DIY repair of a welding inverter mma 250

In detail: do-it-yourself repair of a MMA 250 welding inverter from a real master for the site my.housecope.com.

Compound:
master oscillator - uc3846dw, tl082 and 2 pcs. tl084i, buildup - ao4606, keys - gw45hf60wd, output rectifier - stth60w03cw
They brought it without any signs of life. The check revealed a dead roll at 12 V (exploded) and 4N90C. I changed it, I turn it on. Power supply +24, +12 and -15, everything is stable, there is a saw on the master, the output is silent. I further check the elements for deadness - the diodes are alive, I have not checked the keys yet, there are two small scarves in the key chains on which in the middle there are 2 either a dinistor or a zener diode. in general, I did not find data in the tyrnete. Marking BM1238 and BM1243. Maybe someone can tell me? In the board, one side does not ring at all, the other - as if a capacitor is charged, and then infinity. It should be?

It would not hurt to have a diagram from him, but I can't find something. Found a couple of similar ones, but a little not that. If there is, please share. A device with a vertical arrangement of connectors.

is there a processor? I did not indicate it in the composition, but I cannot understand from the pictures
Check the keys. I personally solder each transistor and check it. It's hard to find a defect there.

Radist morze, BMxxxx? These are bidirectional zener diodes in IGBT gates at 15v, you can set both 15v and 18v. Download information on SMAJxxxxx and make sure. Yes, in principle, any circuit with such a set of circuits as in Gerrard Edon mma-250 is confirmation of this. the numbering will be different.

REKKA, but where does the processor come from? This is not a 20-30 mowing machine.
Irina Slava, thanks for the comprehensive answer. I watched some kind of circuit, and I also came to the conclusion that these are zener diodes, only in that circuit they are inversely connected in series. And I already know about numbering. It's just that the composition is slightly different. Apparently, here is 3846 with external excitation, and this generator is on tl082. After it there are 2 pieces of tl084i, and then 3846. And in that diagram, everything is on tl084.
found a broken diode. one of the counter-parallel included in the tl082 strapping. Now I will look for Old and Replacement.

Video (click to play).

the diode was in a semi-torn state, if you press it down with a probe, it rings. on the board at first it also called, then stopped. I changed it, but it's no use.

Radist morze, the network has a MMA ZX7-225 scheme, here it is. close to required or ZX7200IGBT.

this scheme is suitable for my Dnieper, it is also three-story. and this is a stranger. " e-don ”single board. Well I write above that with a vertical arrangement of bayonet connectors.

REKKA, what have the keys to do with it when the control impulses do not come from the micro? on 3846 there is a saw on leg 8, there is an impulse on leg 10, and the exit is dead.

by the way, I thought that 3846 was dead, replaced - the same thing. tl082 also replaced, there is no sense either. I sin on tl084i, but I don't have them

here the ZX-7 diagram is similar, but not completely identical in details.

REKKA, at first I also thought that dead keys can plant an impulse, but there are still field workers between the microra and the keys. Yes, and I soldered the keys, the effect is the same. on the other hand, broken keys will not send an impulse. there is a trance between the field workers and the igbt. No, there is a problem somewhere in the generator.

I think I get it. The blown up stub microcircuit is most likely 15 volts, not 12. I was confused by someone's post on the Internet that the opamp can have an oblique power supply. Having looked through several schemes, I did not see a single one where it would be +12, -15 and +24. Everywhere food is +15, -15, +24. I don't have any 15 V rolls now, I need to connect from the laboratory power supply unit. I will unsubscribe based on the results. Perhaps later, because the lights are turned off.

Guys, I was right! I changed the roll 12 to 15 and the impulses started running. And why didn't anyone correct me right away? I wrote at the beginning. I am assembling the apparatus. I'll try to cook and unsubscribe.

The sparkler is working, but my opinion about it is a shitty device. In principle, it cannot give out the declared current of 250 amperes, since the keys, working in pairs, are at 45 amperes. in total, each shoulder is also 45 amperes. The datasheet says that this is the maximum current.Suppose that in the impulse mode it is two times more, totaling 90 each arm, which means 180 the entire bridge. The question is, what 250 amperes can we talk about? The Chinese apparatus is the Chinese current. I tried to cook it. My “Dnipro MMA-200” cooks better and produces more current. This is not an advertisement for Dnipro, this is just for comparison. Verdict - don't buy gowns.

- the bridge pumps the primary. in the secondary - its own current and voltage. and the number of turns in the secondary.

KRAB, sorry, I figured it out last night too. I came here to correct the message, and here is a new post 🙂 Outstripped!

but all the same, three-story buildings are better, in my opinion.

I put 110 amperes on edon, I cook a profile pipe. Shit seam. I bet on my own - a completely different matter. In general, I cook it with my apparatus at 75-100 amperes, depending on the place of the seam. And edon on the 110th “shelf” does not warm up, but I am not talking about the rib at all.

You can, of course, write off everything on the nonlinear dependence of the regulator in edon. There is a digital scale in mine, so I don't bother with the position of the regulator and the discrepancy between its nonlinear characteristics and markings on the body. Although the scale can also be incorrectly set if someone clocked it.

So your “Dnipro mma-200” is a 100% Chinese device, don't look at the name, Image - DIY repair of a welding inverter mma 250


If you already want to have a purely native inverter, take Paton, this is a Ukrainian assembly
Read also:  Sewing machine podolsk 142 DIY repair

tynalex, the Ukrainian assembly will now take almost nothing, they do not bring them to us. and according to your first link - an American iPhone is also made in China. Yellow-horned production is cheaper. Norwegian seiners take the caught fish to China for processing, and then the finished products are transported to Norway. Estimate how many man-hours the crew puffs, how much fuel, but it is still cheaper for them, because fish processing is very expensive in Norway. I once wanted to make a bungle for myself, but in terms of details it came out about two thousand hryvnias, and I didn’t take that into account, but I simply didn’t find something and didn’t know the prices. And it still needs to be done. As a result, he rummaged around and bought himself a factory one, in a suitcase, and for another 970 hryvnia, it seems. The delivery cost seems to be 1040. And they are already boiled-overcooked. recently the non-stick has stopped working, but that's another topic. And in general, this topic has been closed for two days already, we will not throw a flood.

These devices have been known for a long time and there are 1: 1 schemes for them (I have for a long time in the folder Image - DIY repair of a welding inverter mma 250