Powerware 9120 Online UPS

[goodpsusearch]

This is another freebie I found recently. It came with 2 12V 9AH batteries that are not completely dead. It seems they still have ~1/3 of their initial capacity.

The UPS works, but displays “DC bus fault” the first time it starts. I looked around and it is probably cap related:
http://oh2mp.ham.fi/misc/powerware9120/english.html
http://www.fixya.com/support/t133894..._fault_problem

I tested every capacitor on it and only the C14 = 100 µF 63V Jamicon TK is bad (capacity: 103uF, esr: 28).

1) Do I need to change the other cap mentioned in the links?
It is a Chemicon SME 2200 µF 16V with capacity: 2400uF and esr: 0.04-0.06, tested on circuit (datasheet)
Is Chemicon SME a reliable series?

2) Can I replace the Jamicon TK 100 µF 63V with Rubycon YXF 100uF 63V?

[japlytic]

Nippon Chemi-Con SME appears to be a reliable series, and has been superseded by SMG or SMQ.
You can replace the Jamicon TK with Rubycon YXF.
Check for shorts on the DC bus as well from C40 and C42 (after discharging them), including checking for shorts on Q11 and Q12 on the power board.

[goodpsusearch]

I replaced the bad 100µF 63V Jamicon TK with Rubycon YXF 100uF 63V and now the UPS works... I'll test it more thoroughly tomorrow.

Btw it has a very useful feature. It can display the Watts and VAs of the devices that are connected to it realtime.

How accurate can I except it to be? Does it mean that I can calculate the power factor by dividing Watts with VAs?

[japlytic]

Originally Posted by goodpsusearch

Btw it has a very useful feature. It can display the Watts and VAs of the devices that are connected to it realtime.

How accurate can I except it to be? Does it mean that I can calculate the power factor by dividing Watts with VAs?

I guess so, and how the UPS calculates Watts and VA should be fairly accurate.

[biserko]

can help
My board is a model 098-12101-04 PSDR-1112NN P.W
P / N :12101-09p-638 (e1)
I appeared on the display DC BUS Fault

[Behemot]

Seems that this usually means shorted transistors and/or dried caps near heatsink. Do you have digital multimeter? Any other equipment?

[biserko]

Originally Posted by Behemot

Seems that this usually means shorted transistors and/or dried caps near heatsink. Do you have digital multimeter? Any other equipment?

Can you write in more detail which transistors, and where they are on the plate ...

[Behemot]

If you provide good enough pictures If it is more or less the same as in the first post, than that would also be capacitor C14, it's the 63V/100uF cap. Check all those transistors attached to the big heatsink in there, if you have DMM, that's a minute to check them all.

[Jooo]

I have the 3000VA version of this. It works nicely, except it starts to make annoying squealing noise after about 5 mins of running. BUT only when connected to mains, and even if batteries are full.

However, in battery mode, this squealing doesn't happen, even after long time.

Any clues what can cause this, and is it fixable?

[goodpsusearch]

Originally Posted by Jooo

I have the 3000VA version of this. It works nicely, except it starts to make annoying squealing noise after about 5 mins of running. BUT only when connected to mains, and even if batteries are full.

However, in battery mode, this squealing doesn't happen, even after long time.

Any clues what can cause this, and is it fixable?

You should check every capacitor in it (please be careful, some of them retain their charge for hours) starting from those 2: http://oh2mp.ham.fi/misc/powerware9120/english.html

[biserko]

Originally Posted by Behemot

If you provide good enough pictures If it is more or less the same as in the first post, than that would also be capacitor C14, it's the 63V/100uF cap. Check all those transistors attached to the big heatsink in there, if you have DMM, that's a minute to check them all.

where it would be located capacitor C14, it's the cap 63V/100uF

thanks

[biserko]

Originally Posted by biserko

where it would be located capacitor C14, it's the cap 63V/100uF

thanks

[biserko]

[Behemot]

Hmm, seems like different design. OK, let's start. You have photos so you know where every cable belongs. So, disconnect everything and remove the main board.

I would personally begin with the section between blue and red wire, there is some silicone with only small heatsink and some even without any. Check it. What are the small capacitors in there? They may be baked, I can imagine some heat is being produced arround those components.

[biserko]

We replaced the components that you have said, but the problem is what are the other of components very very hot, you can not keep your finger on how many of them are hot ...

[Behemot]

Does it work now?

Well yeah, you can think about adding some bigger heatsinks or doing some ghetto mods like resistor fly mod and so. If it is worth it…

[davor]

I have blown one IRF3710, near 2 caps mentioned before, and glass fuse on board (5A).
Replaced IRF and fuse but it blows fuse again. On LCD is "DC Bus Fault".
Any idea what next?

[mariushm]

A blown fuse rarely happens without a cause.

Maybe look at the traces going to the IRF and follow them back checking components in the path?

If something blows the IRF, something is probably shorted or a diode is faulty. I would go from the IRF solder points/holes back and check everything.

Check diodes, check resistors, check for shorted capacitors and if you can their ESR/if they're leaky, check npn transistors (put meter on diode check, test 2 out of 3 pins in both directions as if they're diodes)

If you can, take out the board and get a good camera with macro mode and snap some focused (not blurry) pictures of both sides, maybe someone with good eyes here can spot some components that look burnt out or some broken/burnt copper traces on the pcb.

[davor]

Thanx. Its 22:00 and I've just replaced rest 3 IRF's (S/D=short) but spoted on 2 IRF's on the right, traces of some kind of burns/cuts of on source pins!!!
Now Ive left UPS to filll batts to full and will post results tomorrow.
1st IRF was burnt so hard that pin was in the air!!! Temp. was so high that components around have black marks!
Fuse (5A) is soldered to the board as, by design, it's not ready for easy replace.

[davor]

No change at all - fuse keeps blowing.
Replaced also 2 condensers, checked diode near IRF (first from the left, traces of burning around it).
With power cable connected fan starts (as before) and I can browse thru options on LCD but when I press power - bang.
Tryed to start device just on batts but some smoke came out - didnt have time to spot from where and didnt want to do it again.