Crazy decoder values

[Potrail2378]

Hi, I have two decoders with the same fault. I really don't know how they got damaged. If I may, I will tell you the symptoms, and hopefully someone can help, so I don't damage any more. With the red and black wires connected to the DCC output, I get the following readings on a home made test jig -

Grey - Fwd = constant 26v Rev = constant 26v

White - Fwd = 21v Rev 26v

Blue = positive voltage

Green - F1 off = 0v F1 on = 26v

Yellow - Fwd = 0v Rev = 26v

Orange - Fwd = 26v Rev = 26v

Violet - F2 off = 0v F2 on = 26v

[96RAF]

What voltage you see depends upon which scale you are using on your meter and where you are referencing to/from. I would not expect to see 26v at all.

Red and Black are DCC track voltage and will show approx 15vAC across those two pins. Do not meter these two pins against any other pins, except each other.

Orange and Grey are motor voltage and you will see 0-12vDC in fwd and reverse as output by the decoder. This voltage is PWM, hence 12vDC of variable duration which will show on your meter as the average potential + or -. Again do not meter these two pins against any others.

Blue is a constant 12vDC generated by the decoder. This can then be used to meter the remaining wire colours.

White is switched negative going in hand with Blue when F0 is commanded in the loco forward direction.

Yellow is the same but when the loco is in reverse.

Green is also switched negative in hand with Blue when F1 is commanded. This is not direction dependant.

Purple when found is again switched negative in hand with Blue when F2 is commanded and also not direction dependant.

These function negative wires should only be metered against the blue positive, not each other or any of the track or motor pins.

I suspect your test rig may be crossing voltages and that is where the decoder damage is occurring, e.g. if you apply an external voltage to the motor pins whether the decoder is powered or not you will damage the H-bridge motor drivers. Similarly if you apply an external voltage to the function output wires then you also risk blowing the function output circuits.

Note: Do not confuse decoder function numbering with operating F-key numbering. Take for example a four function decoder - When controller function key F0 is selected it uses decoder function output 1 in forward and function output 2 in reverse. Key F1 uses decoder function output 3 in either direction. Key F2 uses decoder function output 4 in either direction.

[ColinB]

I must admit I have never measured the voltages on my decoders. Funny I have never really seen the point, especially as I used to be an electronic engineer, it is not as if I can fix them that easily. Generally in my experience the favourites to die are the input diodes on the input and the output circuit. If you have the right equipment and like soldering grains of sand, then you can fix the input diodes. DCC is a pulsed signal, so any measured voltages are going to be a bit dubious. I have a super expensive Fluke multimeter and even that gets it wrong. 96RAF has explained all the signals, the only thing I will add is the white, yellow and green/purple outputs are what we call "open collector" outputs which don't do anything until you "pull" them up to a voltage using a load resistor. Perhaps if you explained the fault you have, we might be able to help a lot more.

[96RAF]

The other thing that might help is a schematic of your test rig or at least a photo and description of it if a commercial unit.

The load resistor Colin is talking about on the function wires is of course the lights or whatever service you are switching from them.

I use another fairly expensive device to measure DCC voltage called a RRAMP meter which is simply a meter modified to measure the DCC waveform voltage more accurately. In the great scheme lf things a comparative voltage is all we are normally looking for, not a precise value.

RRAMP Meter showing DCC volts and amps.

[Potrail2378]

Hi, here is a photo of my 'test jig'. Obviously, I only tested the outputs because the grey and orange wires were causing problems.

I have just tested each and every lead for any short circuits, and there are none.

[96RAF]

I would ‘improve’ your rig by connecting the Red and Black wires to a known good DCC source by way of an isolation switch so you can safely plug-unplug a decoder.

Then hang a different colour led and resistor as an indicator on each of the White, Yellow, Green and Purple wires, connecting the Blue as common positive to all led anodes.

If you have a motor then hang that on the Orange and Grey wires, if not then you could use a bi-colour led and resistor to indicate increasing voltage and direction, or you could fit a small panel meter with a suitable DC range in lieu.

Then when you plug in a decoder and operate it from your controller you can simply check all the functionality, and address it to suit your needs.