Adding a Sound System to a pair of DRS Bachmann Class 57/0 locomotives
This page describes the addition of a sound system to the Class 57/0 Bachmann locomotive. The original page was created as I undertook the project and included a number of blind alleys before I eventually homed in on a final solution. This summary attempts to provide a more direct route to the end result.
The Class 57 loco pair double heads a nuclear submarine fuel flask train. The number two loco (57012) is the first to be updated, followed further down the page by the lead locomotive (57011).
Class 57/0 at York station in 2008.
|A New Decoder type:
This exercise gave me the opportunity to try a Zimo sound decoder for the first time: Digitrains sound files fitted to a Zimo MX644D 21 pin decoder.
|The Speaker Location:
There are two options that I can see. A DCC Supplies 20x40x8mm sealed unit can be fitted under the fan grills in a purpose built speaker bay, facing upwards. However, the bass response and power handling of this speaker is not an ideal match for the Class 57. To get a bigger speaker on board, requires modification of the fuel tank/battery box, where the drive unit for an ESU 50334 speaker can be accommodated facing down towards the track, with the fuel tank/battery box forming a sealed enclosure above the speaker. This is the route I decided to follow.
Showing the fuel tank/battery box centre bottom
ESU 50334 4 ohm speaker glued around the edge of the aperture, forming an air-tight seal
Viewed through the top of the fuel tank/battery box, with 1.4mm plasticard filler piece ready to be bonded in place & sealed.
First the speaker wires will be soldered in place (when the glue has fully cured)
(The wires emerge through holes drilled in the filler piece, to be routed through the wiring tunnels of the main casting.)
circuit for the first Class 57 to be modified:
Several of the original Bachmann parts, not required for DCC operation only, have been removed. Extra connections are made to the 21 pin DCC connector, to provide access to the Aux3 and Aux 4 function outputs. All positive connections for the end lighting PCB assemblies are connected in parallel on the modified main PCB. The negative connections are controlled via the decoder function outputs to provide independently switchable directional lighting at each end of the locomotive. A single DCC controllable cab light is installed in the cab that includes the driver. This is referred to as the "number one end". At the number two end of this loco, a miniature connector is fitted to enable a bit of fun in the adjacent KUA involving a bright green LED.
New thin enamelled function wires connected to the 21 pin connector.
The rewired main PCB with a Zimo MX644D decoder module installed.
The original "E" tabs at the ends of the main PCB are no longer needed. They have been isolated from the PCB tracks and used as anchor points for the new function wires.
The Class 57 re-assembled with the new speaker and Zimo decoder on-board.
Sound Operation with the Digitrains programmed Zimo decoder: (Version 2.0 Digitrains Class 57 software)
The new speaker arrangement works rather well with quite a good bass response for such a small speaker. I'm not familiar with the sound of real Class 57 locos, but the output from the new decoder certainly sounds plausible!
After a temporary hiccup with earlier software, the loco was fitted with a decoder using Version 2.0 of the Digitrains Class 57 software. The button mapping for the sound options exactly corresponds with the published list on the Digitrains website. The first CV that requires attention (to maintain a neighbour friendly sound level) is the master volume control on CV266, with value reduced to 16.
Zimo Motor control :
Initially, the motor control using the default settings in the Zimo decoder, appeared to work very well. However, latterly I began to notice a slight judder at speed step 2/28. Reading the manual, CV9, which controls the sampling rate and period time for the back emf feedback system, has a default value of 55, with suggestions for some changes for 3 pole motors exhibiting jerky behaviour. When I checked the value of this CV, it was actually set to zero. Changing CV9 to 87 stopped the problem. Movement is now smooth from the minimum speed step upwards. The acceleration and deceleration delay setting on CV3 and CV4 give a very lifelike impression, although both will probably need reducing somewhat for computer control.
One quick win was to successfully add shunting mode and momentum override mode to the locomotive via CV124 (value 3), CV155 (value 19 for shunt mode via button 19) and CV156 (value 18 for momentum override via button 18).
Zimo Motor and sound
synchronisation.... first thoughts :
Running light engine at lower speeds, the engine revs and brake sounds work well together. The motor revs and volume build up just before the loco moves away.
Running light engine at half speed, is not so impressive, as with a 2 meter test track, I have to bring the loco to a stop after only a few seconds of running, but unfortunately, the engine sounds continue to notch up, while the loco is actually decelerating and continue to run at excessive revs long after the loco actually comes to a halt. This does not seem at all prototypical! Using the F6 button which is intended to bring engine sounds back to idle, does not significantly reduce the time taken for the engine sounds to resume idle operation. This may not be the major show stopper it first seems, when operating on a continuous track, where abrupt stopping, immediately after a max acceleration start is an unlikely scenario. Also, I understand that in the longer term, this may no longer be an issue, with the possibility of an improved SW fit. (fingers crossed)
Zimo Sound and Function
My decoder has been pre-programmed using Zimo's "extended function mapping", With CV61 set to 97, an arrangement in which, although the manual indicates only buttons 0-8 are available to control the hardware function outputs, in fact, I can confirm that CVs 43-46 also work within this regime, to enable buttons 0-12 to be used. As received, all these buttons are already assigned to hardware or sound functions on my Zimo decoder, so I first need to move two sound functions to currently unused buttons, to free up 2 more lighting buttons.
To e.g. re-assign the sound on button 9 to Button 15 (and to bring the volume of that sound down) Zimo use a technique described as Pseudo-programming. With the loco on the main track, "programme on main" is selected on the controller. CV300 is programmed with a value corresponding to the button number you wish to add the sound to (in this case 15). The controller is then returned to normal operation configuration and by selecting appropriate controller keys, it is possible to search through all the available sound slots in the decoder (by listening to them on the loco speaker) and then select the required sound. Repeated pressing of the F0 key regenerates the selected sound, enabling the loco speed control to be used to adjust the volume level to the desired setting. The new key assignment and volume level are then saved by pressing the F8 button. After returning the speed control to zero, normal loco operation can be resumed.
Many thanks to Paul Chetter, who kindly pointed me in the right direction within the on-line manual for the above!
Zimo sound slot mapping for
the Digitrains Class 57 Decoder code:
Going through the sound slot reassignment procedure above, revealed the locations of the accessible sound slots:
Group 1: Spirax valves; Engine rev up & down; Primer Pump; Flange squeal; Air Brake
Group 2: 3 tone horn; single tone horn
Group 3: Air release; door slam; compressor
Group 4: Brakes on
Group 5: engine start; engine stop; engine idle; ramp up to notch 2 (n2); n2; ramp down n1 to idle; ramp up n2 to n3; n3; ramp down n3 to n2; ramp up n3 to n4; n4; ramp down n4 to n3.
During the Pseudo-Programming process, key 2 moves left to right through the active group. Key 1 moves right to left, but also switches the motor sounds on and off (bit of a pain).
key 4 moves up to the first sound in the previous group, key 5 moves down to the first sound in the next group
(so its not essential to use key 1)
Zimo also offer a mute facility that can smoothly reduce the overall volume to zero over a programmable time period. This compares with LokSound's fader, which has a programmable level shift, but a fixed ramping time.
The planned lighting controls:
(1) Directional lights at end one, via button zero.
(2) Directional lights at end two, via button 8.
(3) Cab light at end 1 (the only manned cab) via button 9.
(4) Connector at end 2 to enable some hopefully interesting? effects in the front nuclear flask carrier, enabled via button 10.
Appropriate combinations of (1) to (3) above provide a complete operational lighting capability.
A Table Summarising the proposed mapping changes:
The areas in blue require change action
The corresponding CV updates:
Improving the cab lighting:
The cab light is used in the computer controlled engine start-up sequence, but the original illumination level is a bit on the low side. So the inside of the cab roof was painted white and the current through the cablight LED was increased to 4mA........ This had the desired effect.
Summary of primary hardware mapping process available on
Zimo enhanced mapping of physical function outputs
Set CV61 to 97 to activate this regime…….
For multiple outputs, add the values together
To add Directivity: Use special effects CVs first 2 bits:
(If special effects are also required, add to the above, higher bit values as required)
Additional flasher option
CV117 Defines the on-off times: tens=on, digits=off range 0-99 where 0=100ms & 9=1s
So e.g. 42= 500ms on and 300ms off periods.
CV118 Defines which outputs are affected: Bit0=H/L; Bit1=Rears; Bit2=Aux1; Bit3=Aux2etc to Aux4
Bit=0 means no flashing of that output; Bit=1 means flashBit6=1 makes inverse flash on Aux2, Bit7=1 makes inverse flash on Aux4.
The Navy Flask train
Running the Zimo equipped 57
under computer control:
The loco has been adjusted to set a max scale speed of around 75mph. It proved a bit tricky to achieve the accuracy required to stop with a KUA wagon just clear of the points.
Zimo fitted Class 57 with the latest KUA update in tow. With only a centimetre or so to spare, stopping locations are critical on the small test track.
Updating the second Class 57
to a similar standard:
1) Hardware Updates: Circuit Diagram......
Hardware changes in place (just waiting for the glue, used to bond down the marker light diffusers, to dry, before re-assembling)
The CV changes are identical to those used on the first loco. (Except for the address.)
Good news, initial testing indicates that operation appears to be exactly in line with the revised mapping plan!
Both 57s are now operating correctly and in a similar manner, using the V2.0 Digitrains Class 57 software.
Running both the modified
Class 57s as a double heading consist
under computer control:
The two 57s double heading in consist on the test track.
Using the Traincontroller software to set up the schedule for this "trainset", both lead loco functions and functions common to both locos (e.g. engine and wheel flange noises) can be incorporated into the schedule. (Lead loco by default and common functions using the function forwarding command.)
So far, so good!
Excellent! I've now got a start sequence programmed in: Front driver's door slams, front cab light comes on, front loco engine starts. A few seconds later, the rear loco engine starts. The front lights of the lead engine are switched on and the rear lights of the rear engine switch on. The cab light goes out, the triple horn activates, the engine revs & sound level picks up on both locos and the pair smoothly pull away to run through the simple route sequence on the test track. While rounding the curve on the return run, the engine revs drop to idle and the wheel flanges squeal on both locos. Then when parked up at the end of the run, the lights are extinguished sequentially and the engines turned off.
The maximum scale speed is set to 25mph for this run, keeping the locos below the speed at which the second notch-up is triggered. It sounds very good to me!
I've made a video of a test run that doesn't quite do the locos justice, but here it is: Please click here
One subsequent change: Latest photos on Google images show that white LED lights now appear to be the standard source of illumination for the front marker lights, so I have substituted new pure white tower LEDs for the original yellow types used in the Bachmann model. I also had to fit some additional layers of tissue paper between the new LEDs and the rear of the marker light lenses to get the intensity down to the required level. (An easier option than increasing the values of the surface mount series resistors on the lighting PCBs)
........And that pretty much completes the project!
Oh yes.....and there will hopefully be the opportunity to re-blow the decoders with some really special Class 57 software in the not too distant future!
|Supplier website links:
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.