Bachmann Derby Lightweight - Interior Lights

[Nigelcliffe]

First coach and the lights, single test LED will be dropping the voltage. 

The test LED will drop approx. 3v regardless of how its used.  With only 235ohms resistance (two 470's in parallel), the current through the single LED will be very high, so glows extremely bright, and likely to burn out if left for any time.

 

The collection of LEDs in the coach lights may be in series with each other, which means they drop a different total voltage (some of your measurements suggest 14v).  Or there may be other internal resistors in the coach lights (again sees a different total voltage drop over the coach lights).     So exactly how that behaves will be different.  

Safest is the largest resistor which gives acceptable coach lighting brightness. 

 

 

 

 

For the button control of the six-pin, I think you will want the "manual control" option, which is F2/F3 key.  One key is "forwards", the other key is "backwards".   As LEDs only light in one direction, it depends which way round you wire the lights to the decoder (one lighting contact via resistor to Orange, other contact to Grey).  So you do have a choice of function key.

To use the motor outputs you MUST use both wires to the orange/grey, and disconnect both sides of the coach lighting from other power sources (pickups, etc..).

I'm not sure the decoder blue helps at all.   You definitely don't need the decoder blue for the button control of motor trick (blue mixed could lead to decoder blowing its motor outputs).     Blue has some value if fitting a stay-alive to stop the trailer lights flickering, but that's all.    

 

 

- Nigel

[gordon s]

Managed to take the trailer car apart and the board as expected is different. I started to map out the circuit, but then realised it wouldn't be needed if I'm isolating the two springs and feeding with the grey and orange wires. The voltage across the two springs as it stands is track voltage (18v).

 

I assume changing CV133 to 102 should drop the voltage back down to around the 14.4v of the power car. Without thinking, I stuck an  led across the two springs and couldn't fathom out why it wasn't working. Of course the higher voltage was blowing them as soon as I touched the springs....Doh!

 

The biggest problem is how to program the trailer car as the ECoS won't read the decoder without a motor load. Everything I've read says to put the power car on the programming track as well. I can understand that, but surely then changing the CV's such as 61 and 133 will change values on both chips.

 

The challenge is that the six pin socket is very small and there are no easily accessible wires on the Zen decoder. Pins 1 and 2 are not connected to any tracks on this PCB as it's in a trailer car with no motor. I can possibly solder two wires to the back of the socket, but the contacts are very close together and even my smallest, most pointed bit will still be difficult to access to attach wires. My thoughts were to add a motor to allow me to program it and then remove the motor to connect the two wires to the springs.

 

What do you guys do to programme trailer cars with their own chips and no motor?

 

On the upside the second and third DMU should be easier as I'll know what I'm doing.....

Edited April 8, 2020 by gordon s

[Nigelcliffe]

Programming a six-pin, either a temporary motor, or a temporary resistor of around 100ohms.   In both cases, remove when done.

 

Yes, LED without resistor goes pop very rapidly !  Ooops, we've all done it.    LED with 1k resistor should be OK.

 

 

Nigel

 

[Yardman]

TCS among others do function only chips. Here you put a resistor normally around 100ohms across one function for programming. These decoders have their own address but don't require a motor.

[gordon s]

Getting nearer...

 

Managed to hook up a grey and orange wire to the decoder socket and cut the tracks to the springs. Hooked up a motor to the grey and orange wires and programmed CV61 and CV133 and then desoldered the wires to the motor. Wired in a 1kΩ resistor on one wire and hooked them up to the springs. Placed a test led across the two springs and switched F2 on and it lit up......Whoopee!

 

Had one problem with the first decoder though. Put it into the socket in the trailer car and then onto the programming track to test and set up via the ECoS. All seemed OK and the message came up 'decoder read' as normal, but then the two front lights came on and the motor went off at full tilt. I've tried a decoder reset with CV8 but no different. Spoke to DCC Concepts and they've suggested doing a CV8 reset whilst it's going haywire. Any other ideas?

 

If all else fails this one can go back. No smoke or burns etc, just like a runaway.....

 

The second decoder does what it should, so I've reached the point of putting both roofs on.....Scary.

 

Out of interest, can the front headlights be dimmed via a CV? They really are very bright right now.

 

I'm guessing if test led's light up OK, there's no reason why they would blow up the roof leds. Agreed?.......

 

Edited April 8, 2020 by gordon s

[gordon s]

It's been a long few days for what initially I thought would be an easy task, but job done and it all works as it should......

 

I can now pull into the bay with all lights blazing, let the passengers off and F1 turns off one set of carriage lights and F3 the other.

 

The additional 1kΩ resistor has also brought the lights down a tad, so very happy.

 

Many thanks to Nigel et al for all your help, suggestions and support. Much appreciated.

 

Fresh with that success, I'll have a go at the 105 and 108 now.....

[Nigelcliffe]

Most decoders have CVs to alter the brightness level of lights.   Matter of wading through the manuals to find the relevant settings. 

 

I know its a DCC Concepts Zen in the trailer, but lost track of what's in the power car.  

 

DCC Concepts manual suggests CV64 is a global dim setting, and values of 1,2,3,4 might work.    Or its more wiring changes and increase the resistor values. 

 

 

- Nigel

[gordon s]

It’s a Zimo MX630 in the power car and knowing Zimo, almost everything is adjustable. Off to check the manual.......

[Nigelcliffe]

Zimo has loads of brightness adjustment.  CV60 is the global dimming setting.  If you are only trying to change one light, then CV60, plus the appropriate values in the "dimming mask" determine what get adjusted.

 

 

[gordon s]

I found CV60 in the Zimo manual so put both cars on the programming track and adjusted CV60. No problem with the Power Car and the lights are great. I expected the Zen decoder to do the same and this Is where I really was surprised.

 

I always thought CV numbers were set by a standards organisation and CV60 on one decoder would do the same on any decoder. I now find that CV60 on the Zen decoder is related to ABC braking and nothing to do with lights.

 

I must be honest this was a huge surprise to me as I thought every manufacturer’s CV functions would be the same.

 

That takes me full circle now as I’ll need to set up the trailer car separately and that brings into question the whole motor bit again.

 

What a real pain.....

[Nigelcliffe]

You've just learned:  a handful of the critical CVs for running locos are set in the standard, a further small group are "advisory" (so meant to follow them, but makers don't always do it), and the rest are up to manufacturers to do as they please.

 

Once you've got the trailer in the ECoS as a "loco" (or whatever it is in its interface), then you could use Programming on the Main to alter CV's in that car alone - you only really need the programming track stuff to set the address.  Just remember to lift the power car off the track whilst programming the trailer car (assuming both have the same loco address). 

 

 

There are ways to do decoder locking, which allows one decoder to be Programming-on-the-main changes whilst not affecting the other decoder.   But, usually its more trouble than its worth, and just lifting the item not to be programmed is far simpler. 

 

 

- Nigel

[gordon s]

Tried programming the trailer car on main this morning, but changing CV64 to any setting between the range of 0-20 made no difference at all to the headlights. I reset it to 4 as the default. I've double checked everything in terms of the right loco code etc, but suspect it wasn't actually doing anything.

 

I'm toying with the idea of buying an ESU decoder tester as it would certainly save time and hassle in the long run. Connecting a motor up again in the trailer car would mean dismantling it all again, including the programming that has set the motor output as a function, plus removing the diode and supply wiring.

 

One thing did occur to me though. I have Decoder Pro set up with my ECoS so perhaps I could programme CV's via that, but suspect I will come up against the same problem with no motor installed.

 

Edit: As you suggested earlier this might work with both cars on the programming track.

 

https://www.jmri.org/help/en/manual/DecoderPro/Multi_Decoder_Control.shtml

Edited April 9, 2020 by gordon s

[Nigelcliffe]

ESU tester is a handy device, and certainly solves the programming without a motor issue.  Whether you have enough decoder faffing in your life to justify its ownership I can't say.    I have one, it works well. 

 

I can't say how the DCC Concepts decoder responds to CV64. 

 

JMRI/DecoderPro.   Provides a better/different user interface for programming, but you're still limited by what your ECoS can do.  So it doesn't fix the motor issue for programming track reading.   ( Think of JMRI as having an expert button-pusher who will rapidly push buttons on the ECoS for you.  )

 

Multi-decoders and decoder locking.  Only go there if prepared to do a lot of reading around the topic.  Different decoder makers do decoder locking in different ways (often quite subtly different).   I'd be very cautious with the JMRI tool until you'd understood which CVs it is changing, and that CV's match up to what your decoder maker's have chosen to use.  (Without checking, I suspect its designed for one or two brands of US decoders, and whether it works for a wider range is unsure ).   

 

An alternative approach for programming the trailer cars is to a "I don't care" decoder in a loco which you sit on the track at the same time.   The "I don't care" loco gets programmed with all sorts of garbage settings (which make sense for the trailer car), but you don't care about that as you decoder-reset the "I don't care" loco after finishing.  

 

 

- Nigel

 

[gordon s]

Thanks for the reply Nigel. Decided in the end to buy the ESU decoder tester as with numerous loco's to set up and possibly refine in the future, I'm sure it will pay for itself. I'm always a great believer in having the right tools for the job as they make life so much easier. If they don't get used as much as you expected, then second hand value on eBay reduces the financial hit.

 

I did think about a 'don't care' loco, but as we said yesterday, once you get beyond the basics, It's a minefield as different manufacturers may use the same CV for different options, so unless the two decoders are similar it could be a problem.

 

Setting up the decoder first on the ESU seems a better option. With any luck it will be here in a few days.

 

Spent the day doing two locos. The first a Hornby B1 took 10 minutes, Undo two screws in the tender, plug in decoder and job done. I could have taken out the capacitors, but all seemed fine. I suspect modern locos and modern decoders are far less of an issue re capacitors.

 

The second loco is a Bachmann J39 with a split chassis. Must have been at it for a couple of hours and so far I have had to drill and tap holes to take power feeds to each half of the chassis and grind away part of each chassis half to provide room for the decoder. It's all done and working, but still tight, so I'll split it again tomorrow and grind a bit more away. The body has very little clearance around the split chassis, so you have to make room, whichever way you can.

 

You can guess which gave the most pleasure....

[gordon s]

Just to close the loop, the two car unit is now working perfectly......

 

Whilst working on my Ivatt 4, I found an old chassis, that had no motor, but did have a decoder fitted up front. Another rummage and a motor without a worm gear appeared, which was wired up to the decoder and surprisingly it all checked out on the programming track. Not having a worm gear connected is a plus as the loco doesn’t move up and down on the programming track. Thanks to all who suggested a dummy loco where any programming of CV’s is irrelevant.

 

I can now programme the trailer car without problem, so set about dimming the front lights again. CV64 still didn’t change anything despite what setting was used. Mick (Newbryford)  of this forum kindly told me that you need to change CV 49 and 50 first. Once CV49 was set to 12 and CV50 to 24, changing CV64 to 1, brought down the lights to a much more acceptable level.

 

Very happy with the outcome, so thanks again to everyone. Hopefully the 105 and 108 will prove to be similar. Once you know which tracks to cut and what to connect, it’s actually quite a simple job for anyone to do.

Edited April 12, 2020 by gordon s

[beejack]

Just noticed this thread.

 

I have just modded my own Derby Lightweight.  I put a lenz standard + v2 in the motor trailer and replaced the 6 pin in the dummy trailer with a standard + v1.

 

The green wire for both was connected to the V- light spring and the track to the light spring in each was cut using a 1mm drill in a pin vice.  Using a slightly oversized drill bit ensures the track is cleanly cut. 

 

As the lights were originally fed directly from the track connections, they already had diodes and resistors in the positive circuit so no need for any resistors on the green wire.  

For the dummy I just taped up the blue cable as not needed and the other 6 wires were attached to a modified 6 pin blanking plug.

 

Now I know it all works I have ordered a zimo m617n as the lenz is a little big for the space.

 

As for lighting functions, I left the front lights for each decoder on F0 but reversed the direction for the dummy so the white lights illuminated in the direction of travel , changed the tail lights to F1 for the front car and F2 for the rear car and the interior lights are on F4. 

Used JMRI Decoder Pro to program the chips but I did have to program each car seperately, Then I gave them both the same address.  I didn't try to read the Lenz V1 just wrote the changes using Decoder pro.

 

[beejack]

Having now received my Zimo MX617N pin pin chip I sat down and dismantled the dummy car again.

 

I removed the lenz chip and inserted the zimo and soldered the interior lighting cable from the lighting spring to the chip.

 

To my suprise the interior lighting wouldn't work.  The only way I was able to get it to work was to cut the V+ track and solder a second cable from the common positive tab.