APC SmartUPS 5000 blows fuses (no error messages)

[axero]

I have an older variant of the APC SmartUPS 5000 and the other day it shorted out and kept blowing the group fuse in the distribution board. The batteries are original and have seen better days but they can sustain the load for 5-10 minutes which is about 260 Watts. The group fuse is rated for 10A @ 230V and from the specs it is recommended to plug the UPS to a 16A fuse (as it is rated for max 3750W load):

When I went to the place to troubleshoot the UPS, I begun with replacing the fuse which blew instantly upon replacement. Everything was dead. I removed the load from the UPS and tried once again to replace the fuse. It instantly blew again but now the UPS was alive afterwards (from battery power I assume), so I pushed the self-test button and replaced the fuse once again. Now the fuse didn't blow and the UPS ran the self-tests without issues.

I tried to reconnect the load and now everything works as normal. Never during the malfunction were there any visible error messages but I don't have any powerchute software to check any deeper error messages. I know it has an extra network diagnostics board with IPMI like functionality (Network Management Card) but I don't know how to operate it.

Does anyone know what happened? Is this a malfunction or just a temporary state where it needed more than 16 amps of current? Perhaps the old batteries are causing the issues?

[johansen]

you may have a sticking relay. if one of the relays sticks, the ups should fault, but i'm fairly certain it should be able to detect that..

anyhow a stuck relay could short out one of the secondary coils which are used for voltage buck or boost. this will hopefully blow the fuse before the relay vaporizes.

[axero]

Ok, some updates here. When I connected the load that priorly was in the UPS directly to the power outlet; I noticed that the load now blows the group fuse. There seems to be a short circuit in a part of that load. I have managed to disconnect that part of the load and the rest of the system/load runs without tripping the fuse.

The problem is however, that when conducting measurements on that troublesome part of the load with a multimeter, I detect no short circuit. It seems that this problem is intermittent.

After that discovery, I decided to reconnect the UPS but it still blows the fuse as before. I managed to make it agree with the group fuse in a similar fashion as described in the prior post; I removed the load and powered the UPS up. Then the line voltage LED were blinking along with the battery LED, but after waiting for a while it went to normal status. I tried self-test once again but that lit up the red "battery X" indicator. Perhaps the batteries have given up.

It's a bit strange though because the troublesome part of the load is protected behind a 2 ampere fuse and that fuse got tripped, well at least eventually, not from the beginning. I'm a bit confused about this.

Perhaps if I can get to the network management card, more detailed messages in the event log could reveal what might have happened. Unfortunately, I don't know what IP, the card is bound to or the Admin password. Probably, a factory reset would wipe the event log.

[johansen]

you may be able to pry the cover off the relay and examine the contacts.
it would normally be considered impossible for a 2 amp fuse to blow a 16 amp fuse..

a 2 amp circuit breaker on the other hand might do it.

[rs20]

The problem is however, that when conducting measurements on that troublesome part of the load with a multimeter, I detect no short circuit. It seems that this problem is intermittent.

Not necessarily intermittent; it's just that testing with mains is testing with > 100V AC, whereas testing with a multimeter is < 5 V DC. Not exactly mysterious that you might get different behaviours. A 50V MOV or GDT would exhibit the same behaviour as you're seeing: appears perfect open to multimeter, dead short once mains in applied.

More realistically, it could be a switching converter that has a working UVLO, but then has a fault that causes it to just becomes a short circuit once sufficient voltage is applied.

It's a bit strange though because the troublesome part of the load is protected behind a 2 ampere fuse and that fuse got tripped, well at least eventually, not from the beginning. I'm a bit confused about this.

Again, I wouldn't lose sleep over this -- a 2 amp fuse blowing a 16 amp fuse is a little odd, but we are talking about vaporising pieces of wire here -- hardly a sound basis for logical deductions.

Your load is blowing a 2 amp fuse. That has nothing to do with the UPS, and is a clear fault, so fix that first. Only once you've figured that out, should you move on and think about your UPS again.

[axero]

Not sure how to interpret "vaporising pieces of wire". The group fuse is a porcelain diazed D type or at least used to be. I have now replaced it with an automatic fuse into the Diazed socket. I guess that automatic fuses are a bi-metal construction that breaks the circuit as it "overheats" due to the too high amperage. I take it to understand that the above mentioned abbreviations mean; MOV = Metal Oxide Varistor, GDT = Gas Discharge Tube, UVLO = Undervoltage Lockout.



The shorted part of the system is a 230V outdoor surveillance camera together with a couple of NIR lights. I have put all high-voltage cams behind what I believe are 2 ampere fuses. They look similar to this:

but with a pink flap instead of black. I think the pink color is color code for 2A (like 4A = brown, 6A = green, 10A = red, 16A = gray, 20A=blue, 25A=yellow, 35A=black) as is the same system as for Diazed fuses (https://en.wikipedia.org/wiki/IEC_60269). Perhaps they are circuit breakers, I thought that they were ordinary fuses. All I've seen is that there are slow fuses and fast fuses, never thought of the existence of such circuit breakers as DIN-modules.

What I mean with "intermittent" is that during testing I got it to work at the beginning when I started the troubleshooting. I kept it running for a while normally with the UPS on. Then when I disconnected the UPS and connected it directly to the mains power, the 2A fuse/breaker tripped and the group fuse (which is also slow) was blown to boot. As long as I didn't reset that 2A fuse/breaker, the group fuse didn't blow.

So with that knowledge, I reconnected the working parts of the system back to the UPS but the UPS still blows the group fuse.


The UPS should be "smart" enough to have an overload protection circuit that would kick in should a short circuit occur on the load side of the UPS. So obviously something is wrong with the UPS. I'll try to take the UPS apart and see if I can find something. It's just that it is so ... damn ... HEAVY, even without the batteries.

What I'm losing sleep over is the UPS, the short circuit at least seems to be a more manageable problem.