IGBT high current wiring

[tjwal]

I want to build an AC tig welder circuit. The suggested IGBT is G30N60B3D which is rated for 60 Amp continuous.

when I look at a picture of the device on digikey the leads look too small to carry 60 AMP. Assuming it is good for 60 Amp how do I physically connect a large conductor to it. I think I need about 10 gauge wire which is pretty large. Even if I go with something smaller it will still be larger than the leads.



Although I'm an EE by training I've never done any professional electronic design. On the hobby side it has all be low power analog or digital.

Please help

tjwal

 

[crutschow]

If you are planning on running 60A continuous you should use a larger transistor. For best reliability it's good practice to run a transistor at no more the 50-75% of it's maximum rating.

Also note that this transistor has a maximum collector-emitter saturation voltage of 2.1V when on (from the data sheet). That means the transistor will be dissipating over 120W. You will need to attach the transistor to a very large heat-sink to avoid zapping it from over temperature.

This points to another reason to use a larger transistor. They will have a lower saturation voltage and thus dissipate less power.

You could make a connection to the leads with a short length of smaller wire. It's ok if the wire is bigger than the transistor lead.

 

[TheNthDegree]

Hi,

As crutschow pointed out, this IGBT is basically too small for what you want to do. It is rated 60A continuous **at 25 degrees C** note from the datasheet that the current rating is 30A (halved) at 110 degrees C.

Unless you have an excellent heat sink and good luck, you will find it difficult to keep the IGBT at 25 degrees C. As soon as the temperature rises above 25C, the current handling will go down. The datasheet will contain a case temperature vs. DC current graph, see it for details.

If it were me, I'd try to use IGBT(s) which are rated at least 120A at 25 degrees C to be on the safe side...

Hope that helps,
-Nick

 

[tjwal]

The schematic calls for 4 of these in parallel with a max current of 200 amps, so they might be OK. There will be a big heat sink and fan.

But even if I go with a larger IGBT I still don't know how to physically make the connection. I don't think a copper clad board would handle the currents (or am I wrong?).

I haven't been able to find any resources on the net that go into this.

thanks for any info

tjwal

 

[blueroomelectronics]

Might help if you post your schematic.

 

[crutschow]

But even if I go with a larger IGBT I still don't know how to physically make the connection. I don't think a copper clad board would handle the currents (or am I wrong?).

Simply lap solder the wires to the leads on the transistor.

 

[TheNthDegree]

The schematic calls for 4 of these in parallel with a max current of 200 amps, so they might be OK.

Well, that changes everything. Yes, with 4 in parallel you should be okay, assuming that the IGBTs are relatively well matched so that they equally share the load.

For this kind of current, it would probably be best to ditch the PCB entirely, and use point-to-point wiring or a copper bus for the high-current path(s). For example, you could use short 14 or 12 AWG wire links which are soldered directly to the leads of your IGBTs, then connected to a common bus capable of the full current draw. Keep the links between IGBTs as short and similar in length as possible.

In this case, the smaller wire size would be okay because (presumably) the length of the link wires is short and they are in "free air" so they will be able to dissipate heat better than if they were in conduit, for example.

Hope that helps,
-Nick

 

[tjwal]

welder schematic

I'm planning on using the power section of this design
**broken link removed**

I would go with a micro for the timing circuits.

I have an AC/DC buzz box so I can use DC directly from that.

I'm just over $100 for the major items so it does look doable.

tjwal

 

[crutschow]

Well, that changes everything. Yes, with 4 in parallel you should be okay, assuming that the IGBTs are relatively well matched so that they equally share the load.

IGBTs have a positive temperature coefficient of collector-emitter saturation voltage, thus they tend to automatically share the load. If one starts to carry more of the load, it will heat up and its collector-emitter voltage will rise. This will cause some of the current to go to the other transistors, and thus better share the load.

 

[Ubergeek63]

I want to build an AC tig welder circuit. The suggested IGBT is G30N60B3D which is rated for 60 Amp continuous.

when I look at a picture of the device on digikey the leads look too small to carry 60 AMP. Assuming it is good for 60 Amp how do I physically connect a large conductor to it. I think I need about 10 gauge wire which is pretty large. Even if I go with something smaller it will still be larger than the leads.



Although I'm an EE by training I've never done any professional electronic design. On the hobby side it has all be low power analog or digital.

Please help

tjwal

well for starters junk the IGBT idea!


It sounds like you already have the 2KVA power transformer. You probably only need 50V FETs and need at 120A of current capacity and a heat sink. Now, if you want to avoid the heat sink, you can use a full bridge of FETs of 400A each or more to have a chance (eliminating the diode bridge in the process, since it would also require heat sinking).

Dan