Zener diode what could be the value

[taydin]

Hello everybody, I'm fixing a 3KVA UPS. One of the daughter boards on it had extensive damage, it wasn't even possible to look up the values of the parts. So I took the time and drew the schematic of this daughter board. It seems to be a galvanic coupling circuit between the PWM driver IC (SG3525) and the MOSFET H bridge:

[taydin]

In this circuit, Q5, Q6, Q7, Q8 were blown. One of the D9, D10, D11, D12 was also blown apart. Curiously, one had a chimney hole on top of it, but was still checking out with a DMM check. The designators on the PCB for these are ZD, so these are zener diodes I presume. Here is a picture of it (sorry, don't have a microscope cam, just stuck the digital camera to the ocular )

[taydin]

The code on the zener diodes is 6N or 8N, not very clear, but 6N is more likely. There is also the W17 code next to it.

I desoldered one of the zener diodes (not the one in the picture that has the chimney hole) and have used my power supply to try to find the break down voltage, It didn't break down even though I went up to 60V.

Based on the schematic, the zener diode is used as some kind of overvoltage protection, so I'm guessing a low break down voltage would be appropriate (like 7V or so). And it is a higher wattage part based on the relatively larger case. But in an inverter application that operates at a high frequency, I am doubting that the zener would be able to clamp the voltage in time. I would have expected some kind of a schottky diode arrangement.

So my question is, based on the design and purpose of the circuit, what replacement part would be suitable for the zener diodes? For the transistors, I will just use a switching transistor with high voltage and with a suitable CBE footprint.

[taydin]

By the way, here is the picture of the entire UPS mainboard. I wasn't able to find the schematic of this, but maybe some of you will recognize this unit (probably a generic chinese thing) and point me to a schematic:

[Armadillo]

Transformer 3&4, where is the return line? You sure you trace it correctly?

[taydin]

And here are the pictures of the entire daughter card (designators haven't been matched with the schematic)

[taydin]

Transformer 3&4, where is the return line? You sure you trace it correctly?

Yes that seemed curious to me as well, but that's all there is. I'll check it again just to be sure.

My interpretation was that Q7 pulls the MOSFET gates to ground, while Q8 pulls the gates to Vcc

[taydin]

Transformer 3&4, where is the return line? You sure you trace it correctly?

Yes that seemed curious to me as well, but that's all there is. I'll check it again just to be sure.

My interpretation was that Q7 pulls the MOSFET gates to ground, while Q8 pulls the gates to Vcc

Aah, checked that again, that other transformer pin goes to the ground plane on the top layer. So pin 4 of the transformers are actually ground.

[Armadillo]

your schematic is wrong. Q3 collector is connected to Q4 and both of them are the same device which is not the same as your schematic.
you should spend time correcting the schematic.

[taydin]

your schematic is wrong. Q3 collector is connected to Q4 and both of them are the same device which is not the same as your schematic.
you should spend time correcting the schematic.

As I said I haven't matched the designators on the schematic and on the board. I just let KiCad annotate. But maybe I should do that ...

[JFJ]

The code on the zener diodes is 6N or 8N, not very clear...

Looking at the code the other way up, it reads N8 or N9, which are the respective markings for BZX384-B68 (68V) and BZX384-B75 (75V) zener diodes. Please see page 2 of the BZX384 series datasheet:

https://assets.nexperia.com/documents/data-sheet/BZX384_SERIES.pdf

[taydin]

Looking at the code the other way up, it reads N8 or N9, which are the respective markings for BZX384-B68 (68V) and BZX384-B75 (75V) zener diodes. Please see page 2 of the BZX384 series datasheet:

https://assets.nexperia.com/documents/data-sheet/BZX384_SERIES.pdf

Indeed nothing was showing up when I was searching for 6N, so I probably had it upside down 

The digit after the N is not symmetric, so I think we can safely exclude 8, which means this is the N9 part. Thank you for pointing me at the right direction!

[LuisHertz]

Hi

Can you fix the UPS?

[taydin]

Yes, I changed all transistors on that card. I didn't find the exact zener there, but used another 75V zener in a SOT23 package and just soldered it crooked

I also had to change the SG3525 chip on the oscillator card. But afterwards, the UPS came back to life. There was really extensive damage, and I think it happened like this:

One of the MOSFET's got short circuited. Then, in the next cycle, the other MOSFET of the H bridge turned on and now the entire 310V DC voltage was now across the transformer, and huge amounts of current was flowing through it. Once the MOSFET's melted away under that high current and the circuit opened up, the whole energy that was stored on that transformer winding was dumped out to the entire circuit as perhaps thousands of volts of voltage. In total, I have changed components on 4 separate daughtercards, all MOSFET's, all IGBT's.