UD2.9 Troubleshooting Help

[GKnapp]

Hello all,

It's me again, this time almost with a working driver board, one of Profdc9's ud2.9's. I'm hoping some of you more patient types might be willing to help me troubleshoot! Any help would be greatly appreciated.

No plumes of smoke when I give it 24 volts in, and 24, 12, and 5v outputs are good.

The most immediate problem I notice is that the OCD/UVLO light comes on and stays on. Sometimes, it is intermittently on or off. The OCD tap point voltage reads less than a volt, depending on what i have the pot set at. I'd like it to be about 5 volts. Disconnecting the UVLO part of the board doesn't change the voltage reading, but does sometimes turn off the OCD/UVLO light off.

When I have the inversion of the fiber optic receiver the wrong way round, the light is intermittently on, and the driver runs continuously (as I would expect from it being the wrong way around.) When it does this, one of the output transistors gets hot much faster than the rest.

Now, with the receiver not inverted, nothing happens. Before, it was generating nice square waves (with the uvlo part disconnected) but now enough poking and prodding on my part have made it stop working at all. I'm doing this all without the pulse skipping features enabled (treating it like a ud2.7)

So, I suspect something is up with the OCD/UVLO circuit, but I'm not sure what. The voltage at the ocd tap point is clearly wrong, and I've tested the pot and found that it was working fine. To the best of my knowledge, the diodes on the ocd aren't fried. Also, I would suspect that the problem was solely with the UVLO chip being fried, but even without the uvlo jumper connected I still am having problems. None of the relevant chips are obviously fried (unnatural dead shorts and such) but I haven't yet replaced them all.

In any case, do you have suggestions of what sequence of things I should test, or any ideas of obvious things I'm not thinking of? Any suggestions and help is very much appreciated!

For reference, here's the schematic:

[davekni]

For low OCL trip voltage, one of the following:
    Comparitor input (U6-2) is damaged, drawing current.
    R13 or RV1 is the wrong value.
    C16 is damaged (resistive short) - can happen when ceramic caps crack.

For the opto receiver. send it a square wave and scope from the receiver output through the circuit.  See where the signal stops.

[GKnapp]

Thank you Dave, that's very helpful! I'll check all those and report back at some point Thank you for all the help you've given me and others, it's very appreciated.

[GKnapp]

I tested all those, and put in a new comparator, but it doesn't seem to be any different. The resistors seem to be the correct values, and I replaced both c15 and c16. One thing I noticed is that the voltage at the OCD tap point is variable--it sometimes starts higher then goes down, or fluctuates around. Also, sometimes the voltage is different between J9 1-2 and 2-3. Then if I wait a moment, it isn't different. Is this indicative that the problem is probably some chip flailing around? Sometimes, connecting the UVLO portion via J7 seems to make a difference in the voltage I measure, and then other times it doesn't. It's not at all predictable. Adjusting RV1 makes a very very tiny difference in what voltage I measure. Ugh.

When I send an interrupter signal in, it used to make it all the way to the gdt outputs when the ocd light was flickering on and off. So I assume that at least at some point, the rest of the circuit was functioning. Currently though, the light has been consistently on and it only makes it to the input of U5-B where it disappears.

[davekni]

It sounds like RV1 has a bad internal wiper contact - probably the most common failure of POTs.  It may measure 10k end-to-end, but resistance from either end to the wiper may be high and random.  You could replace RV1 with two 5k resistors for a fixed 5V OCD trip voltage, or replace with a good POT.  (Measure center to end resistance while adjusting  on any replacement part before inserting.)

[GKnapp]

I tried a new pot, and that almost looked like it worked for a second (got a temporary reading of 4v at the tap point), but then I went back to measuring tiny voltages. So then I tried two 5k resistors instead, and that also still leads to a small voltage on the tap point. I noticed that the output of the 24v regulator was 26 volts which was alarming, so I replaced that, with no change, except that in a moment or two something died and the board shut off and started drawing lots of power. So I think maybe I'm going to call it a day :p

[davekni]

Sounds frustrating:(  Unless there is a short on the ECB, there are only those few components in the VR1 circuit.  Either one (or more) of them is bad or there is a short to some other node of the circuit.  (I hadn't listed J9-2.  I presume you aren't adding any jumpers to J9, only measuring the voltage on pin 2.)