UPS and inverter questions

[andrew77]

Re: UPS and inverter questions

Oh yeah, thanks for participating in my thread. I really appreciate this.

[andrew77]

Re: UPS and inverter questions

2 The burnt track is in fact a fuse F1. This is quite a common trick, to reduce the track width to create a fuse. The fact that it has blown indicates a fault current. The circuit needs to be traced from this point to find out what components are supplied. You will find a short there.
3. Maybe your shorted transitor is the cause. Have you lifted the transistor legs to confirm a device short? When you do also check the diode alongside, make sure it is not shorted.

There was a trace that was blown on the other side of the board and I repaired that first before replacing the 94-2358. This caused the F1 trace to blow, because after I fixed the trace, the UPS wouldn't do anything at all. It did actually show some signs of life before.

I soldered up the gap after I replaced the 94-2358 from the other UPS that I have. That 94-2358 was definitely shorted, if you look closely at the first picture at the top of the thread you can see a little chunk missing from the part.

After I replaced the shorted MOSFET and soldered the blown traces, the thing shows signs of life, but it still doesn't work right. Probably some of the other components are bad.

That part (94-2358) is for sure a MOSFET. There's a test that I found on google that shows you how to do a simple test with a multimeter to prove that the gate still works. This part worked with the test. The other part that I pulled last night, 94-4311, doesn't act like a MOSFET when you do this test. Maybe this part that I'm testing is damaged it did get hot enough to make the tin on the leads melt off, but it's not shorted. It's probably not a MOSFET. I'll pull one that isn't heat damaged and retest.

[PCBONEZ]

Re: UPS and inverter questions

I haven't ID'ed 94-4311 but I've been looking.

I emailed IR and told them what I thought about their cross reference not being able to ID their own parts. It wasn't pretty..

[andrew77]

Re: UPS and inverter questions

I emailed IR and told them what I thought about their cross reference not being able to ID their own parts. It wasn't pretty..

[andrew77]

Re: UPS and inverter questions

I don't know what my problem was earlier today. I used this link to figure out if these parts are MOSFET's:

http://www.uoguelph.ca/~antoon/gadgets/mostest.htm

Now the things test like they are supposed to. So they have to be MOSFET's. Maybe I've been inhaling too much lead or something. I did have a really nasty headache when I woke up this morning and it lasted almost all day. Thankfully it's gone now.

So, would it be a good idea to look at some other schematics from APC and maybe use parts that they use on that?

[PCBONEZ]

Re: UPS and inverter questions

I had a bad day like that recently and it turned out to be a bad test lead on my meter.

I hate it when that happens!!!

.

[davmax]

Re: UPS and inverter questions

Unfortunately the APC schematic may be for somewhat different circuitry. I would not use it as a reliable reference or part substitution. The parts in your inverter need to be identified concisely. I will do some further checking.

[davmax]

Re: UPS and inverter questions

There's 3 of these 94-2358, there's 4 of these 94-4311 and 4 of these other p19nb2. Here's some pictures. The first one is the 94-2358. The second and third is 94-4311. The fourth and fifth is p19nb2.

That is a complex circuit. I suspect that this product is a sinewave output job(capacitor and inductor in the output). The P19NB2 is made by ST Micrelectronics but must be a special. A search of their site reveals nothing with that number. You said these are in series connection, I would suggest that there are two pairs of two series connected MOSFETs. Is that right?

You need to try to establish what circuits the F1 fuse supplies for this seems where the fault is. The main fuses are apparently intact (two blade fuses side by side) have you checked these?

[andrew77]

Re: UPS and inverter questions

The fuses are good. I looked up stp19nb2 on Mouser.com and came up with STP19NB20. This might be equivalent. It says 200v 19amp N channel MOSFET.

The STMicroelectonics website doesn't seem to have anything about this part on it, but if it is similar then Mouser.com does have them.

How does the UPS know when it is being overloaded?

[andrew77]

Re: UPS and inverter questions

Is there a way to find out if the MOSFET's are leaky or weak?

[davmax]

Re: UPS and inverter questions

The fuses are good. I looked up stp19nb2 on Mouser.com and came up with STP19NB20. This might be equivalent. It says 200v 19amp N channel MOSFET.

The STMicroelectonics website doesn't seem to have anything about this part on it, but if it is similar then Mouser.com does have them.

How does the UPS know when it is being overloaded?

Very interesting. Why a 200V device? Tell me more about this inverter. Is it a UPS? What is the output voltage? Do you know if it had sine wave output? It uses a 12 volt battery, right! Is there connection to AC power for maintaining the battery and running a load and battery creates load power when normal AC power fails?

Usually there is a current sensing that detects overload and should protect the unit. Clearly there is a weakness somewhere. You do not seem to have traced what that blown fuse supplies. Trace using an ohmmeter. A need to know.

[davmax]

Re: UPS and inverter questions

There is a leak test, but it requires a voltage supply up to 200Volts in this case that has a high internal source resistance. I have such a device , but no help to you. The test takes the MOSFET up to it's max rated voltage and the leakage current is measured to be within spec.

For you, the simplest method right now is to test each MOSFET with a multi meter set to say 2000(2K) range. Place positive lead on the centre leg or metal base and test both the other legs. There should be no circuit. A short or low ohms means a bad device. This does depend on external circuitry, you should however be able to compare devices.

If you wish to check if the MOSFET switches on apply +5v to the gate (pin1) via a 10 to 20K resistor with negative to source (pin2). Then measure as above again. This method should not damage any other circuitry.

[andrew77]

Re: UPS and inverter questions

I apologize for not making it more clear what I'm working on, so here it is:

http://www.shentech.com/1055427sn.html

The output is supposed to be about 115 volts, but I think I was seeing 95 volts. I'm not too good at electronics and I don't think this meter is giving me a totally accurate reading since I don't know whether or not it's output is a sine wave or a modified sine wave. Pretty sure it's a modified sine wave, but it does seem to have more electronics than it should.

I dismantled a Belkin 550VA earlier today and took some pictures to see how it was put together, and it only has 4 MOSFET's for the inverter part in it. There's also only one transformer in it.

I'm trying to find where the blown trace goes to, I guess I'll have to unsolder where I fixed it at and I'll be able to see where it is going. I'll get back with you when I figure it out.

Let me know if you need more pictures because there's nothing stopping me from taking more. Just let me know what part you need to see.

[kc8adu]

Re: UPS and inverter questions

your ups is a smps that makes approx 170v dc and chops it into a squarewave with those 200v mosfets.just like a cheap inverter.

[davmax]

Re: UPS and inverter questions

thanks Kc8adu. I was wondering what design approach was being used. This would explain the MOSFETs of low voltage and the 200V variety. A two stage conversion. 12V to 170V then 170DC to 115AC. The 170V being regulated and protected hopefully.

[davmax]

Re: UPS and inverter questions

Thanks for brief data on UPS. I see it is circuit breaker protected. I guess this breaker is at the output. Can you please confirm.

[davmax]

Re: UPS and inverter questions

By the way. How was the F1 fuse bypassed? It can create further damage if you do not replace witha similar level of fusing. I would have stripped a multistrand wire and used one or more thin strands of copper conductor (to approximate the original blown copper).

[andrew77]

Re: UPS and inverter questions

With the F1 fuse, all I did was bridge the gap where it was blown with some solder. The area that I had to repair the break was really tight. There's a little inductor that looks like a transformer but it only has two wires, then there's the two fuses that are in the way, and the two capacitors.

The circuit breaker is inline with the power cord. I guess the breaker is there to protect the UPS, and it only protects the load from utility power?

Anyway here's a picture of where I had to solder the thing. The capacitor that is near is about 16mm tall and maybe 15mm wide.

[davmax]

Re: UPS and inverter questions

1. I understand that the breaker is in the input power lead , right!

2. It can be fatal to do what you did. ie bridge out a fuse with solder. Use the correct fuse rating at all times. Go under the PCB if necessary.

How is circuit tracing from F1 and MOSFET testing? Needs to be checklisted.

[PCBONEZ]

Re: UPS and inverter questions

http://rocky.digikey.com/scripts/Pro...97&M=STP19NB20