Global AC logic output - XP Power FleXPower

[tony359]

Hi all

I've fixed a couple of FleXPower X7 power supplies. Since they have some logic outputs, I went to test them to make sure all was in order.
"Fan Fail" would work when the fan is manually stopped.
"Global DC" would change state when I manually adjusted one rail beyond 10%

"Global AC" doesn't seem to do anything. To test, I power up the PSU and then I remove power. The 5VSB and Aux rails remain active for quite a few seconds after power is removed. I see the rails collapsing but there is no change in the "Global AC" signal. The manual says that the signal should change 5ms before there is "loss of regulation". I have two of these PSUs and they both show the same. I've tested as the manual says, I read 5V until the 5V line dies. I still read 5V while watching the 24V DC output collapsing.

Chances are they have the same fault - but could it be that I am misunderstanding how this signal is working?

Thanks!

[CapLeaker]

You may have to load the output with something, What is it? Sounds like a medical or a special PSU of some sort.

[tony359]

It's a modular power supply, it can be configured from Digikey etc to whatever you want in terms of currents and voltages.

https://www.xppower.com/product/fleXPower-Series

I don't think the load is the issue here - but I can try, thanks. The idea with the AC Global signal is that if the PSU has a compromised primary, the standby electronics will sense it and won't even try to turn on the secondaries - and will also log/display the issue.

"Global DC" logic works. There is an LM339 where the logic happens which I understand is a voltage comparator. Something might be happening there I guess. I have traced the "Global AC" logic to a pin which is an output of two inputs. I'll check
- Whether the optocoupler is actually getting a signal (maybe the LED is broken? On both PSUs though?)
- Whether any signal is coming out of the LM339
- What kind of signals are coming in the LM339

So far I haven't identified where the signal to monitor is coming from.

[CapLeaker]

Problem is that the logic is programmable or settable. You can set it high or low output. Sounds there has to be something else on the input that makes the output go the other way.

the global AC fail does do nothing to the psu secondary. It should go to a main board that this power supply powers and tell the computer for example stop reading and writing and lift up the heads on a HDD. So there should be an output on the connector. Kinda same thing than the Power OK pin on an ATX power supply.

[tony359]

there are some jumpers at the back, you see them closer to the optocouplers. I suspect they set how the logic works. I don't see anything programmable on that PCB.
Yes, Global AC Fail doesn't do anything to the secondary, it monitors (I guess) the 390V DC and sends a "Global AC Fail" signal when it goes below a certain level to instruct the device connected to the PSU to take action.

OR - it could be one of those DC modules sensing the DC voltage in and reporting it back to the logic.

So far I haven't been able to see where this logic is getting the signal from but all the traces arriving to that area seem to come from the DC modules - so maybe one of them is responsible for sensing the DC voltage in and send a signal back to the logic.

After all the manual says that "Global AC Fail" will be triggered 5ms before the loss of regulation and that level of precision makes me think that maybe it's one of the DC modules (module 1?) responsible for determining that. Alternatively the logic can only probe the 390V DC and take a guess on when the DC modules are going to give up.

[CapLeaker]

Find the optocoupler that is responsible for the ac power fail and go from there. Should be easy to find. It supposed to give 5ms warning of output regulation.

[tony359]

Indeed - the opto is the one with the green mark and I traced the signal to one output pin of the LM339. I know what the inputs are, I need to check what happens there.

Because it says 5ms before "loss of regulation". I'd be tempted to think that this signal comes from the secondary and not primary. The main switching IC might have a feedback signal that goes back to the main logic and tell "I am about to drop the voltage".

"simple" is a big word 🙂 Thanks for now!

Edit by mod: Tony made a video about this fault now, seen here with timestamp to the relevant section: https://youtu.be/jDAyzrxcYdU?t=1609