starter solenoids

[ronv]

For $400 you can have the recommended controller with all the bells and whistles.

 

[tcmtech]

I think you went way overboard on the pre analysis aspect of your design and shot yourself in the foot before you ever began.

When dealing with nearly 100 volts at several hundreds of amps the few watts of power loss difference between a large IGBT power block and a Mosfet block don't add up to enough to worry about especially when comparing the cost and durability differences that go with the two.

If IGBT's didnt work in that, and far higher, power levels they would not be the standard choice for power devices in industrialiual equipment like high powered VFD motor drives and multi tens to hundreds of KW switching power supplies.

Same with losses relating to switching frequency. A big DC motor doesn't a need a multi tens off KHz switching frequency either. A few hundred Hz is more than high enough and even the biggest multi thousand amp 2+ KV rated IGBT's units can do that.

As far as IGBT thermal runaway that again is not an issue if you designed your circuits and heatsinks right in the first place.

Point is I have worked with them for many years in a multitude of different applications and I dont hesitate to replace old Mosfet switching devices with properly rated IGBT ones!

 

[strantor]

I think you went way overboard on the pre analysis aspect of your design and shot yourself in the foot before you ever began. Point is I have worked with them for many years in a multitude of different applications and I dont hesitate to replace old Mosfet switching devices with properly rated IGBT ones!

I think you hit the nail on the head. Next time I will spend some time doing the hands-on part of the homework before putting all my eggs in the book basket. Your expertise is very much appreciated, and I'm always listening even if I don't sound like I am.

If you are running that high of current levels along with mechanical contactors 28 volts wont be anywhere near enough headroom.

Odds are you will blow right through that 100 volt Mosfet and destroy it the first time you use it.

Why is this? I don't remember reading anything in my quest that would indicate that it's not safe to operate a mosfet at or very near the avalanche voltage. Could you explain please?

Thanks

 

[tcmtech]

The general rule of thumb so to speak is that in any switching power supply or motor control system you have to anticipate that there are going to be inductive spikes during the switching process that could be at least two to three times the systems working voltage.

From personal experience in working on big inverter power supplies found in welders, plasma cutters, and VFD units its typical to see the switching devices rated at around 3x - 4x the peak rail voltages and 2x - 3x the peak amps.

For example typically in a plasma cutter, welder, or VFD power supply powered from a 240 VAC source the DC rail voltage would be around 340 volts and the IGBT's would by rated at around 900 - 1200 volts.
Same with 480 VAC powered units. Their switching devices will most often be rated for 1600 - 2400 VDC.

Granted you can find designs with lower peak voltage ratings in comparison to their working voltages but from personal experience I have seen that they tend to be found in the lower quality and cheaper built equipment that gets an honest reputation for not being as well built or reliable.

However if you can handle your Mosfet going dead short at full load and causing your system to go into a run away thats up to you. Personally I wouldn't try relying on the mechanical contactors rated for 12 - 24 VDC to successfully break 72 - 96 VDC at several hundred amps without having a meltdown from the arc flash when they open.

 

[ChrisP58]

I will agree with the need for more voltage margin between the mosfet rating and the applied voltage. At the current you will be running at, the transient spikes just from the wiring inductance will be significant.

One reason that IGBTs are preferred over mosfets for high voltage applications is because the losses due to the saturation voltage of an IGBT are smaller than the RDSon losses of a mosfet.

A rule of thumb that I have been told about mosfets is that for a given size of silicon, the on resistance quadruples as the voltage rating doubles. So, all else being equal, mosfets will have lower conduction losses at low voltage, and IGBTs will have lower conduction losses at high voltage. Where the cross over point is will vary with individual components and applications.