APC UPS SUA3000RMI2U Overloads when exceeding quarter load

[TysonZ]

Hi EEV,

Lurked here a few times but it seems this time I require assistance as I do not have any idea what kind of fault would cause these symptoms.
I have in my possession a 2008 SUA3000RMI2U APC UPS which seems to overload when supporting an output load exceeding about 400w. (Not Quarter I know.. )

Issue:
The UPS successfully transfers to battery on power loss with loads less than about 400w.
With a load of less than 400W while on battery, if I increase the load to exceed the 400W threshold, the UPS "Overloads". Lowering the output load back to 0W does not stop this overload. (Unsure if a feature of APC UPS?)
If the load is greater than this threshold, and the UPS fails over to battery, the UPS "Overloads".
Using a multimeter, I have identified that the output voltage of the device is 240V AC when working correctly (both on mains and on batter under 400w load), although once this threshold is exceeded, the output voltage immediately drops to about 160V AC.

Once the overload is cleared by powering off the device a strange smell can be smelt, from the main PCB, but it does not smell like burning. (Could be burning dust from the 2x large transistors?)

The device can then be immediately powered on and it resumes to work as normal so long as the load is below the aforementioned threshold...

What I have done:

• Complete reset of system settings
• Ensured battery input (2x(4x series 12v 5AH)) are all outputting correct voltage
• I have cleaned the PCB of dust and other debris to ensure no shorting.
• Observed the Mosfets and capacitors for damage (All clear)
• Briefly observed other components on the PCB for damage.

I have attached some images of the device for reference.
If you have any ideas as to what may be the cause of this issue, please let me know. I am open to any and all suggestions.

[Black Phoenix]

Please test the batteries:

What is the voltage of the bank of batteries when:

Fully Charged, No Load;
Under Load, less than 40%;
Under Load, more than 40% (the overload situation).

And if possible by banks of 4 batteries each (If I'm not mistaken it have 2 banks, with 4 batteries in the top bank and 4 on the bottom).

[TysonZ]

Good evening Black Phoenix,

Thank you for your response, I am away from the device at the moment but do have some of the information you are requesting.
Therefore I will post that information now.

Fully Charged, No Load; 54V
And if possible by banks of 4 batteries each (If I'm not mistaken it have 2 banks, with 4 batteries in the top bank and 4 on the bottom):
Correct, there are 2 banks with 4 batteries in each. These both individually read 54v

When I return I will test the batteries under the following conditions and report back:
Under Load, less than 40%;
Under Load, more than 40% (the overload situation).

In case the issue may be the batteries (Testing showed they all reported 13V), I will provide info on the batteries below.
The UPS came with the following batteries pre-installed (apologies for the ebay link, unable to find a datasheet): https://ebay.us/bneVda
Standby Float Use: 13.5v - 13.8v Voltage Regulation, 1.5A Initial Current
Cycle Use: 14.4 - 14.8v Voltage Regulation, 1.5A Initial Current

I forgot to mention that I tried swapping out each bank of 4 batteries with other 12v 7.2AH batteries I had for another UPS to make sure it wasn't a single bank of faulty batteries. (Making 1x(4x 12v,5AH) & 1x(4x12v,7.2AH)).
Doing this still caused the UPS to fail with the previously mentioned symptoms...
Perhaps it's just this brand of crappy battery I was provided?

[TysonZ]

Full response to above questions below:

Fully Charged, No Load; 54V
And if possible by banks of 4 batteries each (If I'm not mistaken it have 2 banks, with 4 batteries in the top bank and 4 on the bottom):
Correct, there are 2 banks with 4 batteries in each. These both individually read 54v

Under Load, less than 40%;54v rapidly drops to 51v in 20 seconds, then hovers around 50v
Under Load, more than 40% (the overload situation): 46.4V, It seems the failure occurred when battery voltage was below 47v.

During this testing it seems output voltage was hanging around 200v during the overload.

[Black Phoenix]

Full response to above questions below:

Fully Charged, No Load; 54V
And if possible by banks of 4 batteries each (If I'm not mistaken it have 2 banks, with 4 batteries in the top bank and 4 on the bottom):
Correct, there are 2 banks with 4 batteries in each. These both individually read 54v

Under Load, less than 40%;54v rapidly drops to 51v in 20 seconds, then hovers around 50v
Under Load, more than 40% (the overload situation): 46.4V, It seems the failure occurred when battery voltage was below 47v.

During this testing it seems output voltage was hanging around 200v during the overload.

Ok let me see if I understand totally:

All batteries read 12V or above alone and the total of all 8 on the battery cartridge is 54V at the cartridge terminal right?

Under load the battery cartridge drops to 50V with less than 40% and close to 46V above 40%, giving the overload error, right?

A new cartridge by APC standards should be 48V with a nominal capacity of 10Ah, for your info. A bank of 4x 12V batteries in series (48V) in parallel with a second bank.

The smell you feel after the UPS is turned off is like a kind of alkaline smell who looks like you can smell with your mouth (I know it's strange what I've written, but is what the smell I think it is looks to me when I feel it)?

The inbuilt self-test of the UPS says everything is OK with the battery cartridge right (Press and hold the I/Test button for a few seconds to initiate the self-test.)? I had a similar problem but with a smaller SUA1500RM2U with 2 banks of 2 batteries each and it was a battery who would drop is voltage to low 9V immediately when provided work on her.

Do you have another similar UPS were you can load the battery cartridge and see if the symptoms are the same? - Forget I saw that you already had:

I forgot to mention that I tried swapping out each bank of 4 batteries with other 12v 7.2AH batteries I had for another UPS to make sure it wasn't a single bank of faulty batteries. (Making 1x(4x 12v,5AH) & 1x(4x12v,7.2AH)).
Doing this still caused the UPS to fail with the previously mentioned symptoms...

[TysonZ]

All batteries read 12V or above alone and the total of all 8 on the battery cartridge is 54V at the cartridge terminal right?:
Yes this is correct, on average they read about 13.3v.

Under load the battery cartridge drops to 50V with less than 40% and close to 46V above 40%, giving the overload error, right?
Yes, the entire cartridge drops to 50v rapidly when load us under the overload threshold, and 46V when above the overload threshold.

A new cartridge by APC standards should be 48V with a nominal capacity of 10Ah, for your info. A bank of 4x 12V batteries in series (48V) in parallel with a second bank.
It does seem like mine follows the APC standards then.

The smell you feel after the UPS is turned off is like a kind of alkaline smell who looks like you can smell with your mouth (I know it's strange what I've written, but is what the smell I think it is looks to me when I feel it)?
I'm not entirely sure, but it's certainly not like anything I've smelt before. 

The inbuilt self-test of the UPS says everything is OK with the battery cartridge right (Press and hold the I/Test button for a few seconds to initiate the self-test.)?
Yeah it does pass according to the management interface.

I noticed that the logs of the UPS show that the %VAOut and Iout spikes during these overload sessions. Amps spike to about 18A out from the normal 0.00A, and %VAout spikes to to about 35% to 70%.
Actually looking at the logs normally it always report 0.0A out and 0% watt out under normal load, I wonder if this is an issue or if the UPS is just reporting output amps while on battery...

Below you will find a spreadsheet of the last test I have done. Note that the starting capacity was low as it took a few tries for me to get it to show both a working self test and two overloads, one of which reporting a low battery condition. This was tested with a small load of about 345W.



Here is an overload I tester earlier, notice how the battery quickly drops to 64% capacity.


I had a similar problem but with a smaller SUA1500RM2U with 2 banks of 2 batteries each and it was a battery who would drop is voltage to low 9V immediately when provided work on her.
I'm not knowledgeable on batteries, but it could be that they aren't good. The high amp output by the UPS when on battery is a bit worrying though...

Thanks for taking your time to read over this for me, I do appreciate that.

With the price of batteries coupled with the fact that I'm not confident it's the batteries, I don't really want to have to buy 8 more just to test them.
(Seems battery stores don't like you returning batteries after you've left the store unless they are actually faulty, so if it's the device and not the batteries I would be stuck with an extra 8 batteries.).
What I might do is purchase another unit... That way I may actually have a working unit that I can test the new unit batteries on the old unit. Seems to kill two birds with one stone, well a diamond in this case given the $$$. 

[Black Phoenix]

Yes all the symptoms points that the batteries itself are the fault. Specially that drop into 30% like 1 minute after the load starts.



You can try, since under 40% loads the UPS doesn't enter in overload to charge the batteries, load it until the batteries are almost dead, let it charge again and discharge, like 2 or 3 times, to try to revive the batteries. It is not guarantee than it will make them turn to life, but you don't have anything to loose with it. By my calculations a full charge of all the batteries should keep a load under 400W power on for at least 30 to 40 min. At full 2400W (3000VA x 0.8P_f) no more than 15 min.

[bombsquad]

Bump. Did you get a resolution to your issue in the end? Was it the battery after all?

I have an SUA2200RMI2U (essentially the same machine with a few less FETs and a few constants being different, as I understand it) that exhibits precisely the same issue. It works perfectly up to about 350W load, then on the cusp it oscillates between an "overload" condition and normal operation. Anything load significantly over this results in an emergency shutdown due to overload. The hum of the transformers becomes something of a rattle during this supposed "overload" period.

Having dismantled the damn thing and tested almost every part, from voltages, diodes, caps, relays, transistors, transformer impedances, and traced a few key signals around the board, I can't seem to find a component or functional fault.

Two distinct sets of batteries have yielded the same result -- on the basis of this thread, the only thing I can imagine is that the internal cabling used to assemble to batteries into their unified pack might be the limiting factor, presenting a limit to flow at higher loads. This may translate into an effort by the UPS to push more through the inverter to maintain that output draw, which it correctly realizes is too much, and thus fails safe.

The resolution to your tale - or any other helpful hints - are much appreciated.