APC Smart-UPS 1500 (SMT1500)

[Bretterson]

Hi there. Occasional lurker, first time poster.

To start, I do have experience with electronics and repairs, but I wouldn't consider myself advanced. I know stuff, but some of what I come across here is totally lost on me. My day job is IT (mostly help desk) and my side hustle is repairing pinball machines.

Ok, so the reason I'm here... I recently purchased an APC Smart-UPS 1500, model SMT1500, on Facebook Marketplace. The seller said it worked fine but would probably need new batteries "soon," so I couldn't fully test it. Well, "soon" really might right away because they're dead, so I bought a set of aftermarket replacements. I installed them and, aside from appearing to charge curiously fast, everything seemed fine.

Alas, everything was not fine. Once the battery reached 100%, the unit threw an "output over volt" error and shut down. This happened no matter what I did or what settings I changed. Strangely, during one test I managed to measure approximately 123 or 125 volts (I can't recall which) in the few second window before it shut down... certainly didn't seem over voltage to me. I also found that it works perfectly fine with the batteries disconnected, but together the batteries measure around 27 volts.

Obviously the guy still says it worked fine for him and won't give me a refund... so I figure I have little to lose and take it apart. I found a section of very charred components on the mainboard. One piece was so fried it just fell out and I have zero clue what it was.

Questions...

• Does this seem repairable or does the board look too far gone?
• If feasible, what components should I replace?
• Alternatively, does anyone have a clue whether I might be able to swap the entire board (640-3081A_REV06) for one from the rackmount version of this UPS (640-3081_REV03)? They look very similar aside from the battery connector.

I think that's all I have at the moment. Hopefully this thing isn't a lost cause.

[fmashockie]

I can try to answer some of your questions.  First, I would start by cleaning that board up.  It probably looks a lot worse than it actually is.  So clean it up the best you can.  Most of that soot will just wash off and then you will get a better idea of what might be damaged.

It looks like a MOV is likely the damaged/charred component.  If it is a MOV, and it took the brunt of the damage, it is very likely that the rest of the unit could be okay.  That is if this blew due to an external overvoltage event.  If it was caused by something internally, then it is only a secondary symptom of a primary fault.  You'll have to do some investigating to find that out.  But at first glance, this appears to be repairable.  Do not let all the soot/charring scare you. 

The battery voltage you measured seems fairly normal for these large 12V batteries in series.

Do some cleaning first, and then get back to the forum with some pictures once it is all cleaned up.

[Bretterson]

You're right, I definitely should've cleaned it up before posting. I agree that it likely appears worse than it is since I was able to wipe soot off that big yellow capacitor.

I thought that very charred/melty bit might be a varistor. I'm pretty sure there's also a missing component that fell out right below that and to the right of the yellow cap. I think it might be another varistor based on pictures of a board that sold on eBay a few months ago: eBay auction: #204631885269.

I didn't mean to suggest the batteries measured high, if it sounded that way. Reading it again I can see how it might be a little unclear.

I'll clean it up this week, maybe tonight, and post a new pic.

[Bretterson]

I had a chance to clean it tonight and the board looks pretty decent. That big yellow cap was ruined so I just cut it out to make it easier to get in there. Looks like a few wires should probably get replaced as well.

[RadioNZ]

Your first photos look catastrophic, but most of that is just soot.
The clean-up photos reveal it's not as bad as first thought.
From what I can see, one of the MOV (MV4 - the charred one inside the heatshrink) has taken a hit.
This could be due to a lightning strike or there's been one hell of a surge on the mains.
I would also check the other MOV to see if any of them have gone short as well.
As to whether you can swap out the board from a rack mount unit to this one, well the part numbers appear the same (apart from the revision) so it's possible.
You would be better off leaving the rack mount unit intact and just trying to repair this one.
Given that the unit is throwing up an over-volt indication, it's likely whatever took out that MOV is responsible for this indication and there
may be more failed components than what is initially visible. It may have damaged the line voltage sensing circuitry.
You mentioned that the batteries measured 27 volts together, that's about right - assuming your UPS runs the batteries in series for 24 volt operation. Many do.

[Bretterson]

Your first photos look catastrophic, but most of that is just soot.
The clean-up photos reveal it's not as bad as first thought.
From what I can see, one of the MOV (MV4 - the charred one inside the heatshrink) has taken a hit.
This could be due to a lightning strike or there's been one hell of a surge on the mains.
I would also check the other MOV to see if any of them have gone short as well.
As to whether you can swap out the board from a rack mount unit to this one, well the part numbers appear the same (apart from the revision) so it's possible.
You would be better off leaving the rack mount unit intact and just trying to repair this one.
Given that the unit is throwing up an over-volt indication, it's likely whatever took out that MOV is responsible for this indication and there
may be more failed components than what is initially visible. It may have damaged the line voltage sensing circuitry.
You mentioned that the batteries measured 27 volts together, that's about right - assuming your UPS runs the batteries in series for 24 volt operation. Many do.

Yeah MV4 definitely looks rough. This thing is also 11 years old, so maybe it's built up failure over the years. Based on the pictures of the board in the eBay auction I linked, there should be another one at MV2. It's probably the piece that fell out... could that be the responsible component? Could the leaky capacitor be to blame?

The board from the rackmount version lacks the "A" at the end of the part number that this one has, that's what I'm not sure about. If it was just the revision I would feel more confident giving it a shot.
I don't actually have the rackmount unit. I asked because there are a handful of eBay listings for that board.

I don't really know enough to know what other components should be tested. It works with the battery disconnected, I can power a device, so I feel like that means the line voltage sensing stuff is ok... but I could be wrong.

Yep, the batteries are wired in series.

[fmashockie]

If you are lucky, that MOV took the brunt of the damage and repairing that board should be relatively easy.

Some advice when working on UPSs: do NOT use the battery to troubleshoot/perform testing.  Those batteries can release uncontrolled amounts of current if there is a short somewhere on the board.  It can be very dangerous.  Use a DC power supply with current limiting capabilities to simulate the battery.

[Bretterson]

I pulled that component off the board to see if the part number would still be intact. I removed the heatshrink and the MOV is actually in good condition; the missing MOV must've just burnt/melted the the heatshrink of its neighbor. I'll have to replace it anyway seeing as the leads are pretty much nonexistent, but oh well. (Part number is ERZ-V14D471)

Do you have any idea what part MV2 might be? I can't see any useful details in any of the pictures I've come across.

Good to know about the battery. Unfortunately I don't have a power supply like that. I might still have one my dad built from a kit, but I'm pretty sure it only allows you to control voltage.

[fmashockie]

MV2 is a MOV (metal oxide varistor).  There are other MOVs on the PCB that you can check to see what kind it is. It is very likely the same as ERZ-V14D471

I pulled that component off the board to see if the part number would still be intact. I removed the heatshrink and the MOV is actually in good condition; the missing MOV must've just burnt/melted the the heatshrink of its neighbor. I'll have to replace it anyway seeing as the leads are pretty much nonexistent, but oh well. (Part number is ERZ-V14D471)

Do you have any idea what part MV2 might be? I can't see any useful details in any of the pictures I've come across.

Good to know about the battery. Unfortunately I don't have a power supply like that. I might still have one my dad built from a kit, but I'm pretty sure it only allows you to control voltage.

[fenugrec]

Good to know about the battery.

You may be able to cobble something with an incandescent bulb (say 100W) in series with the battery as a current "limiter". May prevent the UPS from working properly but would reduce risk of a catastrophic short.

[Bretterson]

MV2 is a MOV (metal oxide varistor).  There are other MOVs on the PCB that you can check to see what kind it is. It is very likely the same as ERZ-V14D471

Right, I know it's a varistor. I'm fairly certain it isn't the same one because the outline for that space is larger and the component looks larger in this picture from one that sold on eBay: https://i.ebayimg.com/images/g/sgEAAOSw~CZltAdV/s-l1600.jpg

You may be able to cobble something with an incandescent bulb (say 100W) in series with the battery as a current "limiter". May prevent the UPS from working properly but would reduce risk of a catastrophic short.

Oh cool, I'll definitely look into that. Thanks.

[fmashockie]

MV2 is a MOV (metal oxide varistor).  There are other MOVs on the PCB that you can check to see what kind it is. It is very likely the same as ERZ-V14D471

Right, I know it's a varistor. I'm fairly certain it isn't the same one because the outline for that space is larger and the component looks larger in this picture from one that sold on eBay: https://i.ebayimg.com/images/g/sgEAAOSw~CZltAdV/s-l1600.jpg

My apologies I made a mistake.  MV4 is the same as those with the heatstrink on them. 

MOVs have a few main parameters:

Voltage rating (AC/DC)
Clamping Voltage
Peak surge current

The MOVs are to protect from surges when the UPS is being powered normally and for protection to the inverter when the UPS battery is being used.  In some models there will be MOVs on the both the neutral and live rails of the circuit.  When a surge occurs that exceeds the clamping voltage of the MOV, it will go low impedence and short the current to ground (that is why you will notice on your PCB one leg of the MOV is on the Live or Neutral rail and the other is on Ground). 

This MOV MV2 not being there isn't going to prevent your UPS from working.  As I said earlier, these things are very dangerous to work on.  You might want to invest in some additional equipment (DC benchtop supply, scope, etc) as well as gain a bit more experience in troubleshooting before proceeding.

[asis]

Hi,

All MOV's parameters:
https://datasheetspdf.com/pdf-down/E/R/Z/ERZV20D471_Panasonic.pdf
It should be taken into account that the UPS design can be either 220V or 120V; accordingly, the ratings of certain components may be different.
It is possible that if at the time of the incident the UPS was in operation and sticking (burning) of the RY1 relay contacts could occur; RY2; RY3; RY4; RY6; RY900.
It would be appropriate to check the condition of their contact groups, or better yet, dismantle them and check them individually.
You should also separately check the status of the MOSFET's of the inverter and the operation of the charger.
There are several threads on the forum that describe how to do this.
-
The rating of capacitor C53 for UPS in AC220V version is 2.2mF x 275V~.

[flaotte]

Some advice when working on UPSs: do NOT use the battery to troubleshoot/perform testing.  Those batteries can release uncontrolled amounts of current if there is a short somewhere on the board.  It can be very dangerous.  Use a DC power supply with current limiting capabilities to simulate the battery.

is there any safe (safer) way to debug mains part of such device?

[fmashockie]

Some advice when working on UPSs: do NOT use the battery to troubleshoot/perform testing.  Those batteries can release uncontrolled amounts of current if there is a short somewhere on the board.  It can be very dangerous.  Use a DC power supply with current limiting capabilities to simulate the battery.

is there any safe (safer) way to debug mains part of such device?

The precautions one should take for checkin the mains portion or SMPS of the UPS are the same precautions one should take when working on any device powered by mains voltage. 

-Be careful when checking HV side or primary side
-It might be benefical to use a lamp series tester to protect from unknown shorts on the primary side
-If you plan to use a scope, make sure the unit is plugged into an isolation transformer.  If you plan to probe mains, you should only use methods that are safe - such as a HV differential active probe.

Those are just a few precautions one should take.  But the main thing that separates the UPS from other devices powered by mains is the UPS's battery which is usually two large 12VDC batteries in series that can deliver large amounts of current.  Keep in mind most of these devices are rated to deliver upwards of 1000W - that's a lot of power.  And a large battery that can deliver a lot of current is required.  If there is a short on the board in the charging or inverter circuit, this power can be delivered uncontrollably which could cause serious damage to life/property.  I've seen APC UPS units with shorts on the inverter MOSFETs that caused catastrophic damage and the battery protection fuses did not prevent this damage (I've done a few videos/threads on these topics).  So that is why I voiced my concern for troubleshooting these units without a DC power supply with current limiting.

[Bretterson]

This MOV MV2 not being there isn't going to prevent your UPS from working.  As I said earlier, these things are very dangerous to work on.  You might want to invest in some additional equipment (DC benchtop supply, scope, etc) as well as gain a bit more experience in troubleshooting before proceeding.

I'd consider getting a DC power supply. It looks like they can be had for around $65, I'm just not sure whether I'd use it much in the future. I think an oscilloscope is more than I want to get into as I have zero familiarity with using them. It's probably a useful thing to have and know how to operate, but the learning curve seems high.

Also, you mentioned that the missing MOV wouldn't prevent it from working... could the leaky capacitor be to blame here? If not, then it sounds like there's definitely something else wrong I'd have to track down.

It would be appropriate to check the condition of their contact groups, or better yet, dismantle them and check them individually.
You should also separately check the status of the MOSFET's of the inverter and the operation of the charger.
There are several threads on the forum that describe how to do this.
-
The rating of capacitor C53 for UPS in AC220V version is 2.2mF x 275V~.

Can those relays actually be put back together after opening them?
I can look up the info on testing that stuff.
This is a 120V unit and the capacitor at C53 is 20uF x 150V~.
I looked at the schematic you shared and the part number listed at MV2 matches the component found at MV4, but MV2 appears to be different.

The precautions one should take for checkin the mains portion or SMPS of the UPS are the same precautions one should take when working on any device powered by mains voltage. 

-Be careful when checking HV side or primary side
-It might be benefical to use a lamp series tester to protect from unknown shorts on the primary side
-If you plan to use a scope, make sure the unit is plugged into an isolation transformer.  If you plan to probe mains, you should only use methods that are safe - such as a HV differential active probe.

Those are just a few precautions one should take.  But the main thing that separates the UPS from other devices powered by mains is the UPS's battery which is usually two large 12VDC batteries in series that can deliver large amounts of current.  Keep in mind most of these devices are rated to deliver upwards of 1000W - that's a lot of power.  And a large battery that can deliver a lot of current is required.  If there is a short on the board in the charging or inverter circuit, this power can be delivered uncontrollably which could cause serious damage to life/property.  I've seen APC UPS units with shorts on the inverter MOSFETs that caused catastrophic damage and the battery protection fuses did not prevent this damage (I've done a few videos/threads on these topics).  So that is why I voiced my concern for troubleshooting these units without a DC power supply with current limiting.

Is this the lamp series tester you're referring to?:

I really hate to give up on it, but I'm starting to question whether this thing is too much for me to take on

[fmashockie]

If this kind of work interests you or you would like to do more of it, a DC power supply is a good investment.  And they can be acquired for relatively cheap both used or new.  Good current limiting control is key.  Shop around and you can definitely find a good one for $100 or less.

The video you shared is unavailable, but this is an example of what I am referring to.
Most people build their own as they are very easy to make but I think they can be purchased assembled as well.

The MOV that got destroyed looks like it is for the mains input.  So typically these will get damaged like this from an external surge from your mains supply.  Based on Asis's schematic, I don't think that capacitor C53 is the cause of the MOV going up in smoke.  But I could be wrong!  Either way, it should be replaced.  They both should be replaced, but I was just saying the MOV is not critical to the unit running.  But it is important for protection.

You can test the relays in-circuit, but you would need a DC power supply to do so.  Look up the datasheet via the part # on each relay to get the pinout.  And place a voltage across the coil (enough to switch it on; for ex. if it is a 12V relay, apply 12V).  You need to make sure that you have the polarity right.  This would not matter if the relay was desoldered/removed from the board, but usually there is a protection diode across the coil in-circuit to prevent reverse polarity.  While you are applying voltage to the coil, use your multimeter to check for continuity to mak sure it is switching properly.  This is the easiest way to do it and to confirm your relays are working properly.

You don't have to give up on it! Just be careful and take precautions.  Use the advice from more experienced members in this forum.  And invest a bit in some additional tools.

[Bretterson]

You don't have to give up on it! Just be careful and take precautions.  Use the advice from more experienced members in this forum.  And invest a bit in some additional tools.

I've decided I probably shouldn't invest the time on this right now and will try to get a different one. Maybe I can revisit it later since I'll be able to open the other one up and compare how components should test out (I'll definitely report back if I do). I do appreciate all the advice though. If someone here is interested in the project I'd be willing to sell it for a good price.