Small problem in high Amps SMPS power supply

[BGAmodz]

Hi everyone .

I have an SMPS welder , giving normally an output of 80 volts and current up to 140 Amps .

The problem is that i get like voltage pulses , and not a constant output , the IGPT cooling fan and the leds keeps turning on and off .

The unit got a quad opamp lm224 amd another 8 pin IC that i guess its the PWM chip UC3845 , am wandering if the problem is on the control board or in the primary side ?


Here is a picture of the card :

 

[rumpfy]

Hi BGA,
I feel for you. These things can be very hard to fault find because the controls run at mains potential. I use a pair of back to back 24 volt transformers to isolate the control circuits when fault finding these things, but when you reach the point of full power testing, the isolating traffos have to go and then one is in the dark again. The current sensing is difficult to simulate.
I have a little 'italian' general purpose 140 amp DC welder which has been sitting in the shed for years cos it is too hard to work on. I think you will just have to grind away at it.
i did an esab 140 amp welder which was quite simply engineered and this used a SG3524/3525? IC. I think the SG has pushpull drivers but the UC does not.
good luck with it.

 

[BGAmodz]

I have repaired a lot of SMPSs before this one , but this welder control area is full with epoxy .

BTW how about the transformers ? i have here two transformers one is an smps pulse transformer and one is current transformer , do you have any experience on how to check them ??

 

[tcmtech]

A brand and model would be very helpful information.

Is it cycling on and off at a consistent rate and duty cycle while welding or surging while welding?

 

[BGAmodz]

A brand and model would be very helpful information.

Is it cycling on and off at a consistent rate and duty cycle while welding or surging while welding?

It is cycling on and off while the unit is simply on without weldind and even when i try to weld there is apparently no output current .

here is a service manual attached :

 

[rumpfy]

Hi BGA,
When I said I have 2 back to back transformers to isolate the mains supply, it is so that I can check the operation of things like the pulse transformers, in operation. I imagine you are referring to the small pulse transformer/s used to drive the mosfet gates? I seem to remember seeing lots of small red colored pulse driver transformers in all the welders I have seen.
I seem to remember doing an ESAB welder, and the current transformer had broken loose and the winding had broken from the PCB. Just a lot of fiddling to get it back together and terminated. It was a strip wound secondary.
If the 'cycling' problem is apparent at 'idle' then you can use the isolation traffos to see what is happening. BUT if the problem is happening during welding use, then this is where the fun starts, unless you cannot isolate the control circuits so that you can get your scope to work on it.
I didnt scan through the manual I have to say, but I hope my comment helps you.
Now that I read your post again, have you checked through the current limit/current control section. One small welder I used to weld up an exhaust pipe was VERY impressive with its current control. I used a 2 mm rod to weld up some 18 gauge steel and it was easy to do without blowing holes in the steel. To me that would be the place to start. BUT, there should be output voltage on no load, yes??

 

[BGAmodz]

I am checking the current limiting area and it seems that all transistors and diodes are good , still 2 ics i didn't check LM224 , and a CURRENT-MODE PWM CONTROLLER 2845B , those are controlling the output current modulation right ?

 

[tcmtech]

Start from the beginning electrically. Are the primary DC rail capacitor banks charging up to their correct voltage?

If so are the power supplies for the lover voltage control circuits powering up and putting out their correct DC voltages?

If those check out good then keep following the circuits forward until you find which IC is in charge of triggering the cycling.

 

[rumpfy]

Following on from tcmtech, there is a voltage regulator device U2 which generates some low voltage supplies. I dont know the device U2. But these things often startup by pulsing the internal switch. It is clear that the Vdd for U2 comes from diode D9, after the IC starts correctly, so if there is a starting problem you could see what you are seeing. To check this I would look at V1 (12 volt) and Va and Vb. My guess is that Va is the same as Vb and are 24 volt.
This is tcm's recommendation;
"Check all DC supplies"

 

[BGAmodz]

OK thank you so much , i will check the voltages and see where the problem could be .

 

[BGAmodz]

Hi rumpfy am back on this topic .

I have checked the power supply phase the high capacity capacitors have about 300 volts DC on them .

Now i have found a short in one of the two secondary rectifying diodes ( hyperfast dual diode rectifier ) , can this be responsible for that cycling on and off ?? i guess those are directly related to the output ??

 

[tcmtech]

Yes it would.

The power supply tries to power up but the shorted diode triggers it's overload protection function so it turns off and ties again.

 

[rumpfy]

Hi BGA,
Not to ignore what tcmtech says.
Is that the part D6 or D8? the BYV27-200 diode.
If so they are describes as "Ultra fast Low loss Avalanche controlled rectifiers. These are fairly fancy and you'd need to get the correct replacements.
There is a bit of a discrepancy with what you say, because the diodes D6 and 8 are single diodes but you refer to a "hyperfast dual diode".
But in any event, a shorted diode is a NONO.
Just check the identity of the diode again!!.

 

[BGAmodz]

Actually there is a small discrepancy , because am working on a very similar welder .
Maybe those dual diodes are not drawn as dual on the schematic , but anyways the one that i found short is 60cph03

Link for the datasheet
https://pdf1.alldatasheet.com/datasheet-pdf/view/447629/VISHAY/60CPH03-N3.html

 

[rumpfy]

In the drawing you sent, there are two diodes like you say at the actual welding output. These diodes are duals, there are two of them, and they are reverse biassed across the output. (see page 8 of the drawing; D22 and D23).
On page 8 of the drawing there is a current transformer T2. A faulty diode could (would) upset the current control arrangements and cause this 'at rest' problem you are having. A code number "60CPH03-N3" sounds exactly like a code number "STTH6003CW" to my way of thinking. Looks like you are on the right track.

 

[BGAmodz]

Hi again

i have successfully started the welder , but after a while a friend of mine was looking at the card , suddemly a big blow has happened and the circuit breaker has gone off , but the weird thing is that the welder pcb looks in good state , there is no burns .

any ideas ?

 

[rumpfy]

So I gather you replaced the diode then went to power up, and !!!
back to basics again. Have a re-read of post #2.

 

[tcmtech]

Good chance one or more of the primary switching devices shorted out.

Were you actually welding or putting a load on it when it popped something or was it just sitting there?

Try disconnecting the DC output from the main HV capacitor bank going to the main switching devices and see if it will power up.

 

[BGAmodz]

Yes indeed i got a bunch of primary diodes ( zener diodes and skottky diodes ) shorted after that blow as well as an IGBT .
I have found some 1N5408 diodes , can i use them instead of the BYS 045 ? and i have only 0.180 volt drop on them when i use diode mode , but the 1N5408 has 0.500 v .


I also remember that before replacing the dual diode , i have replaced by mistake a small sot-23 zener diode by a normal one at the control section , maybe that was also a factor for the shorts ? because i don't really see how the shorts have happened , really weird indeed .

 

[rumpfy]

I'm sure tcmtech and I will be able to help a bit more, but YOU have the hard grind ahead.