Debugging motor decoder

[berry120]

So after waiting forever for a bunch of 2W resistors to arrive from China I've finally got around to building one of these - but alas, my electronics side has failed me and I'm not entirely sure what I've done wrong! I've checked everything over as thoroughly as I can and can't see anything immediately obvious (and it doesn't help that my multimeter appears screwed since it's giving me very high (~200V) voltage readings that are clearly wrong!)

Essentially, the thing seems to work fine on DC but not on DCC - when a DCC signal is pushed through it seems to steadily crawl along at a bit under half speed, no matter what packets are sent. I don't know if it's taking the average of the DCC packet and essentially rectifying it to produce a DC voltage? Just a guess.

Is there some test that I can do or a likely component that's gone wrong in this instance? Tried more than one PIC, so I don't think that's the issue. The l7805 and 2W resistor both get rather warm during operation, but I'm assuming that's normal? Thanks!

[berry120]

Ah, scrap the above - turned out a) I had a short that I'd missed and b) my multimeter's battery had run out! New battery, built a new board, double checked shorts and *sigh* have a different set of problems.

Firstly, the good news - the decoder seems to be recognising and processing packets - however, the voltages appear way off. At full speed one way, I get around 3V on the output, at full speed the other way, the same but inverted (-3V) - and correspondingly less voltage as I drop the speed. So it seems to me there's a problem in the power department, but I can't for the life of me find where. The 5V reg is kicking out bang on 5V, the supply voltage is 12.6V.

In terms of the l272:

Pin 1: 2.95V
Pin 2: 12.6V
Pin 3: 0.16V
Pin 4: 0V
Pin 5: 5V
Pin 6: 0.61V
Pin 7: 0.61V
Pin 8: 4V

The PIC:

Pin 1: 5V
Pin 2: 0V
Pin 3: 3.95V
Pin 4: 5V
Pin 5: 5V
Pin 6: 1.13V
Pin 7: 1.13V
Pin 8: 0V

Not sure if there's anything untoward in there that could be of any help, but I was always taught the first rule of troubleshooting was to check voltages! Any ideas would be much appreciated!

[berry120]

Oh and no joy just with a DC voltage - tried firing the 12.6V power supply straight at it and nothing on the output at all.

[suzie]

The voltages on pins 1 and 8 of the L272 and similarly the voltage on pin 3 of the PIC are the interesting ones. This should be the PWM signal that controls the motor speed so if you can get a scope on it you should see a variable mark space ratio as you adjust the speed.

You might need to adjust CV9 to match the motor before you get maximum performance out of the motor.

The 2W resistor will get quite warm so don't worry about it if your 5V is OK. If it gets hot and starts smoking though suspect that the tantalum capacitor has failed if that is getting hot too.

You might like to try temporarily installing the function decoder firmware to do static testing, that will allow you to control the voltage on the grey and orange wires using functions F0 (direction dependant) and F2 to verify that the full track voltage is getting to the outputs, and you should also be able to test on DC too.

[Paul Harman]

Just to add to what Suzie has said make sure that you are not using 126 speed step mode on the motor decoder. Only 14, 27 or 28 steps are supported.

[berry120]

Good news - it works! Thanks Suzie and Paul for your suggestions - turns out like a complete muppet I'd accidentally changed CV5 to a rather low value when programming the PIC, hence the low track voltage and the fact the loco wasn't going anywhere. When I bumped that back up to the default and adjusted CV9 (165 seems like the best value for an old X04 loco I was testing with if that helps anyone) it ran like a dream.

Thanks so much for making this circuit design available Paul - definitely building another few of these up soon! Along with a homebrew controller using (for now) an Arduino, CMDRArduino and the minibooster circuit, it makes for an incredibly cheap yet very effective DCC solution.