circuit Diagram with some original parts removed plus rear light & front lights switching transistors added:

The new PNP transistors enable or disable the positive voltage to the rear light LEDs and/or the forward lights, under control of the green and purple decoder wires. The decoder white wire is programmed to go negative when the locomotive moves forward and the decoder yellow wire is programmed to go negative when the locomotives moves in the reverse direction. Several original components serve no useful function in a DCC environment and have been removed to give more room for the new circuitry (Diodes D1 and D2,  plus the two tiny 220k resistors, R11 and R12). The large 470 ohm SMD resistors R13 and R14 are replaced with leaded parts to allow access to the circuit without breaking tracks. The parallel resistor pairs at the collector of each transistor are simply there as I didn't have the correct value 220 ohm and 330 ohm leaded parts.

 

End 1 and End 2 Cabs (my definition):

I'm regarding the cab with a driver in place as the number 1 end cab.  Hence the opposite end of the locomotive, with no driver, becomes the number 2 end.

 

Main PCB modifications:

Removing components not needed for DCC operation:

To simplify fitting the switching transistor parts, Diodes D1 and D2, large 470 ohm SMD resistors R13 and R14 plus the two tiny 220k resistors, R11 and R12 are all removed.

Fitting the switching transistors circuit:

Any general purpose PNP transistor will do the job. I used the BC557b.  The wiring arrangements on the main PCB are shown below. Not hugely tidy, but it'll do!

 

Unmodified PCB

PCB with unwanted SMD parts removed

New circuit in place

 

Adding cab lights:

The cab mouldings are withdrawn from the upper body shell. Small holes are drilled near the top of the cab rear walls. Warm white Lighthouse LEDs, are fitted with their leads inserted through the holes. The leads are formed to keep the LEDs pressed into the cab roof and the cab mouldings are refitted to the body shell. The positive leads of each LED are wired together, with a tail, to connect to the chassis assembly. A series resistor (10k) is soldered to the negative lead of each LED. The other end of the resistor is wired to a tail to join the white function wire at end 1 and to a tail to join the yellow function wire at end 2. The three wires are formed, ready to solder to the points on the main PCB indicated below. 

The blue common positive wire is connected to the collector of the transistor controlling the positive supply to the forward lighting, so the cab light only comes on when the front lights at that end of the locomotive are illuminated.

Wiring inside the body shell for the cab lights

The main PCB attachment points

(blue wire is soldered to collector of upper transistor in the circuit diagram)

 

Fitting the decoder:

The blanking board, used purely for DC analogue operation, has already been removed from the 21 pin connector on the main PCB. The new decoder is carefully aligned above the 21 pins before being eased down into place.  The decoder is programmed using the cv values indicated in the table below.

 

Chassis and body shell ready to re-assemble

 

Programming the CVs for the ESU LokPilot V4 Decoder:

Button zero activates the forward lights (via aux2 and headlight + rear light), while button 9 activates the rear lights (via aux1 and headlight + rear light).

 The colour sections must be programmed after cv32 is set to the value indicated in the 1st line of each colour section

CV Purpose

CV Number

Value

 

General

Sets to 28/128 speed steps (default is 14 speed steps)

0 + 2 + 0 + 0 + 0 + 0

29

2

Motor Control

These are the defaults, no delay = 0 max delay = 255...adjust as req'd

Acceleration delay

3

32(def)

Deceleration delay

4

24(def)

 

Lighting control

Set index variable (cv31) to 16

31

16 (def)

 

Mapping of function buttons:

Set index variable (cv32) to 2 (and leave for all beige shaded area)

32

2

Map White & Yellow plus Purple to button zero    

White wire + Purple (end 1 Forward only)   on via button zero

330

1+8=9

Yellow wire + Purple (end 1 Reverse only)  on via button zero

346

2+8=10

Remove unwanted Green & Purple button mapping defaults:    
Cancel green wire defaults 362 0
Cancel green wire defaults 378 0
Cancel purple wire defaults 394 0
Cancel purple wire defaults 410 0
Map White & Yellow plus Green to button 9:    
Set index variable (cv32) to 3 (and leave for all blue shaded area)

32

3

White wire + green (end 1 Forward only) map to button 9

362

1+4=5

Yellow wire + green (end 1 Reverse only) map to button 9

378

2+4=6

 

Set function definitions:  

(Note these did not in fact require attention as their default conditions were OK)

Set index variable (cv32) to 0 (and leave for all orange shaded area) 32 0
White mode (dimmable H/L) 259 1(def?)
White brightness (max) 262 31(def?)
White undimmed LED 263 128
Yellow mode (dimmable H/L) 267 1(def?)
Yellow brightness (max) 270 31(def?)
Yellow undimmed LED 271 128
Green mode (dimmable H/L) 275 1(def?)
Green brightness (max) 278 31(def?)
Green undimmed LED 279 128
Purple  mode (dimmable H/L) 283 1(def?)
Purple brightness (max) 286 31(def?)
Purple Double strobe LED 287 128

 

 

Programming the CVs for the TCS EU621 Decoder:

Note: Re-mapping limitations mean button zero is forward lights (white & yellow plus purple) while button 1 is rear lights (white & yellow plus green).

TCS function mapping only allows buttons 0 to 6 to control white and yellow function wires, hence the need to change the rear light button from 9, as used on the LokPilot, (to button 1).

CV Purpose

CV Number

Value

Lighting control

White wire (end 1 Forward only)   on

49

0

Yellow wire (end 1 Reverse only)  on

50

16

Green wire (forward and reverse)  on (aux 1)

51

32

Purple wire (forward and reverse)  on (aux 2)

52

32

Button Mapping

White wire map to buttons 0 & 1 (forward)

33

1+4=5

Yellow wire map to button 0 & 1  (reverse)

34

2+4=6

Green wire map to button 1 to enable rear lights (aux 1)

35

4

Purple wire map to button 0 to enable front lights (aux 2) 36 1+2=3
BEMF
config

61

1

     
     

Motor Control    (starting point for optimisation)

Acceleration delay

3

16

Deceleration delay

4

12

 

 

Outline Test & debug strategy:

The loco is checked for basic movement as loco number 3 in its default state. If basic functionality is confirmed, the lights are checked and non default cvs are programmed. The lights are then rechecked. Next, the dynamic movement properties are explored and optimised. Finally, the definitive address is programmed.

When any issues are identified, these are studied until understood and then resolved.

 

Issues with ESU Decoder equipped Class 57:

 

There were no ESU control or functionality issues, in fact all went remarkably well in the decoder department.  But.....looking more critically at the Bachmann marker lights, they are very yellow, rather than warm white and their relative intensity varies in an undesirable way with the angle of view. A diffuser, made of white tissue paper, folded to give 4 paper thicknesses, was fitted between the Marker light LEDs and the rear of the lens light pipes. This improved matters quite significantly and also dropped the light intensity to a more believable level. It's still very yellow, but I'll do a bit more research before resorting to a change of LED devices. (A swop to "prototype" or "warm" white LEDs might be necessary.)

Also, when taking the test photographs, it became apparent that the Bachmann locomotive had a distinct list to starboard, which could not be cured by manipulation of the bogies.  Removing the bogies revealed a Bachmann quality problem on the number 2 end bogie. One of the two screws holding the top central bearing moulding to the bogie was missing, leaving the bogie bearing part at a very slight angle and causing the list to starboard. Fortunately I had a screw of the right size, so was able to fix the problem. We are now back on an even keel!

 

Headlight and marker lights at the forward end

(+ cab light) switched on/off by button zero.

(Also switch on button 9 for trailing end rear lights)

The lights have saturated the camera

The markers are more yellow than the picture suggests

Rear lights at the trailing end

switched on/off by button 9.

(Also switch on button zero for forward end lights)

The rear lights have saturated the camera

The lamp centres really look red

 

A second DRS Class 57 was modified in a similar way, but fitted with a TCS EU621 decoder. The second loco was renumbered to 57012. The two locos were then trialled on a 2m straight test track to compare performance.

       

Issues with TCS Decoder equipped Class 57:

Again, there were no TCS control or functionality issues (apart from the need to use button 1 for the rear light switching). The same fix was applied to the second Class 57 to improve the behaviour of the front marker lights.

The second locomotive sat upright on its bogies OK. (This time, all the relevant screws were in place.) However, another quality problem was apparent, even during DC testing. As the speed was increased, at a critical point around 40% of top speed, an undesirable metallic vibration could clearly be heard. This was quite tricky to pin down, but appears to have been due to one of the motor terminal electrical contact tabs vibrating against the plastic under frame, which acted as a sound board, amplifying and distributing the sound. I had to strip the loco down as far as the metal chassis block incorporating the motor installation. The motor contacts were then bent and wedged with pieces of plasti-card so that there was no longer any contact with the plastic underframe. The motor capacitors and the now redundant rear light switch were also removed, but it is not thought that these were part of the noise generating mechanism. The motor shafts were lightly oiled and the loco re-assembled. There is still a hint of a noise increase, but the effect is now largely masked by the normal gear train sound and is now considered acceptable.

The final digit of the loco number (1) was painted over and replaced with the number 2, hand painted and tidied up when dry with the odd touch of background colour. The side numbers are just about OK, but the front and rear panel numbers are not to be studied too closely!

 

ESU - TCS Decoder comparison:

 

Test ESU Decoder Class 57/0  57011 TCS Decoder Class 57/0   57012
1 Slowest smooth running

Identical default minimum speed

Identical default minimum speed

2 Control feel during slow running Very smooth transition between speed steps with default CV3 & 4 cvs 3 & 4 need to go to circa 16&10 to achieve smooth speed step transitions, but that extends accel/decel time (& hence distance) of TCS loco compared to the ESU loco at speed. 
3 Mid speed running Smooth, with shorter braking distance than TCS To get smooth slow speed behaviour, deceleration  distances get a bit big at medium speeds
4  Fast running No obvious issues, but need bigger test track to properly check this out. Long accel/decel time and distance becomes excessive.
5 lights intensity steadiness

No problem

No problem

6 CV programming issues No physical indication on loco of a successful programming event.

Important issue for simple programmer users! (Less of a problem if your programmer can also read cvs)

Loco responds to a successful programming event with a brief forward movement.

It's not possible to map buttons higher than 6 to basic white/yellow decoder wire operation so I could not exactly duplicate the ESU control arrangements and had to use a different button for the rear lights (1 used instead of 9).

7 Setting short address Bachmann Dynamis failed to programme the short address using its "change address on the programming track" routine. Instead it was necessary to program CV1 with the new address as a normal cv change action. The Dynamis routine programmed the new short address in the TCS decoder without any problem.
8 Programming efficiency Looks more complicated, but defaults cover a lot of the basic lights options. Fewer cv settings are required to set up the basic lights options on the TCS decoder, but  the ESU solution appears slightly more flexible.
9 Documentation Have to use the English manual available on the Web, as the manual included in the decoder pack is German only. The English manual is comprehensive but has several numerical errors. Only top level info appears in the outline (English) manual in the pack. Need to access web based info for e.g. button mapping info etc. No documentation errors noticed.

 

 

 

 

Comparison Test Conclusions:

Motor control seemed rather better using the LokPilot decoder. The ESU decoder exhibits very smooth speed step transitions. To get even close to this behaviour using the TCS decoder, CV3 and CV4 needed to be set high enough to make braking and acceleration distances a bit on the long side, when at speed. It is probably possible to overcome these issues on the TCS decoder, but an in depth knowledge of the more subtle coding options for the TCS part becomes necessary, whereas the ESU LokPilot 4.0 works extremely well straight out of the box.

In general, I found the TCS function output programming easier to get to grips with than the ESU, although neither presented any significant problems. However, a big plus for the TCS decoder was how it reacts to a programming event: by a brief forward movement of the motor drive. This gives immediate confidence that the programming event has been actioned. No equivalent feedback is provided by the ESU decoder. Also, I had to re-programme the (short) address for the ESU decoder by directly changing the cv (CV1), as my Bachmann Dynamis seemed unable to programme the address in the normal way (which works fine with TCS decoders). (This anomaly is strange as I believe ESU designed the Dynamis unit!) 

For the Class 57 project, the motor control is the deciding issue, as the very simple function programming requirement was successfully implemented on both decoders. So the TCS decoder goes back to the parts store and a second ESU decoder goes into 57012.

 

The Class 57/0 locos 57012 and 57011 on the test track

 

A Much Simpler Hardware Solution for the Lights:

Once again, with the benefit of hindsight, an alternative approach becomes apparent :

Because both the ESU and TCS decoders have direction options incorporated into the function programming, it is possible to take an approach that requires very few changes to the original locomotive circuitry, with no need to remove parts or to add extra switching transistors:

In fact, the only changes required are as follows:

1) Make sure that the rear light switch on the underside of the locomotive is in the "on" position, or solder a link on the main PCB between tags C and D to permanently set the switch to "on".

2) Disconnect the orange wire at Tag A on End 1 of the main PCB and connect this wire instead to pad "Aux 2" on the main PCB.

3) Disconnect the orange wire at Tag A on End 2 of the main PCB and connect this wire instead, via an extension wire, to pad "Aux 1" on the main PCB. (Use sleeving to insulate the joint). 

4) Cab lighting can be added as shown above, except that the blue wire is connected to the end of the parallel pair of large 470 ohm SMD resistors, nearest the decoder.

 

Photoshop diagram illustrating the changes

 

This leaves a circuit diagram as shown below:

 

The revised CV programming for The ESU LokPilot4 decoder would then be as follows:

CV Purpose

CV Number

Value

 

General

Sets to 28/128 speed steps (default is 14 speed steps)

0 + 2 + 0 + 0 + 0 + 0

29

2

Motor Control

These are the defaults, no delay = 0 max delay = 255...adjust as req'd

Acceleration delay

3

32(def)

Deceleration delay

4

24(def)

 

Lighting control

Set index variable (cv31) to 16

31

16 (def)

 

Mapping of function buttons:

Set index variable (cv32) to 2 (and leave for all beige shaded area)

32

2

Map White & Yellow to button zero

 

 

White wire (end 1 Forward only)   on via button zero

330

1 (def)

Yellow wire (end 2 Reverse only)  on via button zero

346

2 (def)

Remove unwanted Green & Purple button mapping defaults:

 

 

Cancel green wire defaults

362

0

Cancel green wire defaults

378

0

Cancel purple wire defaults

394

0

Cancel purple wire defaults

410

0

Map Green & Purple to button 9:

 

 

Set index variable (cv32) to 3 (and leave for all blue shaded area)

32

3

Green wire (end 2 Forward only) on via button 9

362

4

Purple wire (end 1 Reverse only) on via button 9

378

8

 

Set function definitions:  

(Note these should not in fact require attention as their default conditions are OK)

Set index variable (cv32) to 0 (and leave for all orange shaded area)

32

0

White mode (dimmable H/L)

259

1(def?)

White brightness (max)

262

31(def?)

White undimmed LED

263

128

Yellow mode (dimmable H/L)

267

1(def?)

Yellow brightness (max)

270

31(def?)

Yellow undimmed LED

271

128

Green mode (dimmable H/L)

275

1(def?)

Green brightness (max)

278

31(def?)

Green undimmed LED

279

128

Purple  mode (dimmable H/L)

283

1(def?)

Purple brightness (max)

286

31(def?)

Purple undimmed LED

287

128

Button zero activates the front lights and button 9 activates the rear lights

 

The corresponding TCS CV programming (allowing button 9 to be used to switch the rear lights this time) is as follows:

CV Purpose

CV Number

Value

Lighting control

White wire (end 1 Forward only)   on

49

0

Yellow wire (end 1 Reverse only)  on

50

16

Green wire (forward only)  on  (aux 1)

51

0

Purple wire (reverse only)  on (aux 2)

52

16

Button Mapping

White wire map to buttons 0  (forward)

33

1

Yellow wire map to button 0  (reverse)

34

2

Green wire map to button 9 to enable rear lights (aux 1)

35

0

Green wire map to button 9 to enable rear lights (aux 1)

37

16

Purple wire map to button 9 to enable rear lights (aux 2)

36

0

Purple wire map to button 9 to enable rear lights (aux 2)

38

16

BEMF

config

61

1

 

 

 

 

 

 

Motor Control    (starting point for optimisation)

Acceleration delay

3

16

Deceleration delay

4

12

Button zero activates the front lights and button 9 activates the rear lights

 

I've now updated 57011 using the ESU LokPilot decoder, in line with the simple hardware approach and updated CV values described above, and it works just as it should!.....

 

 

Nuclear Flask Train with the two 57s in consist.

 

Military Flask train with commercial flask carrier standing in for the 16 wheel navy waggon....... (pending!)

 

Supplier website links:

 

Rails of Sheffield    The Class 57 units above were originally purchased from this excellent e-shop.
Southwest Digital Ltd    Helpful U.K. source for ESU LokPilot decoders.
Bromsgrove Models    Excellent source of TCS DCC decoders and specialist LED devices for this type of project.
TCS (Train Control Systems)    A U.S. DCC decoder specialist company with comprehensive programming advice on their website.
Rapid Electronics of Colchester    A good reasonably priced source for electronic components.

 

The photos of the model were taken using a hand held Canon Ixus 220.     The photos of the real 57012 loco were taken back in 2008 at York Station.

 

Click to move to Model Rail Index Page