eltitulo

7

hace 5 años

First of all thank you for your help with these data sheets. Capacitors C18-C22-C23 and C24 are 10 nF True ?.
I tell you what happened, I was welding at low power with 2mm electrode (I think that forced it I have no idea) and I heard a small noise like an arc inside the machine.
good to disarm !!! The first thing I could see was that of the 2 510 ohm resistors that one had, it was open. from there following the naked eye 2 igbt in short desolde the other pair and also, in short. So I have 4 igbt in short and testing the resistance of 5.1 and 15 I saw them open.

eltitulo

7

hace 5 años

That's why those smd capacitors generated me doubt, I have basic electronic knowledge, I'm electromechanical hehe

marcelo_moy

16

hace 5 años

I can not really know what happened to it, the capacitors are very rare to burn, I do not think it is the 2 mm electrode, what if it influences the point of operation of the machine are the ground and support cables of the electrode that must be those that came with the machine or the same section without a significant part of cut wires, and the power cable that does not cause a very large voltage drop, and of course the supply voltage. The 510 ohm resistors are the ones charged by the large capacitors when the machine is turned on and then sticks the relay.

marcelo_moy

16

hace 5 años

When a short circuit is produced because 2 igbt of the same branch are burned, the voltage drops, the 24 votes fall, the relay opens and the short-circuit current passes through the 510 ohm resistors burning them. Verify that the fan works well when changing everything. the first test do it with a 100w incandecent lamp in series, it will turn on a short moment while charging the capacitors then you must paste the relay and the fan, and start investing if you get there is a good sign. Take out the lamp turn on again and measure the output voltage that should be close to 70v.

marcelo_moy

16

hace 5 años

It would be good before hitting the electrode again, to test it with a small nicrome resistor that supports approximately 2500w and to see or with an oscilloscope that the cuasisenoidal output wave of several khz decreases the useful pulse width. when connecting the resistance and also when varying the potentiometer. the maximum current will not be reached due to the high resistance value with respect to the sticking of the electrode. Or measure the current with a DC RMS amperometric clamp

marcelo_moy

16

hace 5 años

I am designing a test to simulate the uc3846 (integrated of the control of the source of current) the detection of the short circuit to verify that it stops sending pulses to assure that reaction when it occurs, and the igbt are not burned.
The sensor or system that detects the overtemperature may have failed and the igbt did not work properly supporting the current.
This circuit does not feel the current that the igbt take from the capacitors, it has no control over whether the mosfets of the same branch are fired at the same time, this destroys them instantly.

eltitulo

7

hace 5 años

The truth is that you are a teacher and I think the ground wire was cut with filament. For now I will wait because those igbt are expensive but just change everything I warn you the truth thank you very much !!

eltitulo

7

hace 5 años

Now tell me why I see you are a megaelectronic hehe where can I start? Any good course or electronic book? Thank you!

marcelo_moy

16

hace 5 años

It depends on your previous knowledge, there are many, even videos, the basis of electronics does not change, youtube looking for a video I think is the fastest, as if you had your own teacher