Bachmann Derby Lightweight - Interior Lights

[halsey]

58 minutes ago, Nigelcliffe said:

 

 

Zimo publish an extensive decoder manual, with diagrams, page55 onwards in my English language version.

 

The convention on 6-pin and 8-pin is that pin1 is orange.  

 

 

 

I have it but it didn't help me (with my very limited techy ability) with non-wired decoders

 

All good now though...…………….

 

I don't intend any new locos for a long while now once I get my last one sorted, which is an LMS diesel, when the replacement 21 pin decoder comes back - as with my increasing knowledge I'm sure it is faulty

[halsey]

On 01/01/2020 at 13:40, Nigelcliffe said:

 

Motor unit first:   no lights means one of two things.   First, the decoder could be fitted backwards - putting an 8-pin in backwards will give motor control, but no lights, so try turning the decoder plug through 180 degrees.   Secondly it could be electrical contacts within the unit, ie something has become detached. 

 

Trailer unit:   On its own, it is unlikely to be read on the programming track (not enough current through lights to generate the "read" pulse back from the decoder).   So, you'll have to write values without reading, or extract the decoder and put it into a loco with a motor which has enough current to be read.   
One of the features of CV29 is to change which way is "forwards", which will solve the directional lighting issue by "fibbing" which way is forwards for that decoder. 
First, calculate new value for CV29 - use this tool I wrote about a decade ago:   http://www.2mm.org.uk/articles/cv29 calculator.htm    Once you've got the options you need selected, the choice of direction changes things by "1", so there are only two possible values you'll need to try out.

With just the trailer unit on the programming track, write the new value for CV29.  

 

Feeling very pleased with myself now - thanks to the above I've just managed to re-programme CV29 on 2 locos to alter direction so now my loco yard Jintys go in the same direction which helps!

[gordon s]

Search and you will find.....

 

I have my Derby Lightweight in bits on the table to look at switching the interior lighting. It comes off two springs either side of the decoder. These make contact with copper springs in the roof.

 

 

Having checked the voltage across these contact springs, it's showing as 14.4v. I did think of cutting the tracks to these springs and then hard wiring, but don't want to b*gger up something else. Second thought was to add insulation tape to the springs and wiring the blue/geen wires directly into the roof via a miniature 2 pin connector.

 

Having checked the voltage across the blue and green wires for F1 it's showing as 17.9v. Is this going to be a problem or will I have to add a resistor to drop the voltage?

 

The last thing I want to do is blow the LED's.

Edited April 14, 2020 by gordon s

[Nigelcliffe]

If the lights are originally wired to the track (pickups), then they are getting full track voltage.  There is probably a rectifying device in the lights to convert the input to DC (and to deal with running on analogue, when the DC will reverse depending on direction). 

 

I don't know how you are measuring track voltage, but if just using a multimeter, unless its a very posh one, it won't give an accurate value.  

 

So, my view, is you'll be fine on the green+blue wires, provided you have fully insulated the contacts from the pickups.   (But were it me, I'd have measured the current drawn by the lighting strip to be confident of the current needed from decoder ).  

 

 

Nigel

 

[halsey]

all good stuff I' learning from this - thanks

[gordon s]

Thanks Nigel, much appreciated. I was planning to take two wires (green and blue) directly from the decoder to the roof contacts. The only way to get to the roof contacts is via the vertical spring strips and I will put insulating tape on the ends to prevent that happening. I could detach them completely, but didn't really want to do that as that limits any possible sale to a DC buyer.

 

I have a RRampmeter I bought years ago, so I could dig that out to take measurements. Those I have quoted were via my Wavetek DM9.

 

'Scuse my ignorance, can you show me how to connect to check the power draw of the LED's. Is it just in series between the two spring contact strips.

 

 

Edited April 5, 2020 by gordon s

[gordon s]

Well this is really strange. I checked the voltage across the two rails with my RRampmeter and that came in, much the same at 18v.

 

Using the RRampmeter across the two spring terminals on the Power Car, the voltage was 8.5v, not the 14v as measured from my Wavetek.

 

The green DCC light was illuminated.

 

I then plugged in the two pairs of leads into the RRampmeter and connected it in series between the two sets of spring strips and it read 8.5v and 0.01A with the roof lights illuminated.

 

I guess the final stage will be to check the Function output across the decoder with the RRampmeter to see if that is also vastly different from earlier readings.

 

Still hoping I can connect the blue/green wires directly to the roof lights, but until I get a true reading of the output voltage, I'm reluctant to hook them together. 

 

 

[Nigelcliffe]

I've not looked inside the Bachmann unit, so, this is how I'd approach it without having handled one:

 

Looking at the PCB photo, there are a collection of diodes, possibly resistors, and capacitors in the middle.  It should be possible to track the tracks from the sprung pickups back to these items,  that will tell you which are in the circuit with the roof lights.   If its just diodes and capacitors, then nothing to worry about - they're just to get the DC current the "right" way round for the roof lights, the capacitors are anti-flicker. 
If its resistors as well, then need to think a bit more, they may be current limiters for the lights, so adding equivalent to your direct wiring may be needed. 

 

I'm not sure what your RRampmeter tells you when its not reading a DCC signal. 

 

Current consumption of 0.01A means that isn't a worry for the decoder output. 

 

 

- Nigel

 

[gordon s]

Morning...

 

Managed to strip it all out again and I'm in the process of buzzing it through. Here are the components.

 

471 = 470Ω 5%

224 = 220kΩ 5%

152=1.5kΩ 5%

 

 

Had a thought last night (if the circuit will allow) and that was to cut the track before the components that feed the springs and then run the blue green wires back to the decoder.

 

Would that work?

[Nigelcliffe]

Yes, cutting the track feeding the springs would work, and would keep all resistors (etc) in circuit with the lights.  You need to trace tracks very carefully, and only cut the one feeding the lights.  The little round rings probably indicate a track swapping sides from top to bottom of PCB.

If one side of the PCB tracks comes back to loco pickups, then you could leave that in place, and just break the other track.  Feed that track from the function output (green), and the positive comes from the track (ie. don't need the blue).  This is completely normal, its how N-gauge stuff is usually wired, and will be in the decoder manual.  Downside is the lights may flicker, if they do, break the track to pickup, and feed it with blue. 

 

D6 looks like a diode feeding one side of the spring contacts.   But not clear where the other spring contact ends up - I'm guessing the ring just off shot to right of C5 may connect to ring to left of D5, but that's a guess. 

 

 

 

[gordon s]

It's OK Nigel, I started my working life on PCB design, so am familiar with feed throughs or plated through holes. Just hope I can remember what I learned 50 years ago.....

 

This appears to be the circuit to the lighting springs. Does that make sense? 

 

 

 

I'm enjoying the detective work.....

 

[Yardman]

I answered this in a previous question on coach lighting which is copied below. The original roof PCB needs to be altered so control of interior lights can be passed to the DCC decoder. The picture shows a class 108, but a Derby Lightweight would be similar.

 

 

For a start mine aren't battery powered, it's harder to find space for a battery and switch, multiple LED's do start to draw a bit of power. Mine are DCC powered operated by function only decoders, the coaches operate in sets so I can turn all the lights in a set on or off with a single command. The accompanying picture should explain it. I use small SMD (Surface Mount Device) LED's Usually one to a compartment, wired in series sets of three with a current limiting resistor. The picture shows a DMU power car which only required three LED's and one resistor. Marker lights and route indicators required others. This was all driven from the loco sound decoder. The trailer car is similar, but used a Function only decoder with pickups on both bogies. Remember LED's are polarity conscious so install them the right way round.

One thing to note. The coach interiors need to be painted to prevent light leaking through the plastic. I give mine a coat of black then a coat of an appropriate interior colour. The interior colour has a bearing on the light colour so experiment. I use “Warm White” which I think most represents incandescent lamps. “White” looks more like modern strip lights and “Yellow” more like oil or gas lamps.

[gordon s]

Thanks for that Yardman. As always, I was hoping there would be a simple solution. Where does battery power come into it? I know you say yours aren't battery powered, but surprised that could be a solution.

 

A most impressive set up.

[Nigelcliffe]

This is what I'd do with that diagram - cut one connection (perhaps just insulate at the spring contact) marked with red X, and connect to the decoder function output via a resistor of between 470 and 1K ohm.    The positive comes from the pickups via D5 and D6.   There is a chance that D5/D6 are the reverse orientation for this to work, in which case, break track leading to D5/D6, and connect wire from L contact to one of the pickups.   

 

There are other options, such as connecting the L spring to decoder green, and relying on the path via R3 to Pin 7, but not sure of direction of diodes marked and how they interact.  So my thinking is to remove D1, D2 and R1,R2,R3, R4 from consideration.

 

 

 

Edited April 6, 2020 by Nigelcliffe

[gordon s]

Thanks Nigel. I'll take another look tonight/tomorrow and add the directions of the diodes.

 

What wattage would you recommend for the resistor? 1/4W?

 

That all seems too simple.....

[Nigelcliffe]

You measured 0.01A at the lights.  Assuming 15v (which it is give or take), then Watts = V*A = 0.15W, which is less than a 0.25W resistor.

 

I think the diodes D5/D6 will be the wrong way round, so the bypass necessary.  

[Yardman]

3 hours ago, gordon s said:

Thanks for that Yardman. As always, I was hoping there would be a simple solution. Where does battery power come into it? I know you say yours aren't battery powered, but surprised that could be a solution.

 

A most impressive set up.

The person that originally started the post was trying to add battery power, but if you have DCC and thus power on tap there's not much point.

[gordon s]

Morning all, 'tis me again.....

 

I appreciate you know far more than me re electronics, but I'm trying to learn without making too many mistakes, so bear with me....

 

Picking up on your second option, Nigel. Here's the full circuit diagram re connections to the springs. There are other components to control the directional led's, front and rear, but I've left them out of this diagram.

 

If the blue common wire brings into play all the resistors etc, can I simply isolate the connection of the left spring from D5 and D6 and just connect the green wire directly to that spring?

 

I'm naturally being very hesitant on this project as i'm on the edge of my current learning curve and much as I'd like to switch the internal lights on and off, I don't want to blow them all as no spares are listed.

 

 

 

Edit: I'm currently working through the whole circuit and that may have a bearing on the connections. I'll be back.......

 

 

Edited April 7, 2020 by gordon s

[Nigelcliffe]

Yes, from what I can see, just going green to Left Spring should be OK.    The diode voltage drop (over D5 or D6) will be inside the decoder, so the volts should be near enough the same. 

 

If worried, start with green via series resistor of approx. 470ohm to 1kohm, that will ensure its not over-current, lights I expect to come on, but not full brightness.  If they are bright enough, leave resistor in place.

 

I'm slightly puzzled by two 470ohm resistors in parallel (R3/R4), meaning 235ohm resistance over them. 

Edited April 7, 2020 by Nigelcliffe

[gordon s]

Here's the complete circuit including all the lighting. I've just checked R3 and R4 again and they are definitely wired in parallel.

 

Does this whole circuit make sense? I keep looking for a second connection to the front led's.

 

Thank you Nigel for all your help on this. Very much appreciated. It will be great if it all works....

 

 

[gordon s]

Well the good news is that nothing has broken, but they are not working either....

 

I cut the track before the left hand spring and connected a 1k resistor fed from the green wire, reassembled the unit and put in on the programming track. The existing direction led's work and act correctly with a change in direction and the motors is fine. I cut the green wire away from the 8 pin plug and soldered it directly to the 1k resistor.

 

Rather than put the roof back on and risk blowing up the led's, I thought I'd take a voltage measurement across the two springs, hopefully looking for around 8.5v. This is where it's gone haywire. My RRampmeter is simply flashing continually. Both the volts and the current display plus the DCC and AC lights. As it's still on the programming track, nothing has blown or burnt out, but something clearly is wrong.

 

Any ideas?

[Nigelcliffe]

I assume you've checked that the track is definitely not connected to the pickups (ie. the break is complete). 

 

With the green wire connected, but no path between the spring contacts (no lighting strip),  and the coach on a DCC track,  then you may see DC volts if the function key associated with green is active.    I'd be using a multimeter to check this, not the RRampmeter (because I don't know what that will show me for DC volts).   
Alternatively, if you've used the series 1k resistor, then a LED could be put across the spring contacts, with its positive wire towards the blue, negative to the green.  That should light when relevant function key for green is turned on.

 

 

I'm not sure what your system is, and how makers implement programming tracks varies.   However,  a basic programming track has no voltage present except during programming pulses, which are very short.  

 

 

Edited April 7, 2020 by Nigelcliffe

[gordon s]

You're a star!.....

 

No short between track and spring. Tick.....

 

DC measurement of Function 1 to the springs.....14.4v across two spring contacts. Tick....

 

Put a LED across the two springs and it lit......

 

Switch F1 on and off and the LED went on and  off.....

 

Question 1. Before I put the roof back on,  is 14.4v OK for the led's in the roof?

 

I assume taking the resistor down from 1kΩ to 470Ω will increase the brightness.

 

Thanks for sticking with me on this Nigel. I couldn't have done it without your help. Feeling a real sense of achievement to sort this.

 

Question 2.... and that's the trailer car, which should have a 6 pin decoder. They arrived today. I have yet to take the trailer car apart, but will need the decoder to switch the head and tail lights.

 

I assume the options are to fit the decoder as normal, but cut the tracks to the two springs and then run two wires, blue and green from the power car and connect them to the trailer car springs.

 

What would you suggest I do?.......

 

The 6 pin diagram is here....I have a DCC concepts Zen Decoder.

 

 

 

 

 

[Nigelcliffe]

Sounds like it's working.

 

The 14.4v measured is with nothing over the contacts.  If you put your single LED there and measured whilst lit, it would be 3v to 4v (because that's the "forward voltage" of the LED), the remainder of the 14.4volts being dropped over the various resistors in the circuit.   
So, in terms of your coach lights, with the resistor in place there is no way you can be over-current (you didn't blow the single LED!), but will have to adjust resistor downwards (possibly remove and only have the two parallel 470ohm originals) to get the brightness.  But see what you get going down in resistance in steps. 

 

 

 

Six-pin trailer cars.  With a 2-Function decoder, you can't officially do the interior lights as well as the cab-end head/tail lights.   BUT, as DCC Concepts are still copying TCS features,  you might be able to....    In a TCS decoder there is an option to change the motor output to be controlled by a function key (rather than the usual response to the speed packet on a throttle).   Connect the coach lights (with appropriate resistor in series) to the decoder motor output, and control by the function key.    Reading the DCC Concepts manual, the option is there, limited to FnKey2 or FnKey3.     Might be easier than wiring lights to the trailer if the choice of function key for interior lights is acceptable ?    There's also a CV for setting the motor "speed" which would be a proxy for "lighting brightness" for the function key operation.   
If you go that way, you need to wire motor outputs to both the lighting contacts, with enough resistance in series to protect the LEDs from over-voltage  (speed setting doesn't change voltage, it changes the "on time" of the PWM output at full voltage, averaged over time the average volts are lower, but a LED responds instantly, so gets hit with the short bursts of full volts).

 

Otherwise, if doing it via a jumper wire,  yes, what you describe works.  Or, if there is another spare function output on the decoder (say brown or whatever colour), then might choose to use that instead.   But, in this case will need to ensure that there is a connection on one-side to the track as a substitute for the blue wire, unless you're also going to move blue across (unnecessary to jumper blue as well in my view). 

 

 

 

- Nigel

 

 

 

 

 

 

 

 

[gordon s]

Morning Nigel....a great day yesterday, so I wonder what today will hold in store. Thanks again for the info.

 

Checked the test led again this morning as I wanted to see what effect changing the 1kΩ resistor would make. Once I could see it would change the brightness, curiosity got the better of me, so I tried it without the resistor completely. The test led was full brightness but it had the effect of pulling down the directional led's to something like half power. Switch F1 off and they came back again to full brightness. I wasn't expecting that, but I'm not an electronics guy. Is that because the test led was taking more of the  power when there was no resistor in circuit to limit it?

 

I found a very comprehensive draft manual on the web for DCC Concepts decoders. A veritable treasure trove of info....

 

https://www.dccconcepts.com/wp-content/uploads/2019/02/Zen-2019-Manual-DRAFT.pdf

 

You were correct about using the motor control and the bits that refer to that are here.

 

 

I will take apart the trailer car this morning to see the circuit board. Not sure what to expect but I'll map it out as per the Power Car to see if that provides some clues. 

 

Edit: Just read your reply again and you say grey to one and orange to the other. I guess this means isolating both springs. It's not clear which one will be F2 and which F3, but I guess that will become clear once wired.

 

I'm quite happy having F1 for one car lights and F2 or F3 for the second as that will mean no inter car wiring.

 

The power or brightness of the motor function can be set here...

 

 

As the voltage across the springs on the Power Car was 14.4v and the track voltage 18.0v, this would suggest 80% of power, so if I set the value to be 80% of 128 (102), I should be somewhere in the right area.

 

DCC Concepts also talk about using the 'stay alive' wire on the 6 pin decoder to act as a blue wire. Not sure it helps, but let's see how today goes. They talk about the red/black track power only providing half power, so maybe there is a benefit to use the substitute blue wire. Isolate both springs and then the stay alive wire to one spring and get or orange to the other via a similar resistor.

 

Edited April 8, 2020 by gordon s