If you want a tough problem look no further...

Hey,
I was wondering if somebody could shed some light on the problems I am having.

I'm trying to build a variable speed drive. Because of some unique requirements I have been forced to make it myself instead of buying a cheap OEM version from somewhere else. Oh what a drama this has been.

Well, my latest problem was that when a motor was connected the motor would spin for maybe 1 or 2 seconds, then pause for a second or so and then keep going. Sometimes it would stall completely. Now, when I was using rectified 240VAC the motor would run no problem and exactly like I wanted. When it was powered off 415VAC (which is what I need) it had the stalling behaviour. Weird.

I tried a lot of things to fix the problem. In my efforts I connected a much larger smoothing capacitor to the DC bus. Well, actually 2 x 220uF 400V capacitors in series, giving me 110uF total. When I was running it at 240VAC I was using a 40uF capacitor.

When I turned the drive on the motor ran fine for probably 4 or 5 seconds and didn't appear to stall (it normally would have at this point). Then there was an explosion. A fairly big explosion. It tripped the main breaker (300A) to the building. It also did something I've never seen before - the IGBT driver chip (which is a 44pin PLCC) now has a nice round hole in the casing.

This was about the last thing I expected to happen from adding a larger capacitor.

And just for everyone's information, everything is rated for 1200V.

Maybe post a schematic?

was it the motor that exploded or the cap

Here's the schematics attached - everything but the control circuit. Nothing interesting there.

The motor and caps are fine. The IGBT's and the IGBT driver chip exploded.

You are lacking any let alone adequate snubber circuits on the output and If you tripped a 300 amp three phase breaker you are pushing far more power than what a 110 uf capacitor covers.

What power level of supply are you running and what size of motor on the output?

You might want to look at this; https://www.electro-tech-online.com/custompdfs/2010/03/an-985.pdf

It seems like some of your driver circuit doesn't match what the application note shows. Also what is the single IGBT "B" do?

Only a tiny load - a 0.18kW 3 phase motor. The supply was 415V 3 phase, and there was a 10A C curve circuit breaker connected just before the rectifier.

shortbus= said:

You might want to look at this; https://www.electro-tech-online.com/custompdfs/2010/03/an-985-1.pdf

It seems like some of your driver circuit doesn't match what the application note shows. Also what is the single IGBT "B" do?

Click to expand...

I'm using the IR2233, which is basically (or so I think it is) the IR2133 but rated at 1200V not 600V. Also, I will admit that I am missing any kind of overcurrent protection in the circuit (which is why the connections are different in my circuit on the CAO/CA-/ITRIP/VSS pins).

The IGBT I am using comes in a special pack that also includes a braking IGBT ("B"). I don't need it for my application.

If the braking IGBT is connected in the circuit(it is shown in your schematic) With no snubber diodes present if the back EMF from one of the phases turning off could overpower the diode wired to the braking IGBT. That would turn on the brake IGBT causing a dead short. This is just my thoughts on your problem, I'm no expert but I'm sure one will chime in soon.

Do you have a part number for the IGBT? (and whats the ic hooked to ir2233?)

Sounds like a bus shoot through. Have you looked at the gate drive signals going to the igbt with a scope? Are you giving enough deadtime between switching? Also, some pics of the setup could be helpful.

Just wanted to point out these ICs made by Fairchild: https://www.electro-tech-online.com/custompdfs/2010/03/FS2FFSBS3CH60.pdf but they're only good for 600V. (They have free samples of them too)

shortbus= said:

If the braking IGBT is connected in the circuit(it is shown in your schematic) With no snubber diodes present if the back EMF from one of the phases turning off could overpower the diode wired to the braking IGBT. That would turn on the brake IGBT causing a dead short. This is just my thoughts on your problem, I'm no expert but I'm sure one will chime in soon.

Click to expand...

That's a good point. I will take a closer look. Unfortunately, I'm not using the IGBT pack that is shown on the schematic for my prototype because it's hard to wire into a breadboard. The circuit that exploded was using separate IGBT's and so there was no brake IGBT at all.

jrz126 said:

Do you have a part number for the IGBT? (and whats the ic hooked to ir2233?)

Sounds like a bus shoot through. Have you looked at the gate drive signals going to the igbt with a scope? Are you giving enough deadtime between switching? Also, some pics of the setup could be helpful.

Just wanted to point out these ICs made by Fairchild: https://www.electro-tech-online.com/custompdfs/2010/03/FS2FFSBS3CH60-1.pdf but they're only good for 600V. (They have free samples of them too)

Click to expand...

The IGBT I was using at the time was the IRG4BH20K from IR. The IC hooked up to the IR2233 is an IXDP630. This is an IC that takes 3 PWM inputs and outputs 6 PWM signals (3 the inverted version of the originals) and adds some deadtime. The deadtime I added was 900nS. Going off the datasheet I need 640nS at 25 degrees.

I'd like to take a picture of the setup, but I seem to have lost my camera... Bummer... Anyway, I'll try and describe it. Right now I have the 6 IGBT's on a small breadboard mounted next to a small breadboard with the IR2233 and IXDP630 chip on it. I've tried to keep the 2 as close to each other as possible and have run multiple grounds to key points to avoid noise problems.

Thanks for the Fairchild IC link. I will try to find a 1200V equivalent. I have been getting a bit suspicious of the IR2233, but yesterday I pulled open a small VSD we had sitting around the workshop and it uses an IR2233. I guess that shows that it *may* work...

KingKrak said:

I'd like to take a picture of the setup, but I seem to have lost my camera... Bummer... Anyway, I'll try and describe it. Right now I have the 6 IGBT's on a small breadboard mounted next to a small breadboard with the IR2233 and IXDP630 chip on it. I've tried to keep the 2 as close to each other as possible and have run multiple grounds to key points to avoid noise problems.

Thanks for the Fairchild IC link. I will try to find a 1200V equivalent. I have been getting a bit suspicious of the IR2233, but yesterday I pulled open a small VSD we had sitting around the workshop and it uses an IR2233. I guess that shows that it *may* work...

Click to expand...

WHAT!!! You are putting 415 volts through a BREADBOARD? You better read up on breadboards before you kill yourself or burn down your shop.

Are you basically trying to make a VFD(variable frequency drive)? For one capable to run that motor will cost you a lot less money than what you have already spent and WON'T ELECTROCUTE YOU or someone else.

Here is the price range and related capacities you need to beat.

**broken link removed**

**broken link removed**

**broken link removed**

**broken link removed**

There are other ones that are cheaper and bigger but I have bids on those so I am not showing them since I dont want any competition!

looks like the inductor is wired wrong... you need 600V caps ... i can probably find more stuff to mention but need to get going...

Hey guys,
We're actually a company that has been selling VSD's (along with other things) for a long time. We can buy them pretty cheap. The problem is we're going to be making about 5,000 of these a year. We've talked to a number of manufacturers and can buy a 240V VSD on a card for $80. But the cheapest we can get a 415V VSD for is $350. Seeing how I have kind of gotten the thing to work so far and the total cost of components is only about $110 I think it's a better deal.

I'm not too worried about using a breadboard at those voltages. The pin spacing on the 1200V IGBT's is the same as the holes in the breadboard. And I always wear safety glasses before I turn it on...

Ubergeek63 said:

looks like the inductor is wired wrong... you need 600V caps ... i can probably find more stuff to mention but need to get going...

Click to expand...

I'm using 2 x 400V caps in series. They are the same capacitors used on a 415V VSD I opened from a major and well known manufacturer. Can you tell me what's wrong with the inductor wiring? I'm not sure about the size and whether I made it large enough.

Hey guys,
We're actually a company that has been selling VSD's (along with other things) for a long time. We can buy them pretty cheap. The problem is we're going to be making about 5,000 of these a year. We've talked to a number of manufacturers and can buy a 240V VSD on a card for $80. But the cheapest we can get a 415V VSD for is $350. Seeing how I have kind of gotten the thing to work so far and the total cost of components is only about $110 I think it's a better deal.

Click to expand...

So you want US to debug your product so YOU can make money off of it?

5000 units at $100+ profit each. $500K+ in your pocket and we get what for our part of the work?

KingKrak said:

I'm not too worried about using a breadboard at those voltages. The pin spacing on the 1200V IGBT's is the same as the holes in the breadboard. And I always wear safety glasses before I turn it on...

Click to expand...

Quote from Wikipedia, " Typical specifications
A modern solderless breadboard consists of a perforated block of plastic with numerous tin plated phosphor bronze or nickel silver alloy[5] spring clips under the perforations. The spacing between the clips (lead pitch) is typically 0.1" (2.54 mm). Integrated circuits (ICs) in dual in-line packages (DIPs) can be inserted to straddle the centerline of the block. Interconnecting wires and the leads of discrete components (such as capacitors, resistors, inductors, etc.) can be inserted into the remaining free holes to complete the circuit. Where ICs are not used, discrete components and connecting wires may use any of the holes. Typically the spring clips are rated for 1 Ampere at 5 Volts and 0.333 Amperes at 15 Volts (5 Watts)."

Breadboard - Wikipedia, the free encyclopedia

In the schematic you posted, you also show "COM" pin#21 wired to the Vss pin. It, COM, is the return from the IGBT's not to be wired to signal ground,Vss.