APC Smart UPS SUA1500 does not detect AC / doesn't turn on

[CapLeaker]

Did you remove both of the transformer from the board and did you notice anything unusual about either transformer winding connections to the pins on rare occasions I have seen bad soldering joints on the connections pins after removing the solder and reapplying solder to the joints corrected the issue I am not saying it the issue you have

No, I didn't remove these transformers from the board, I did look at the wire connection to the pins and them are o.k. Also the sizzling, buzzing noise is coming from the inside of T1. We had a very large event going on last week at work, with all the extra noise created for prepping this place over the last 2 months, I couldn't hear the faint fizzling noise inside the transformer. I suppose I could “hack” another transformer in. I could do that no problem if it would be my own. Since this thing is for my workplace, I can't do that.

I am not sure if they let me have that broken UPS or not, the manager went on vacation right after the event.

[sam_sam_sam]

Did you remove both of the transformer from the board and did you notice anything unusual about either transformer winding connections to the pins on rare occasions I have seen bad soldering joints on the connections pins after removing the solder and reapplying solder to the joints corrected the issue I am not saying it the issue you have

[CapLeaker]

OK. not sure what went on here (it probably didn't help working on this only on certain days when I had time), but I verified T1 against T2 transformer again. T1 is making this sizzling sound and the I measure 1.7kOhm across the primary winding. It's game over with this, as I don't have a replacement transformer.

[CapLeaker]

AC-OK signal is asserted high, looks OK to me. The LED's coming on might be part of its power up/self-check sequence, an LED test then it sees something confusing with its inputs and shuts down.
If you hear something sizzling, I would go after it, it might be a poor connection on the transformer bobbin (crappy soldering) or just gently tap a plastic pen on things to see if you can flush it out. Even a relay could be fizzling inside.

I am not sure how I missed that fizzling noise. Keep in mind this thing is on my bench at work, where it isn't a quiet place. This “fizzling” noise appears now as soon as I plug it in with the batteries connected and the relays kick it. No battery, no relays, no fizzling.
I did think the AC-OK is good. I remember early on checking these little transformers and compare them to each other measuring resistance and they were identical. Anyway… I'll see Friday if I can figure out where this noise comes from.

[redwire]

AC-OK signal is asserted high, looks OK to me. The LED's coming on might be part of its power up/self-check sequence, an LED test then it sees something confusing with its inputs and shuts down.
If you hear something sizzling, I would go after it, it might be a poor connection on the transformer bobbin (crappy soldering) or just gently tap a plastic pen on things to see if you can flush it out. Even a relay could be fizzling inside.

[CapLeaker]

The output relay is normally always on (R5/Q2) and the MCU gives the SHUTDOWN command (Q1) to turn it off.
So why is the MCU not enabling the mains output? What does AC-OK signal look like?

I replaced D13 with two regular 1N4148 signal diodes. I did do a couple of measurements see attachment. I did notice something today while I tested in with AC. One time when I plugged it in I saw for a split second all LED's light up. It still behaves the same way. Short press on button and I get a beep, but nothing happens. Here is something new and I don't know if it was there before and just didn't hear it. I hear little continuous arcing somewhere where the 2 little transformers are on the main board.

[redwire]

The output relay is normally always on (R5/Q2) and the MCU gives the SHUTDOWN command (Q1) to turn it off.
So why is the MCU not enabling the mains output? What does AC-OK signal look like?

[CapLeaker]

You mention hearing a relay click for a second. It's doing a self-check. Relays off - no mains seen, relays on - mains seen etc.
The transfer relays seem weird - it's using two in parallel RY5, RY4 for Line, RY7, RY6 on Neutral. One must make, the other break or something, but the contacts are in series as well for safety.
I wonder how sophisticated the self-test is. Again, with mains applied and some kind of load, I would see why it's not making it through. Trace the path.

The blown diode might be from insulation failure or arc from contacts to coil on that relay. There are MOV's on either side but not in the middle, so a spike from the VR or UPS output maybe caused this. Just a guess.

I did. The problem is the output relay isn't getting the command on the coil for it to be energized. All relays besides the output relay need 24V from the battery, otherwise the UPS doesn't work. Check. ✅
As soon as I plug it in or turn the inverter on, I have 24V on the output relay coil. I also have AC waiting for it to be switched via the relay. The path of AC power through the relays isn't the problem.
I am not sure why that diode D13 went totally open, maybe I damaged it myself when I tried to fire the relay up with 15V instead of 24V.
The main IC isn't dead. After a brain dead, plugging in the battery and AC power, it beeps every time I press the ON button. It is some sort of a permissive problem or the MCU doesn't see any AC mains and thinks it is unplugged all the time.
As a temporary measure I swap the diode out wit something more common over 100V rated and just solder it directly over the relay coil on the back of the board. A 1N4002 or better should be good just fine.

[redwire]

You mention hearing a relay click for a second. It's doing a self-check. Relays off - no mains seen, relays on - mains seen etc.
The transfer relays seem weird - it's using two in parallel RY5, RY4 for Line, RY7, RY6 on Neutral. One must make, the other break or something, but the contacts are in series as well for safety.
I wonder how sophisticated the self-test is. Again, with mains applied and some kind of load, I would see why it's not making it through. Trace the path.

The blown diode might be from insulation failure or arc from contacts to coil on that relay. There are MOV's on either side but not in the middle, so a spike from the VR or UPS output maybe caused this. Just a guess.

[CapLeaker]

scratch my last post. It is the nature of that beast. But my original blown diode D13 went again. but this time fully open.

[CapLeaker]

Hmm... not sure if I found something or not. The site fault wiring circuit works, but at the same time IC20 the opto gives weird readings if I poke around on it. Not sure if it is due to this specific optocoupler being used, but it doesn't react like a normal one.

SFH6156-4

[CapLeaker]

Or the sensing circuit has some kind of issue or the MCU has a issue that could be causing the issue but how to troubleshoot the issue is the real question I not even sure where to begin if I were doing this repair this is the reason I paying close attention to this post to see if he can figure out what wrong with it

The way on how to go about this is already explained. Basically I have to get it into a working powered state, then figure out all the permissive crap being good and going to the MCU. Now if everything is good to the MCU and it still doesn't work I am shit out of luck, unless I replace the MCU. And exactly this is what I am afraid of a bad MCU, as I already went over the permissive circuits and detection circuits all the important diodes, transistors, Fet's, and some resistors too. So far besides that SOT-23 diode I found squat.

[sam_sam_sam]

Or the sensing circuit has some kind of issue or the MCU has a issue that could be causing the issue but how to troubleshoot the issue is the real question I not even sure where to begin if I were doing this repair this is the reason I paying close attention to this post to see if he can figure out what wrong with it

I have always one thing going through my head: Why is it that there is a small beep when I short press the power button and absolutely no LED is working on the panel and the 120v sockets on the are disabled. Why is there nothing on that front display. I mean if it has a fault, it should display it. At the other end, the 12v should be

( I do not understand this part to why it would display the status when you do this part and not the other function )


OK because I can turn the inverter on and it works fine as kind, with display and all. Why no LEDs when on AC?

This is reason I have to wonder if the MCU is dead and this is reason why you have no response on the LED light status panel but like I said earlier I am not sure how you determine if the MCU is dead or the sensing circuit is dead in a section not letting it power up or something short out and destroy it

One question I have dose it charge up the battery because if it does then this really does not make sense at all unless it is a sensing issue or something wrong with the MCU programming issue or it corrupted some how

This might not be the same thing but we had a machine at work that it would boot up but it would not respond to switch commands or touch screen functions and come to find out from technical support they said that the Erom chip was corrupted somehow and once it was replaced and the parameters were reinstalled it was functioning again so I have to wonder if it could be something similar

if it does not charge up the battery then I would lean more to charging circuit issues

Dose this battery backup have the cue function on where you can see the status of the device and adjust settings and what not does this function correctly this might be one way to know if the MCU is functioning at all

Just keep in mind that I just thinking out loud and this is how I play out in my mind how and what to check sometimes I am on the right track and sometimes I just going down a rabbit hole and going nowhere and probably going nuts as well

I hope this helps you

[redwire]

With mains present, I think the MCU is seeing something the firmware can't understand, undefined, or a safety-circuit has activated. Or it is a legal department reaction...

UPS auto-destruct is when mains is present yet the inverter starts up. That is very bad, if it were to do that backfeed into the inverter would blow it up. It would burn grandma's house down. I don't see fuses for that situation.

So to protect against that happening, I see it uses two transfer relays- one on Line RY5, other on Neutral RY7. It seems to be a redundant scheme so if one relay had welded contacts, it would not blow up.
Also the signals for the transfer relay coils has an extra safety circuit. XFR-GATE is an analog permissive that can override the MCU IC11.
It might be not permitting the line transfer relay (Line) to engage, the MCU sees this and shuts off. Or the H-FLT, ALL-EN2 are asserted?
"SMKLS" could mean smoke less lol.

You'll have to poke around and see which permissive saying no to the party.

[CapLeaker]

I usually follow signal verses testing parts when I'm kinda lost what is not working.

The MCU gets these signals (that I can see):
SITE-FLT is ground-fault, but I don't think it matters the UPS works anyway?

Analogs:
AC-IN is mains from voltage sense transformer T1.
AC-OUT is transfer bus from from voltage sense transformer T2.
I-OUT is load current from CT2.
I-CAP is PFC capacitor current from CT1.

I would put a small load like 25W lamp on it and plug it into mains and trace the AC path with a multimeter. It might be an open relay and they can carbonize and test good... but they aren't. Or the autotransformer is open for BOOST/TRIM etc.
Is the 12V rail ok, IC2 will stop it from transfer relay from activating I think. This thing has so many permissive signals. Some relays have a permissive on each coil wire, so like an AND gate both have to approve before the coil pulls in.

I am not concerned about the site fault, even at that mine works correctly as I had tested it at some point before,
the transformer should be good, as it isn't open. I had another look at this schematic and it always takes me a while to wrap my head around it. Since I can't work on it every day at work, I might put it back together and bring it home next week.

I do understand what you're saying with the 12v rail and all the permissive jazz. That's why it is such a pain to work on after a few days, as I have to go through the schematic again. For sure I am at the end with it in this state and have to get it back together in order to power it up. The cables are too short to power it up while being apart. I can't connect the transformer like that.

I have always one thing going through my head: Why is it that there is a small beep when I short press the power button and absolutely no LED is working on the panel and the 120v sockets on the are disabled. Why is there nothing on that front display. I mean if it has a fault, it should display it. At the other end, the 12v should be OK because I can turn the inverter on and it works fine as kind, with display and all. Why no LEDs when on AC?

[redwire]

I usually follow signal verses testing parts when I'm kinda lost what is not working.

The MCU gets these signals (that I can see):
SITE-FLT is ground-fault, but I don't think it matters the UPS works anyway?

Analogs:
AC-IN is mains from voltage sense transformer T1.
AC-OUT is transfer bus from from voltage sense transformer T2.
I-OUT is load current from CT2.
I-CAP is PFC capacitor current from CT1.

I would put a small load like 25W lamp on it and plug it into mains and trace the AC path with a multimeter. It might be an open relay and they can carbonize and test good... but they aren't. Or the autotransformer is open for BOOST/TRIM etc.
Is the 12V rail ok, IC2 will stop it from transfer relay from activating I think. This thing has so many permissive signals. Some relays have a permissive on each coil wire, so like an AND gate both have to approve before the coil pulls in.

[CapLeaker]

Went on this again today and chased things around checking diodes, transistors, fets and the transformer. I couldn't complete the AC-OK cirquit as there are a couple of IC involved, so I have to put this thing back together hook it up and check it from the backside.

[CapLeaker]

redwire
I checked components R72, Q10 and D14; D18 and D28 are O.k. That's all I got for today. I miss my microscope at home.

[sam_sam_sam]

Verify your wiring with a circuit tester. If it sez it's good and the APC sez you got a wiring fault, there is something wrong with that detection circuit inside the UPS. The site wiring fault LED on this APC i am repairing (or trying to) is not on. Anyway that's how I would go about it.

Yes it is a APC brand I not sure exactly what the model number is at the moment

[redwire]

Using the schematic, SITE-FAULT activates if over 20VAC is seen between PE GND (chassis ground) and Neutral.
The LED is powered separately, from mains-in instead of the SMPS/battery etc.- so it can light even if the UPS is totally dead/blown.