What type of SMPS is this? (Harbor Freight Plasma Cutter)

Looks like a two-transistor (IGBT) forward converter with the diodes shown wrong.

Best regards, Spehro Pefhany

To keep it cheap. There's nothing wrong with a half bridge.

What's 'wrong' about the diodes ?

Graham

The bottom connection on BG3 and the top connection on BG4 should go to the opposite sides of the primary from where they are shown.

Best regards, Spehro Pefhany

I believe that BG3 and BG4 are needed as shown to protect emitter to collector breakdown voltage (V(br)ECS) of the IGBT's which is about 15V. The question is why they do not need the diodes you suggest Spehro, clamping transformer primary to the opposite rails. When the IGBTs are turned off there is a lot of residual energy remaining in the primary that will flyback and the additional diodes will clamp this voltage at the opposite rails. Maybe they use 1200V IGBTs and the transformer primary has a lot of parasitic capacity to absorb that energy before 1200V is reached. IGBTs can absorb lots of reverse voltage avalanche energy (Earv) so maybe just a big heat sink will do. I as you would add those diodes. Regards, Harry

Hmm, if it's supposed to be half bridge, it's missing a ground. As is, the transistors act in series; why they would draw a switch at each end is far beyond me. Engrish schematic, as it were, y'think?

Tim

-- Deep Fryer: a very philosophical monk. Website:

Normally inverse-parallel diodes are built into the IGBT half-bridge module. The parts shown are outside of the module, based on the part numbering and the quantity of diodes shown in the assembly drawing, unless some of the diodes shown are paralleled.

Also, why use two switches and two sets of driving circuitry, doubling the cost and power dissipation?

I had just assumed that the only way of making a 2+kw SMPS was full bridge (or half bridge), and just assumed the schematic to be a simplification/mistranslation. I found a near top of the line plasma cutter schematic here with similar ratings:

I'm sure this is a stupid question, but could anyone explain the purpose of R1,2 C1,2?

Thanks

It's a "simplified" schematic.

"Tim Williams" a écrit dans le message de news:BWrFf.165$ snipped-for-privacy@fe07.lga...

the

Make the switches switch together. When the switches are on you transfer power to the output, and also rise current into the magnetizing inductance. Then open both switches. The primary polarity reverses and with the right pair of (omitted here) diodes the core resets by sending back its stored energy to the supply.

If the inverse parallel diode is not built into the module it must be placed externally as shown to protect the collector breakdown voltage (V(br)ECS) of the IGBT's which is about 15V. The rail clamping diodes (D1+D2) in your link are not need if the IGBTs can absorb the flyback energy at turn off. Lots of primary transformer capacity will help. Two switches are used as your article should point out to reduce the voltage stress on the IGBTs. One of the IGBTs may also function as the pilot light. Regards, Harry

This is a different subject, but are they actually *needed*? The info I have implies that it's just a power dissipation thing with IGBTs. I don't see where much reverse current would come from, either way.

Best regards, Spehro Pefhany

Bingo.

Best regards, Spehro Pefhany

That is probably the same circuit, but with MOSFETs rather than IGBTs.

That's the power supply filter, or voltage doubler, depending on how the input voltage is set up. The resistors are to equalize the voltage across the caps when they are used in series. RL2 controls the 115/230 configuration (see CP). A similar arrangement is used in PC power supplies with a manual switch on "CP".

Best regards, Spehro Pefhany