APC BR1500GI F02 Output Short

[paul_g_787]

I have an APC BackUPS Pro BR1500GI which I bought new in 2013. It has been sitting in the cupboard faulty until now.

The problem arose after a power spike The AC voltage went to 315V for a couple seconds (I have no idea why to this date). The UPS went to battery, then when it went to switch back to AC I had the error code F02 displayed with a constant screech for help that cannot be muted. There was no funny smells or magic smoke at the time of the fault.

From the sound the UPS is making it appears to be in terrible pain!!

This screeching error code happens constantly. With batteries and without. With or without devices connected to the outputs. The unit cannot be powered off or silenced apart from unplugging the AC.

I of course, tried new batteries, which did not help so I bough a second-hand unit (which I have been using since) and put the batteries in that one which is fine now 2 years later.

All the manual says regarding F02 is:

F02 On-Battery Output Short
Turn the Back-UPS off. Disconnect non-essential
equipment from the Battery Backup outlets and the turn
Back-UPS on.

This advice does not help at all.

I got onto APC three times in 2019 and they basically said "naf off" to me when I asked for customer support. I even said I was prepared to send it to them for service and they said I would have to post it to the USA and the cost is $950 plus shipping and they won't guarantee the repair. I definitely will not be buying new products from APC ever again!

Searching the F02 fault online shows up dozens of other people with the same F02 fault code, some with other models, some saying their UPS were brand new!!! So it is a common issue! Especially with the BR1500 models which are mentioned a few times

So now I have a little more electronics knowledge I have been brave enough to take it apart. I found two faults.

The first was C83 100µF 25V. It was bulging and had an ESR of 124Ω and capacitance of 1.6µF. I have replaced this with a new capacitor.
The second fault was another capacitor that had one leg never soldered to the pad from new! It was just flapping in the wind! I tested it's value and ESR and it is good so I soldered it in place correctly. (see photos).

I tested the ESR of every other Electrolytic cap on the board and they are good.

I have tested the MOSFET transistors on the big heat-sinks for shorts. They appear to be fine!

So I am hoping someone here may be more knowledgeable than I am regarding this issue and might be able to help me find the fault.

[fzabkar]

Something must be sensing the current draw. Look for low value, high power resistances. Alternatively, the unit may be sensing the current via a winding in a transformer.

I would think that the battery charging circuit must have some current sensing circuitry.

[johnkenyon]

Something must be sensing the current draw. Look for low value, high power resistances. Alternatively, the unit may be sensing the current via a winding in a transformer.

I would think that the battery charging circuit must have some current sensing circuitry.

To the best of my knowledge APC use current transformers to measure AC current, in this case on the output, and in some cases on the AC input feed as well,
It is possible that the circuit that conditions the current transformer output has been damaged - either the load current is being reported as too high, or maybe reported as being zero when the the UPS knows full well that it isn't.

My advice would be to try and find a schematic*, and try to trace how power gets from the line to the sockets, keeping your eye open for a transformer marked "CT" (=current transformer) rather than "T" (=input power transformer or inverter output transformer)

If the BR1500 is anything like my SU1400, then look for any transformer marked "CT" rather than "T". For example on the SU1400 there are CT1, CT2.
Then look at what each current transformer feeds - look for clamping diodes, burden resistors across the output, then look at where the signal goes afterwards.
On the SU1400 CT2 senses the output current (it's the last component before the "HOT-OUT" connector, and the transformer output goes through two op-amps before arriving at an ADC.

If you have a schematic, you have a fighting chance of at least locating the fault. Without one (IMHO), you might as well give up...

/JK

* Use the elektrotanya.com search, search for APC then manually look through the results.
 

[Dinesh6252]

I had the same output short issue with this ups and have successfully fixed by the replacing the 4 mosfets at the output. This reply is for all the future visitors looking to fix the issue. The Apc back-ups pro 1000 uses four Irf740 mosfets, whereas the 1500va model uses fdp22N50N. Replace all four mosfets .

[paul_g_787]

I had the same output short issue with this ups and have successfully fixed by the replacing the 4 mosfets at the output. This reply is for all the future visitors looking to fix the issue. The Apc back-ups pro 1000 uses four Irf740 mosfets, whereas the 1500va model uses fdp22N50N. Replace all four mosfets .

Hi Thanks for that info! I still have the faulty BR1500GI UPS. I will dig it out and have a look at the four mosfets.

Don't suppose you have a photo of the one you repaired to help me locate these four mosfets?

I remember I tested all the large transistors for shorts and couldn't find an issue before. Maybe I missed one or didn't test properly. 

Any tips how to test them? would you test a mosfet in the same way as a BJT?

I have had a look and there are two different IRF740 mosfets I can get. The IRF740PBF or the IRF740APBF.
Hopefully it is clearly labelled on the package. I will have a look and see.

[Dinesh6252]

I had the same output short issue with this ups and have successfully fixed by the replacing the 4 mosfets at the output. This reply is for all the future visitors looking to fix the issue. The Apc back-ups pro 1000 uses four Irf740 mosfets, whereas the 1500va model uses fdp22N50N. Replace all four mosfets .

Hi Thanks for that info! I still have the faulty BR1500GI UPS. I will dig it out and have a look at the four mosfets.

Don't suppose you have a photo of the one you repaired to help me locate these four mosfets?

I remember I tested all the large transistors for shorts and couldn't find an issue before. Maybe I missed one or didn't test properly. 

Any tips how to test them? would you test a mosfet in the same way as a BJT?

I have had a look and there are two different IRF740 mosfets I can get. The IRF740PBF or the IRF740APBF.
Hopefully it is clearly labelled on the package. I will have a look and see.

Opening up my 1500 va UPS and showing you the mosfets at the out output. You can check all three legs of the four mosfets with multimeter without removing them. Mine 1000va ups had two mosfets shorted but i replaced all four.

Here it is :

[paul_g_787]

This is the inside of my BR1500GI, there are a whole bunch of MOSFETs by the looks of it from the picture.

I do remember I did test between the pins on each one of these for shorts originally but never found a problem before.

I will try them again. Not sure which ones are the output stage MOSFETs though.

[Dinesh6252]

The ones with two legs are diodes and the others with 3 legs are all MOSFETS. Have you checked the output relays for any short?

[paul_g_787]

The ones with two legs are diodes and the others with 3 legs are all MOSFETS. Have you checked the output relays for any short?

OK thanks for pointing that out, I have tested all the diodes and they are not shorted.

I was measuring a possible short on the 4 MOSFETs but could not get a definite reading So I have de-soldered them and will test them out of circuit and will post back later.

There are 4 MOSFETS, they are all IRF2805. I believe this is the datasheet here:
https://www.infineon.com/dgdl/irf2805pbf.pdf?fileId=5546d462533600a4015355de928518eb

I also tested all the relay switches using my bench power supply. I set the power supply to 24V 30mA and used my multi-meter probes plugged into the power supply to apply power to the relays. All 6 of the relays switched just fine when power was applied.

The 5 black Relays are a 793-P-1C-S and this is the datasheet.
https://eu.mouser.com/datasheet/2/378/793p-347351.pdf

The White relay I cannot read the part no as it is worn out but the rated voltage is 24V also.

[Dinesh6252]

The ones with two legs are diodes and the others with 3 legs are all MOSFETS. Have you checked the output relays for any short?

I also tested all the relay switches using my bench power supply. I set the power supply to 24V 30mA and used my multi-meter probes plugged into the power supply to apply power to the relays. All 6 of the relays switched just fine when power was applied.

Did you tested the NO NC points of the relays with COMMON point? Most of the times the relays switches perfectly fine but its NO NC points are Burned . To test the relays you should first test all of them with the multimeter on continuity mode. The COMMON and NC point should give beep. When you apply power to the relay , COMMON and NO points gets connected and should not give a beep on COMMON and NC and vice versa.

[paul_g_787]

I also tested all the relay switches using my bench power supply. I set the power supply to 24V 30mA and used my multi-meter probes plugged into the power supply to apply power to the relays. All 6 of the relays switched just fine when power was applied.

Did you tested the NO NC points of the relays with COMMON point? Most of the times the relays switches perfectly fine but its NO NC points are Burned . To test the relays you should first test all of them with the multimeter on continuity mode. The COMMON and NC point should give beep. When you apply power to the relay , COMMON and NO points gets connected and should not give a beep on COMMON and NC and vice versa.
[/quote]

I tested the common to NC connections with no power and they all have continuity.
But I did not test between common and NO. I will definitely do that when I get home! Good shout.

[paul_g_787]

I also tested all the relay switches using my bench power supply. I set the power supply to 24V 30mA and used my multi-meter probes plugged into the power supply to apply power to the relays. All 6 of the relays switched just fine when power was applied.

Did you tested the NO NC points of the relays with COMMON point? Most of the times the relays switches perfectly fine but its NO NC points are Burned . To test the relays you should first test all of them with the multimeter on continuity mode. The COMMON and NC point should give beep. When you apply power to the relay , COMMON and NO points gets connected and should not give a beep on COMMON and NC and vice versa.

I tested the common to NC connections with no power and they all have continuity.
But I did not test between common and NO. I will definitely do that when I get home! Good shout.
[/quote]

Alright I tested all the relays between the Common and NO pins and used my bench supply to turn the relays on and yes all of them switch and have continuity in the NO position.

The 5 black relays are all SPDT, the white relay is DPDT and I tested both poles and can confirm they are working.

Interestingly one of the black relays stays on for 5 seconds after switching it, probably a capacitor to hold it in the on position to stop it turning on and off too fast I reckon.

[paul_g_787]

This is the first time I have properly delved into MOSFETs so please be patient with me on this one.

So I found this article very useful: https://www.homemade-circuits.com/how-to-check-mosfet-using-digital/

And I followed this procedure for testing the four IRF2805 MOSFETs which are N channel.

The pinout is Gate, Drain, Source when looking from the front of the TO-220 package.

How to Check N-Channel Mosfets

1)  Set the DMM to the diode range.

2) Keep the mosfet on a dry wooden table on its metal tab, with the printed side facing you and leads pointed towards you.

3) With a screwdriver or meter probe, short the gate and drain pins of the mosfet. This will initially keep the internal capacitance of the device completely discharged.

4) Now Touch the meter black probe to source and the red probe to drain of the device.

5)You should see an "open" circuit indication on the meter.

6) Now keeping the black probe touched to the source, lift the red probe from drain and touch it to the gate of the mosfet momentarily, and bring it back to the drain of the mosfet.

7) This time the meter will show a short circuit (sorry, not short-circuit rather "continuity).

The results from the point 5 and 7 confirms that the mosfet is OK.

Repeat this procedure many times for  proper confirmation.

For repeating the above procedure each time, you will need to reset the MOSFET by shorting the gate and the drain leads using meter probe as explained earlier.

The results are just as the article describes:
I first short Gate and Drain with a screwdriver.
Then using my meter in continuity / diode test mode, I measure from Drain to Source it reads OL (overload).
Then I momentarily move the red probe from Drain to Gate keeping the black probe on Source at all times.
Then move the red probe back to Drain and I get a continuity reading.

All four of the MOSFET switch in this fashion as expected and I have repeated the test 10 times for each MOSFET to confirm and wrote down the readings..

The average readings my meter give for each MOSFET is as follows:

MOSFET 1: 0.17
MOSFET 2: 0.28
MOSFET 3: 0.25
MOSFET 4: 0.37

I am wondering why such a difference in the readings? Does this perhaps indicate a fault?

IRD2805 Datasheet:
file:///home/paul/Documents/Electronics/Datasheets/Transistors/MOSFET/Infineon-IRF2805-DataSheet-v01_01-EN.pdf

[TheMG]

I am wondering why such a difference in the readings? Does this perhaps indicate a fault?

Well that's not exactly a very "scientific" method of testing, so the results will vary a bit depending on how much charge you put on the gate of the MOSFET. The voltage supplied by the multimeter might not even be enough to fully turn on the MOSFET. Also it's normal to have some variance between components, no two MOSFETs of the same type are going to be perfectly identical.

The important thing is, that the MOSFETs are not shorted and appear to switch on/off normally, which this simple test has proven, so its safe to say these MOSFETs are not at fault.

Realistically, checking for short circuits actually would have been good enough, as power MOSFETs virtually always fail with a short circuit between two (or all 3) pins. If there is sufficient energy to blow the MOSFET open-circuit, there will be visual clues.

[paul_g_787]

I am wondering why such a difference in the readings? Does this perhaps indicate a fault?

Well that's not exactly a very "scientific" method of testing, so the results will vary a bit depending on how much charge you put on the gate of the MOSFET. The voltage supplied by the multimeter might not even be enough to fully turn on the MOSFET. Also it's normal to have some variance between components, no two MOSFETs of the same type are going to be perfectly identical.

The important thing is, that the MOSFETs are not shorted and appear to switch on/off normally, which this simple test has proven, so its safe to say these MOSFETs are not at fault.

Realistically, checking for short circuits actually would have been good enough, as power MOSFETs virtually always fail with a short circuit between two (or all 3) pins. If there is sufficient energy to blow the MOSFET open-circuit, there will be visual clues.

OK that is good to know. This was what I was thinking too but as I said I haven't had much experience with MOSFETS so was just making an educated guess. So at least these appear to be working OK which is good news.

Question is now what to check next for the F02 Output Short error? A previous post mentioned current transformers so I will have a look see if I can find anything like that next unless anyone has any other ideas.

[paul_g_787]

Something must be sensing the current draw. Look for low value, high power resistances. Alternatively, the unit may be sensing the current via a winding in a transformer.

I would think that the battery charging circuit must have some current sensing circuitry.

To the best of my knowledge APC use current transformers to measure AC current, in this case on the output, and in some cases on the AC input feed as well,
It is possible that the circuit that conditions the current transformer output has been damaged - either the load current is being reported as too high, or maybe reported as being zero when the the UPS knows full well that it isn't.

My advice would be to try and find a schematic*, and try to trace how power gets from the line to the sockets, keeping your eye open for a transformer marked "CT" (=current transformer) rather than "T" (=input power transformer or inverter output transformer)

If the BR1500 is anything like my SU1400, then look for any transformer marked "CT" rather than "T". For example on the SU1400 there are CT1, CT2.
Then look at what each current transformer feeds - look for clamping diodes, burden resistors across the output, then look at where the signal goes afterwards.
On the SU1400 CT2 senses the output current (it's the last component before the "HOT-OUT" connector, and the transformer output goes through two op-amps before arriving at an ADC.

If you have a schematic, you have a fighting chance of at least locating the fault. Without one (IMHO), you might as well give up...

/JK

* Use the elektrotanya.com search, search for APC then manually look through the results.

OK so there are indeed two transformers here labelled CT10 and CT11 just as on your model. See the image attached, they are right in the middle of the image between the black relays.

So these are used to sense the current? One then I assume is used to sense the output current which is how the UPS knows the load. What is the other for?

Unfortunately I cannot find a schematic for the BR1500GI. ElektroTanya was where I first searched but despite looking manually as you suggested I still could not find one.

[SoundTech-LG]

Curious that anyone would dream they could find a UPS schematic...

[tomar1]

Hello, I have many schematics, just the one in this one, no.