3 phase inverted- ac lines to rectfier linked together?

Thread Starter

Matt Traven

Joined Apr 17, 2018
8
I am working on repairing a motor controller for a washing machine motor (120vac -> 220 VAC (3 phase). After replacing 2 shorted IGBT's (and that not fixing the board), I have started to trace paths and see what else was wrong, power input seemed like a good place to start. Right away I noticed that the two ac inputs for the bridge rectifier (d10xb60 https://www.mouser.com/datasheet/2/366/J534D10XB-1203178.pdf) are shorted together. To clarify, I don't mean the chip has been damaged, I mean those center pins are intentionally soldered to the same trace on the board. I have 2 of these drivers, different boards and it is done this way on both. This also leads to some very strange readings on my DMM (strange enough that I grabbed another DMM to ensure mine wasn't broken). By strange I mean: there is sometimes a volatge with only one lead connected, pin->pin i have 0v (as expected by the short), if I connect to one of the pins and then touch the other probe to the frame of the washer(presumable ground) or even just to my body, I see 120VAC. If I look at the dc output pins, I see like 345vDC.

So my questions are:
why is this done this way?
how does it even work with those two lines shorted together.
 

dendad

Joined Feb 20, 2016
3,073
It may be to increase the current capacity. As you have a 120V input and 220V drive, the circuit is probably a voltage doubler.
That has 2 diodes and 2 capacitors, so as a bridge has 2 pairs of diodes, why not just hook them in parallel, as it sounds like they have done.
VoltageDoubler.png
 

Thread Starter

Matt Traven

Joined Apr 17, 2018
8
thank you, I did not recognize that, but with your help see it now. That helps me understand the board a bit, but doesn't get me any closer to fixing it.

When I first pulled the board, the 12A sloblow was toast and putting in a new one popped as soon as it turned on. So I pulled the 6 IGBT's and found 2 (same bank) were shorted, I replaced them....the fuse no longer pops, but the inverter will not spin my motor. Now the igbt's were shorted c->e and g->e, so I am concerned that short to the gate went back into the logic circuit. Each of the igbt's gates is run through a 220 resistor(a very small one), that is still in tact. Does that mean my logic circuit didn't get hit?

Also the board has no burnt spots and all of the traces look to be fine. So I am not really sure where to go from here in terms of diagnosis, open to suggestions! thanks.
 

dendad

Joined Feb 20, 2016
3,073
With most of these type of circuits, check the electrolytic caps too. Often the caps drying out cause lots of problems. If you have to replace one cap from the doubler, change both.
The drive circuit could be dead. Can you measure the low volts supply to the control circuit?
What info do you have on the board?

Remember! Be careful when working on the board. These bite!
 

Thread Starter

Matt Traven

Joined Apr 17, 2018
8
I meant to include: none of the caps are popped, although I did replace the pair on the second board I have (lets call that board B). and I have probed all the resistors to ensure no shorts / open circuits. I have done a limited amount of probing at the live circuit, cause you are right they do bite. Could I probe the logic circuit, sure, but it is difficult to know what to look for. I have next to no documentation on this. I did notice I slow building voltage on what I think is the balancing line to the motor (goes to motor chassis), dont know what that means, maybe someone else does.

board: sole 22187000.0 it is a motor driver designed for motor type: 20585000 (3 phase induction motor, wye configuration)

any other details I can try to answer, just ask.

Thanks.
 

Thread Starter

Matt Traven

Joined Apr 17, 2018
8
yes that is the one, here are a couple images.
if you see the "1" in the top left corner, thats what I called igbt1. Using that convention, igbt1 & igbt2 were shorted as described previously.

The molex near the top is : from bottom to top: AC hot, AC nuetral, ground/balance line, 3x motor windings
-so thats power in & output

the one near the bottom is trickier(10 pin connector, 9 used), these are the lines that instruct the driver to run at what speed & direction. But I really have no idea what goes where or even what values the driver is expecting. I have a diagnosis flow chart that says if pin 1 or 9 are less than 50vac (with respect to ac neutral), then replace the timer. Now those values are fine, but from this I think we can infer pins 1 & 9 tell the driver what to do.

pins 3& 4 are easily traced to the tach, if I measure continuity over those and spin the motor manually, i flashes of continuity as I would expect, so the tach seems ok, but I dont know anything about the circuit that reads it.

While I wait and see what you have say, I think Ill try to trace those inputs and see if I cant make more sense of the logic circuit.
 

Attachments

dendad

Joined Feb 20, 2016
3,073
There is a fair chance the FET drivers (U7, U8 and U???) are dead to.
How are your surface mount repair skills?
Poking around on this board is potentially very dangerous and unless you know what you are doing, it is best to just replace it. I see there are some on Ebay if you can't get a new one from the manufacturer.
 

Thread Starter

Matt Traven

Joined Apr 17, 2018
8
There is a fair chance the FET drivers (U7, U8 and U???) are dead to.
How are your surface mount repair skills?
Poking around on this board is potentially very dangerous and unless you know what you are doing, it is best to just replace it. I see there are some on Ebay if you can't get a new one from the manufacturer.
Yah I was thinking those might be bad, looking at those chips, I have the ability to replace them, though I would like to confirm they are bad before I do so.

I have already replaced this board, and a lot of other parts on this washer, not dying to spend another $100 when its probably a $3 component thats bad.

As for poking around, for live measurements mostly I have been soldering wires on so I can do safely measure things. I have done some fractal burning that works with 2000-10000vac, which will kill you as a sure as a gun shot to the head. My point is that I understand the risks.

But poking around the disconnected board last night did reveal a problem: one of the AC connections was broken between the pin & board on the board connector, I fixed that last night. The first thing I do after I have tried the board in the machine and need to work on the bench is discharge those big caps...this time I though it wise to measure the voltage on them before I discharged. I found one was charged to ~140vAC, but the other had nothing on it...fairly certain thats the problem I need to fix now.
 

dendad

Joined Feb 20, 2016
3,073
Have you changed the bridge rectifier?
There may be a load on one cap more than the other, like one bleed resistor has gone open.
 

Thread Starter

Matt Traven

Joined Apr 17, 2018
8
Have you changed the bridge rectifier?
There may be a load on one cap more than the other, like one bleed resistor has gone open.
I pulled the bridge rectifier tested it, swapped it with the one on board B(also tested fine). It made no difference, so I put the original back in. Now that I think about it, that was all before I repaired the faulty pin, though the bench test should still have been valid.

I am unfamiliar with the term bleed resistor, could you please clarify?
 

Thread Starter

Matt Traven

Joined Apr 17, 2018
8
never mind, I just googled it. So its seems your concern was the one left with a charge, whereas I was more interested is the one with no charge. I think even with the new board, a charge was left on the caps. I was measuring very soon after the board was live, and the voltage was falling (slowly), I thought that was due to my repeated measurements, but maybe its just a pretty high value resistor...
 

Hymie

Joined Mar 30, 2018
711
I would venture that the two inputs of the bridge rectifiers that are shorted together are configured as indicated in the diagram below – converting two or more phases to DC.

3 phase rectifier.jpg
 

Igor58

Joined Apr 10, 2019
1
Hi Matt,
I have the same board with obvious damage to Q11 and Q12 (both are MPSA42 transistors) and Q9 (ZTX458). These are easy parts to find. My problem is that SMD resistors R203 (near Q11) and R74 (near Q9) have burned holes in them preventing me to read their value.
If you still have your boards around, could you kindly get the information I need (3 digits)? I can see them on the photos you've provided but, can't read their values. These components are all located in the center of the board, near the big 680 uF filter cap. One needs sharp eyesight or a good magnifier :). Thanks!
 
Top