Class 158 Supplementary Page covering the addition of sound, passenger compartment lighting and hazard lights to the new-tech lighting Bachmann 158 pair.
The Class 158 is a good test vehicle for the addition of a sound system. While the unit is dismantled, the opportunity will also be taken to add working hazard lights and cab lights. illumination of the passenger accommodation will also be attempted.
Note that the smaller pictures can be enlarged by clicking on the image.
Northern Rail Class 158 at York, fitted with new LED marker / rear Lights
|The proposed sound system:
I have decided to use an ESU LokSound V4 decoder for the hardware and I am sourcing the unit from Howes Models of Kidlington near Oxford. Some research on YouTube revealed an impressive video demonstrating a Bachmann 158 fitted with the LokSound using Howes sound files. (I already have good experience of the ESU LokPilot decoder motor control, and this is duplicated on the LokSound decoders.) In order to maximise the low frequency output, I have also obtained a 4 ohm base reflex speaker assembly from DCC Supplies as a substitute for the small circular speaker that comes with the decoder.
I was originally rather sceptical of the value of sound systems in OO scale models, due to the limitations of the frequency response that is physically possible in a speaker small enough to fit into the locomotive body. However, a visit to Ely station revealed that the bass sounds of the locomotives become much less significant as the distance from the locomotive increases. Even for Class 66 freight locomotives, the low frequency rumble is barely discernable unless the locomotive passes very close to the observer.
|The first little problem:
I downloaded the manual for the Loksound V4 from the ESU website, to see how the function control CVs might be used to control the lights, without compromising the available sound options. I also want to be able to use the existing FL4 function only decoder in the trailer car if possible. However the Loksound V4 manual is in desperate need of revision and I had to get some help to find out just how flexible the function mapping can be, as long as you don't need to use all the possible 26 function buttons! Thanks to Bryan from Howes and Marcelo from Brazil for pointing me in the right direction. With the help of the ESU LokProgrammer software (free download from the ESU website) the buttons and function outputs can be remapped to meet the requirement, except that to add passenger compartment lighting, I need to use 5 outputs.... So the trailer car FL4 has to return to the spares box, to be replaced by a LokPilot V3 FX 6 function decoder.
The Howes Class 158 programming of the LokSoundV4 decoder allows for lighting control via buttons 0, 11 and 12. The out of the box button mapping is shown (copied from the Howes website) below:
The proposed lighting solution is shown below. (This replaces an earlier scheme using switching transistors to interface with the unmodified Howes 158 LokSound V4.)
Operation of the lights is as follows: Button 0 provides day running lights. Button 11 provides night running lights, including the leading cab light. Button 12 activates the passenger compartment lights. Button 4 activates the hazard lights and the doors opening sound sequence.
|The proposed arrangement
for lights control:
|Programming the CVs:
The LokSoundV4 decoder in the power car will require the following cv updates with the Howes cv values as the starting point:
The LokPilot V3 FX function decoder in the trailer car will require the following programming, with the factory defaults as the starting point:
|The Circuit diagrams:
|The Power Car:
Dismantling the car reveals that the speaker and decoder pcb will not quite fit into the roof underside. So the first step is a little surgery to remove the problem areas of the side mouldings.
Next, the hazard light apertures are drilled out and shaped with needle files to accommodate pre-wired amber chip LEDs (DCCC Nano-lights). The chip LED wires are then passed through the aperture til the chip is almost in place. A piece of Blu Tack is used to hold the chip within the aperture, where it is fixed with a small drop of super glue. When the glue has dried, the exposed LED diffusers are painted with orange water colour paint. Connecting the two LEDs in series, around 10K of series resistance looks OK in low light conditions, but needs to be rechecked in daylight!
Both LEDs with a 10k series resistor under the kitchen lights
The 158 cab is full of LED bodies and resistors, with no model driver in place, so a cab light may not be entirely appropriate, but I'll fit an Amber chip LED in the roof of the cab, to try and generate a suitable glow. Around a 10k series resistor looks OK in low light conditions.
The LokSound decoders have 6 function outputs, but the Aux3 and Aux4 outputs source very low current, requiring an additional buffer amplifier to provide a useful output. ESU sell a small interface PCB with a 21 pin DCC socket and the two amplifier stages to provide normal Aux3 and Aux4 output levels. Some additional wires need to be connected to this PCB before it is ready to mount in place.
Underside of ESU Interface PCB with wires in place
The decoder board and the speaker are fixed in place using double sided foam pads. The wires are then connected as shown in the power car circuit diagram. Heat shrink sleeving is used to insulate the solder joints. The wires are held in place under the roof, using insulating tape and Blu Tack.
With the power car chassis housing the central motor, flywheels and gear train bogies, the idea of providing night lighting in the passenger compartment is a bit dubious. So for the moment I'll put this on hold. So.... Solder the wheel pick-up wires (black & red) plus the motor wires (orange and grey) to the cleared PCB above the forward bogie and re-assemble the upper body shell to the chassis.
Time to see if it works!
|Programming, Testing &
A few issues to deal with first: The sound chip master volume control is cv 63. The Howes default value of 135 was eventually reduced to 25! The dogs are much happier now and the wife has stopped glaring at me ! Next problem.... The motor is wired back to front, so disassemble top from bottom and swop the grey and orange wires... done, problem solved.
Now, programme the changes to the LokSound decoder cvs: Great! the hazard lights come on O.K. with the door open sound! BUT... night running rear lights don't work at all...... Re visit the LokProgrammer software on the computer and ah ha!.... a new version of the software has just been released. Download the new version and recheck the function mapping..... Two cv changes are now required to the reverse F(11) operation. Magic! we now have working reverse night running rear lights! Now looking good!
But rear light red glow is noticeable through the cab side windows.....I can't get at the top of the LEDs to provide local shielding without separating the front of the car from the body.. so try reducing brightness. CV 31=16, CV32=0, change CV270. Dropping from 31 to 5 gives a more realistic level of rear light intensity and reduces the impact of the light leakage in the cab quite a lot.... that'll do!
Only remaining major issue is the cruising speed engine sound, which has a low pitched whine I've not noticed on 158s before..... could be a resonance effect due to the bass reflex speaker? I need a longer test track to get to the bottom of this one, as we're still in the middle of our kitchen makeover, waiting for the new worktops (fitting scheduled for next week) My test track is only 5ft long! The sound levels could also do with a bit of a tweak. Perhaps push the master volume up then pull down the engine sounds, to make things like the horn and flange scraping a bit more conspicuous? Hours of fun a bit later on! First I'll modify the trailer car........
|Trailer Car Modifications:
With a fairly realistic looking seating arrangement (albeit minus tables) the trailer car is completely uncluttered with engine, flywheels and loudspeaker. So this will be the the first of the two cars to be fitted with internal lighting, using a Brelec light bar system.
The original FL4 decoder was first removed and the underside of the car roof was painted white, in the hope of improving the light levels with a light bar active.
Trailer car top and bottom, with white painted roof underside
A new ESU 6 function LokPilot FX decoder will be fitted. This time, I'll try to mount it vertically between the front doors, to leave the roof as clear as possible for the lighting bar. First I need to get the hazard light and cab light LEDs into position. These will be fitted in the same way as described above for the power car.
Light bar, decoder, hazard lights and cab light fitted
The Brelec Lightbar was constructed and fitted in place with double sided foam pads. Two white LEDs beam down a circular section light pipe, with a white reflector behind. A board is provided with a bridge rectifier and adjustable current source, so that the unit can be powered directly from the DCC fed track contacts. However this is not needed in my application, where the unit is powered and switched via the decoder. A 220uF 25V capacitor is also provided to form a reservoir at the output of the bridge rectifier. I'll keep this ready to fit across the lightbar LED connections in case of flicker problems. The LEDs needed circa 20mA of current to generate a useful light output. I hope this is within the LED spec. but no figures are given in the Brelec kit, so if the LEDs die, I'll have to look for some more appropriate replacements!
To avoid obstructing the light bar, I mounted the LokPilot FX decoder on the back of the cab rear bulkhead, using a double sided foam pad. The wires are a bit of a jumble, but hopefully they are predominantly hidden in the roofspace.
Close up of front area
The track wiper connection PCB has been removed from the chassis, to avoid a mechanical clash with the decoder. The red and black decoder track connections will be soldered directly to the wheel wiper wires, using unshrunk sleeving to protect against shorts.
|Programming, Testing &
Bug Fixing on the trailer car:
After programming with my Dynamis controller, the trailer car lighting operation was exactly as planned, so no fixes required so far, except for the removal of some finger prints on the inside of a few windows (no doubt generated during hazard light aperture filing). The light bar gives a brighter glow at each end (possibly due to inadequate covering of the LEDs by me) but the overall effect is not too bad for 20mA. The internal lights are not very noticeable in daylight, through the tinted windows, so I'll have to wait until this evening to evaluate the appearance in dusk conditions. Then, if all goes well, I can start to plan equivalent illumination in the power car...... ongoing.................... Not good news! Its now after dark and the car is not really bright enough in the centre, while it's a bit over-bright at the ends.
However, fitting passenger compartment lights definitely looks a worthwhile addition..... if its done well! See below for plan B!
Oh yes, the trailer car would also benefit from winding down the intensity of the rear lights. (Same issue as spotted in the power car)..... CV113 to around 3?
|Plan B Lighting on the trailer car:
I've been pondering how to make an effective anti-flicker solution (as this was also a very obvious problem with just one trailer car bogie equipped with pick-ups). So plan B uses initially 4 nano-light LEDs (arranged in two pairs) bouncing their light off the white painted car roof. Each pair of LEDs will be powered via a 9mA constant current source and a pair of 1000uF back up capacitors will maintain power during any brief outages. (see below for the circuit.) I'll add extra LED pairs with additional current sources if this proves necessary to give a more uniform light.
Plan B lighting circuit 18mA total load for 4 LEDs & about 1Second of anti flicker protection
Just realised that the transistor bases could have been connected together and one set of resistor and 2 diodes could have been omitted!
When first tested, the anti flicker circuit performed spectacularly well, with approx. 2 seconds+ of constant light when power was removed........ looked too good to be true, and of course, it was! For test purposes I was using an old Hornby analogue DC power supply "12V DC Output". This gives an average DC output of around 16V which is usually near enough to the DCC decoder 14.2V for setting up LEDs. However, when you stick 2000uF of capacitor on the end of it, the capacitors charge up to the full 22V peak output voltage from the full wave rectifier in the power supply. So the discharge time for the caps was just about doubled! I've now invested in a small bench DC power supply, which I can set to 14.2V and sure enough, 1 second is the discharge time at around 20mA down to circa 6.5 volts when the 2 series LEDs start to drop out. Hopefully it will be enough.......
The 4 LEDs face towards the roof, behind each of the cross beams
The new lights circuit at the end of the car, with a pair of 1000uF 25V capacitors
(The pillars at the rear of the chassis had to be cut back a little, to avoid a clash with the capacitors.)
The flicker-free circuit works really well, but since I've cleaned the excess oil and fluff from the rear of the trailer car pick-up wheels, the power continuity has improved dramatically!
Next, is it possible to fit similar lighting in the power car?
lights for the Power Car:
First thoughts are to apply a capacitor supply protection system to the LokSound decoder and use 2 pairs of lighting LEDs with 1k series resistor per pair, which can be replaced by constant current sources if necessary to improve anti-flicker performance. The decoder capacitors should hold up the function output long enough to provide anti-flicker protection. After fitting the caps, the car roof was covered in white paper to diffuse the LED light and the speaker enclosure was painted matt white. The anti-flicker capacitor circuit is now shown in the power car circuit diagram above. The capacitors are kept charged via a 220 ohm resistor and in the event of a brief break in the DCC signal, the capacitor continues to power the entire decoder circuit via the diode, for a few hundred milliseconds.
Caps fitted, white speaker enclosure & paper covering the roof
(The internal side shelves within the roof were cut back to enable the caps to sit further into the roof to avoid clashes with the metal pillars in the power car chassis. The caps were glued to the body shell bulkheads and a plasticard insulator was glued on to the capacitor ends to ensure that no accidental contact was made with the metal chassis pillars)
The pre-wired daylight white chip LEDs are glued on to suspended roof beams fabricated from plasticard:
Suspended roof beams with LEDs fitted
The beams are glued in place and the wires formed and soldered with 1k resistors in series with each series LED pair:
The fourth LED is mounted on the speaker enclosure.
Night running and hazard lights on
Hazard lights off & we're clear to roll
Hmm..... better than no lighting at all in the power car, but the light levels are no where near as even as in the trailer car. (Although the motor and flywheels don't help!)
The good news is that the lights are completely stable in both cars, without any hint of flicker, as the train moves.
I think I'm going to have to cook-up a better power car lighting distribution arrangement though........
I've added a couple of extra LEDs beside the speaker to fill in a dark spot and then dropped the LED current all round via a combination of hardware (series resistor values) and software (adjust CV 294) to even out the lighting to approximately match the trailer car levels and to reduce any highlights from the motor and flywheels. Its a a balancing act to get a general level of subdued lighting without highlighting the internal motor mechanics too much, and I think I've probably got an acceptable compromise.
Power Car Passenger Compartment lighting Circuit
I've also reduced the headlight and marker light levels to a more believable intensity, using CVs 262 and 278 in the power car and CVs 114 & 115 in the trailer car.
I think "that'll have to do"
|Back to the power
car.......Beginning to adjust the relative
In addition to the master volume control (CV63), each sound slot can be independently adjusted in level. However, the Howes defaults for all the sound slot levels, except the master volume, are at maximum. This is no problem if you want turn down one of the sound options.... just one CV to adjust. However, if you want to make a sound slot louder e.g. flange scraping, you have to turn down the eight other sound slots & wind up the master volume!
I've established which CVs to tweak.....
I'll probably be tweaking the levels for some time to come, so I'll close the page at this point.
Further comment: The cruising speed engine sound is not quite the same as my recollections of the real East Midlands 158s at Ely Station. So I tried a 158 decoder from another source, which I believe was closer to the real trains in that respect, but unfortunately, it did not have realistic engine sound level changes as the loco powers up upon departure from the platform, so I've gone back to the Howes decoder, despite misgivings about the cruising speed engine sound.
Was the addition of sound a worthwhile feature?
I would have to say that on a diesel multiple unit, the answer has to be YES!
The sequence of sounds corresponding to a station stop certainly adds a lot of interest to operation of the train..... doors open and hazard lights on; pause; door warning buzzer sounds and doors close; hazard lights off; guard's whistle; drivers buzzer acknowledgement; maybe a two tone horn, then engine revs build; brakes release and the train moves smoothly away. Now with that sequence automated via a computer programme, it would definitely add an extra dimension to layout operation.
After a further look & listen to "the other LokSound decoder" prior to making a short video, the engine sound changes are so subtle that they are barely noticeable...... so back to the Howes decoder once again.
However, when dismantling the power car, the front panel glue bond gave way...... so I've taken the opportunity to add a driver, a bright white cab light (not used during driving) and to plug the gaps around the LEDs that leak red light into the cab during reverse operation, using black paint. The new cab light will be independently controlled using the available Aux4 output, via button 20. Once this is all checked, I'll re-bond the front panel to the upper body shell.
A white nano-light with 2k7 series resistance is used for the new cab light, glued next to the existing orange cab light, which still provides "instrument glow" during night running.
For Aux4 to switch on via button 20: Using function programming line 26..... CV32=3 ; Check CV F = CV406=16 (OK); CV k = CV410=32
Northern 158 number two, with the latest modifications
A YouTube video showing the 158 on my PC controlled test track can be seen by clicking here
|Supplier website links:
The photo of the real East Midlands Class 158 was taken at Ely during April 2010. The Northern Rail Class 158 was photographed earlier, at York. The work in progress photos of the model were taken on the kitchen worktop using a Canon Ixus 220HS.