Heljan O gauge & Zimo MX696KS/LOKPL96KS

[Ray H]

I've been tasked by a colleague with the fitting of a Zimo MX696KS to a couple of his O gauge Heljan diesel locos. The Heljan PCB is to be replaced by the Zimo LOKPL96KS (which came with the decoder) so that there's minimal (if any) soldering.

 

I've made a start on "Falcon". This runs OK (on a 36" length of track) and the sound also seems OK. However, the front & read lights don't work. I can but assume that the lights did work before I removed the Heljan PCB.

 

I've transferred the wires from the "R LED" terminal on the Heljan PCB to the Rear Headlight connection on the LOKPL96KS, the "F LED" wires to the Front Headlight terminal on the LOKPL96KS and the four wires in the Heljan PCB's "COM" socket into the Zimo's Ground socket.

 

I have yet to replace the body but I can't see anything thereon that might impact on the front and rear lights.

 

Can anyone suggest what I may have done incorrectly before I turn my attention to the Class 45?

 

I haven't yet transferred the wires to the roof fans to the LOKPL96KS as it looks as though I need to cut the plug off the lead. I presume that the black wire will go to one of the "Ground" connections on the LOKPL96KS and the red wire will go to the Smoke Fan terminal on the LOKPL96KS. Is my presumption correct?

 

These are the first Zimo MX696's that we've used; previously all of his other Heljan diesels were fitted with ESU decoders.

 

Any guidance would be appreciated.

 

Thanks.

[Ray H]

There are a number of what I believe to be 0.33 ohm resistors on the Heljan PCB but my limited electronics knowledge suggests that they don't play any part in the supply to the LEDs, for example to reduce the voltage to them down from 12V. There are a couple of diodes which I presume are used to determine the direction the loco is travelling so that the appropriate lights are illuminated.

[Nigelcliffe]

The normal wiring of lighting outputs on a decoder will have the decoder function output as "Negative", and the "common" to the lights should be connected to the Decoder Positive (typically blue wire on a wired decoder). 

 

If your mentioning of "decoder ground" is correct, then you're probably going to the wrong places. 

 

 

( And you don't reduce voltages to LEDs with resistors.  You limit the current flowing through the LED with a resistor.    Its a fundamentally different thing !!). 

 

 

- Nigel

[Ray H]

The normal wiring of lighting outputs on a decoder will have the decoder function output as "Negative", and the "common" to the lights should be connected to the Decoder Positive (typically blue wire on a wired decoder). 

 

If your mentioning of "decoder ground" is correct, then you're probably going to the wrong places. 

 

 

( And you don't reduce voltages to LEDs with resistors.  You limit the current flowing through the LED with a resistor.    Its a fundamentally different thing !!). 

 

 

- Nigel

 

Thanks Nigel.

 

I did wonder if it was simply a case of swapping the connections over but I didn't want to risk doing any further damage if I changed them over. I will now try swapping them around and report back.

 

Apologies for the bad terminology. I have always thought that the resistor through which the current flowed had a voltage drop across it otherwise the LEDs were effectively being connected to a 12V supply. This probably accounts for why I can never get my head around electronics.

[Ray H]

A straight wire swap hasn't made any difference, I'm still in the dark so to speak.

 

However, on the basis that the Heljan PCB connections were to R LED, F LED and COM - with R LED & F LED (presumably) being more positive than COM, I wonder whether having now swapped each light's wires around, the wires that I previously had connected to R LED and F LED (and which were more positive than the COM connection on the Heljan PCB) should be connected to the LOKPL96KS's Common Positive rather than the LOKPL96KS's Ground connection. That way they would still be more positive than their other wire.

 

Is that possible?

[Ray H]

It is a pity that the loco uses SMD LEDs (and several of them) as otherwise it would be reasonably easy to experiment in the knowledge that you could replace any blown conventional LEDs quite easily.

 

Anyway, within the above constraint I have now undertaken some voltage measurements using a digital test meter. The first set of measurements are with the meter's +ve lead on the LOKPL96KS's Common Positive and -ve lead on the relevant LED terminals:

 

Loco Direction Forward, Lights (Switched) On: Front 9.91v, Rear 0.34v

Loco Direction Forward, Lights Off: Front 8.81v, Rear 0.35v

Loco Direction Reverse, Lights On: Front 8.82v, Rear 10.68v

Loco Direction Forward, Lights Off: Front 8.83v, Rear 0.33v

 

The other set of readings are with the meter's +ve lead on the LOKPL96KS's Ground connection and -ve lead on the LED terminals:

Loco Direction Forward, Lights On: Front 0.15v, Rear -2.00v
Loco Direction Forward, Lights Off: Front -0.45v, Rear -2.14v
Loco Direction Reverse, Lights On: Front -0.45v, Rear 0.14v
Loco Direction Forward, Lights Off: Front -0.45v, Rear -2.13v

 

The significantly higher voltage readings when using the LOKPL96KS's Common Positive suggests to me that this is where the connection should be made. However, readings in the region of 10v are still much higher than a LED usually appreciates so I'd welcome support for this approach before I jump in with both feet and risk blowing the lot - if I haven't done so already.

 

I should add that all the above readings were taken with the leads to the LEDs still connected to the LOKPL96KS's Ground connection and the relevant LED terminals which may be the reason for some of the seemingly stranger readings.

[Nigelcliffe]

Your measurements on the Rear Light terminal makes sense.  But your first table on the Front Light doesn't make sense, it should be flipping between about 11v and about 0v (similar to the rear, but opposite on direction of travel). 

 

 

 

You need to establish (a) whether the LEDs have an appropriate resistor in series with them, or if you need to add one, and ( b ) which of the leads is positive and negative. 

 

For that, I'd disconnect the decoder (IMPORTANT!!!), and then use a battery (PP3, 9V) with a series resistor of approx. 2kOhm in series from one battery terminal, and work round the wires to establish which is positive and which negative to make the LED illuminate.  That helps with wiring them the right way round to the decoder.   The resistor is to ensure that the LEDs are not subjected to too much current; if there is a resistor in the loco as well, they LEDs should glow, but perhaps dimly. 

 

Next, deciding if the LEDs already have a series resistor.  You may be able to see one in a circuit board and identify it, or you have to start some measurements to find whether its present.  

 

Then it should be possible to connect things up. 

 

 

 

As I said before, its not the volts which matter to LEDs, it is the current.  You ALWAYS need a series resistor with a LED to control the current.  The question is whether the maker has left one in the lighting between the wires you have and the LEDs or not.   As I don't have a wiring diagram or manual,  I'd be looking for the resistors presence or absence.      

  

(  LEDs quick explanation:  A LED will always drop a fixed voltage over its terminals when illuminated. This is the "forward voltage" and is in the LED maker's spec sheet.  Its typically around 3v.   A LED also has a maximum current before it begins to burn out, again in maker spec sheet, but typically its 0.025A.   Connect an LED to more volts than its forward voltage and it lights up.  How brightly is determined by the current; the LED will suck all the current it is offered until it goes "pfffff" in a bright brief flash.    To limit the current to below the maximum, a resistor is added in series, calculated by Ohms law (V = IR ).   We can work out V, it is the volts from the decoder function output  (typically track voltage minus 1V), subtract the LED forward voltage (assume 3), and that's the volts over the resistor.  So, at 16v track voltage, V is 16-1-3=12.   The max current we want flowing is the LED max current, assume 0.025A.  Move equation round to get the minimum value for R=Volts/Current, and it comes out as 480ohm.  In reality, I'd start a lot higher (2Kohm) as a LED burning at max output is way too bright, and likely to fail if our estimates are a fraction wrong.   

)

 

 

 

- Nigel

[Ray H]

Nigel

 

Thanks again. I'll try your suggestion tomorrow - I'm busy this evening.

 

However, since penning my previous post (for which I am grateful for your further response) I have experimented with a 5mm Green LED - the first I came upon - and a 1k resistor in series wired between the LOKPL96KS's Common Positive connection and each of the LOKPL96KS's LED terminals in turn, adjusting loco direction and lights on and off as appropriate.

 

There was definitely an instance with the connection to each terminal when the LED was visibly alight, although not that brightly and when linked to one of the two LED terminals - I don't recall which - the LED was just alight regardless of either the direction setting or whether the lights were on or off.

 

The above was carried out with the decoder connected to the LOKPL96KS into which the various wires from the model are connected and of those wires one of each lighting circuit's pairs was connected to one of the LED terminals and the other of each lighting circuit's wire pairs was connected to the Ground terminal.

[Ray H]

I have investigated things a bit further. I know which pairs of wires operate front & rear LEDs and which wire of each pair is +ve (and which is -ve).

 

I had to disconnect the relevant wires from the LOKPL96KS to ascertain the above so I took the opportunity to check the voltages again (with the decoder in situ) using the Common Positive as +ve and the respective F LED & R LED as -ve.

 

Loco Direction Forward, Lights (Switched) On: Front 10.36v, Rear 0.12v
Loco Direction Forward, Lights Off: Front -0.23v, Rear -0.11v
Loco Direction Reverse, Lights On: Front -0.20v, Rear 10.56v
Loco Direction Forward, Lights Off: Front -0.20v, Rear -0.11v

 

As I indicated previously, the Heljan PCB has several 0.33 ohm resistors on board. Other than a couple of diodes, the supply connection for the (roof) fan and what I take to possibly be a couple of capacitors, there are no other components. This suggests that any resistors that are associated with the SMD LEDs are on the same board as the LEDs. However, the brightness of the LEDs when checking their wiring polarity with the PP3 & resistor combination suggested above wasn't that dim but nor was it excessively bright.

 

There are two red & two white SMD LEDs at each end and on the same PCB as the LEDs there are 4 SMD resistors with an in circuit resistance of 10k ohm.

 

I'm not sure what else I can try other than to connect up the various circuits to the LOKPL96KS using the Common Positive connection and see what happens.

Edited October 31, 2017 by Ray H

[Nigelcliffe]

The decoder measured voltages are now sensible, so that's working as I'd expect. 

 

I think you've got all the information to wire it up and it will work:  common positive lead and LED negative to each of the function outputs.   But, if wanting to be super-safe, include a series resistor with each of the lighting connections to the function outputs.     

 

 

- Nigel

[Ray H]

Nigel

 

The loco is now up and running with all the lights working (in the correct directions). I've fitted 1.5k resistors in series - I haven't got enough 2k - as a safeguard as you suggested.

 

Interestingly, I've since taken the body off the Class 45 and the wiring (and motors themselves) look somewhat different. The 0.33 ohm resistors on Falcon are 1 ohm resistors on the Class 45 and where the wire colours are similar at both ends on Falcon, the colours are different on the Class 45. However, armed with the knowledge gained from fitting the decoder to Falcon, I'm hoping the Class 45 will be both easier and quicker.

 

Thanks for your assistance.

[Nigelcliffe]

Glad its working. 

 

I suspect the 1.5Kohm resistors are unnecessary belt+braces, but without inspecting things I can't be sure.   In most models, the supplied lights are way too bright anyway, so no harm in being a bit closer to "scale brightness". 

 

I wonder if the 0.33ohm resistors are inductors for motor noise suppression ?   If that is the case, they will be in series with the motor, and replacing with wire links will slightly improve control as the decoder has a better chance to get a clean BEMF sample to work from.    

 

 

- Nigel

[Ray H]

Fortunately (?) the LOKPL96KS replaces the Heljan PCB so the resistors/inductors are no longer there.

[Ray H]

The next loco on the bench was a Class 45.

 

The internal arrangement with two separate motors - which differs from Falcon - appears to leave plenty of room for the LOKPL96KS, the decoder and for the speaker(s).

 

I began by working out which wire(s) did what and transferred them to the appropriate terminal on the LOKPL96KS one set at a time.

 

I assumed - wrongly as it turned out - that the wiring arrangement for the lights would be similar to Falcon so I spent a while tracing the pairs of wires so that I could do as I did in Falcon and swap the -ve and +ve wire groups over. It didn't help that one wire of each pair was blue.

 

With wires duly swapped around I put the loco on the test track. The motors worked, the speaker sprang into life at the appropriate times but did the light  work - No!

 

I spent a little while checking my sanity and the wiring but couldn't find anything amiss. I then looked at the marking on the Heljan PCB with an eye glass and discovered that the common terminal on the Class 45 is +ve whilst Falcon's original common terminal was -ve. I could simply have done a one for one transfer from the Class 45's Heljan PCB to the LOKPL96KS.

 

With the wires swapped back again all the lights worked as they should.

 

I even managed to read the resistor values on the Heljan light panels - each resistor is 4k7 ohms so I didn't bother to add any additional resistors.

 

Hope this helps.

[Phil S]

I'm puzzled reading through all of this to keep seeing references to the common terminal(s) of a DCC compatible PCB or decoder being apparently negative, whilst they are normally positive !!!

 

Although there are some decoders available which offer the ability to 'source current' from the switched output ( let us call it function x set high), 'the DCC norm' is for the switched function to pull the line low....and to 'sink current' into the switched function 'output'. .... I.e. when active it is switched low (0V) and draws current from the common positive supply.

With the smaller decoders using eg the 8 pin NEM plug/socket, the common wire is blue (constant +volts) and white, yellow green and violet wires are ' pulled to ground. ' to switch their function 'on'. Eg white pulled to 0V to switch the front lights on during forward travel.

It may help, when working with modern wire-free decoders, to think back in terms of those individually coloured wires.

 

Remember when measuring, that we are talking about a local positive and negative 'outputs' ( connections ) because the voltages on the decoder are beyond the diode bridge of the input stage... Which fullwave rectifies the track voltage to create the local supply used by the processor and driver stages. Avoid potential confusion by measuring to track voltage at any time (except perhaps with 6pin decoders where the lights use the rails as a return path.

Edited November 6, 2017 by Phil S

[Ray H]

Phil

 

The LOKPL96KS is effectively a replacement for the Heljan PCB. It is not a decoder. The decoder is an MX696KS that plugs into the LOKPL96KS which avoids any soldering. The speaker and the Stay Alive capacitor are also connected to the LOKPL96KS.

 

The original problem was that I hadn't realised on Falcon that the Heljan PCB had a common terminal for all the LEDs that was -ve and required the +ve connection to the LEDs to be grouped together for the LOKPL96KS with the +ve connection split into forward and reverse.

 

The Class 45 wiring was a straight one to one switch between the Heljan PCB and the LOKPL96KS - I had initially thought it was wired the same as Falcon, which it wasn't.

Edited November 6, 2017 by Ray H