mosfets getting too hot / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

mosfets getting too hot

1 2 next

Move Thread

LAN_403

IamSmooth

Thu Mar 24 2011, 05:12AM

Registered Member #190 Joined: Fri Feb 17 2006, 12:00AM
Location:
Posts: 1567

I'm confused.

I originally had some spikes at the transitions on the gate signal to my mosfets

I originally tried 2200pf mica caps in series with a 4.7R resistor across the drain-source. Now the spikes were gone and the gate signal looked good from 0 to 700vdc. However, even with an input current of 8A going to the half-bridge the mosfets were getting very warm.

I changed the snubber to just a 1000pf cap across the drain-source. The gate signal still looked good, but the mosfets were still getting warm at low current.

Finally, I removed the caps. Now, I had spikes on the gate at low voltage and this disappeared as the voltage increased. Now the mosfets were cool to the touch.

What am I missing? Should I keep the cap but use a higher resistor? How is having the cap across the drain-source causing the mosfets to get warm. I feared letting the current get high because I'm allergic to semiconductor smoke.

James

Thu Mar 24 2011, 05:34AM

Registered Member #3610 Joined: Thu Jan 13 2011, 03:29AM
Location: Seattle, WA
Posts: 506

How are you measuring the input current? Have you tried scoping the current waveform? You may have a perfectly reasonable RMS current with enormous spikes.

IamSmooth

Thu Mar 24 2011, 05:47AM

Registered Member #190 Joined: Fri Feb 17 2006, 12:00AM
Location:
Posts: 1567

current waveform looks good with and without the snubber at high voltage.

Maybe I'm dumping power back into the mosfet when the switch closes.

Electroholic

Thu Mar 24 2011, 12:05PM

Registered Member #191 Joined: Fri Feb 17 2006, 02:01AM
Location: Esbjerg Denmark
Posts: 720

Yes, energy stored in the snubber capacitor is discharging into your mosfet. Try an rcd snubber circuit.

Experimentonomen

Thu Mar 24 2011, 01:38PM

Registered Member #941 Joined: Sun Aug 05 2007, 10:09AM
Location: in a swedish junk pile
Posts: 497

And make the snubber resistors 10 ohms, may help a bit too.

IamSmooth

Thu Mar 24 2011, 02:39PM

Registered Member #190 Joined: Fri Feb 17 2006, 12:00AM
Location:
Posts: 1567

Yes, last night I ordered 24R resistors after I made the post. I am figuring there may be a point when both mosfets are conducting and the switches are allowing 330vac/9R to pass through momemtarily.

The RCD snubber wouldn't help as this does not change the maximum current it will pass. I will see how this fairs.

IamSmooth

Sat Mar 26 2011, 10:55PM

Registered Member #190 Joined: Fri Feb 17 2006, 12:00AM
Location:
Posts: 1567

Well, I just tried 2200pf/24R snubber and the mosfet still gets warm with fan cooling. The current going into the half-bridge is 8A. If I remove the snubber it is cool to the touch at the same settings.

The weird thing is that the gate signal (gate to source scoped) looks better than the signal without the snubber. I am going to try 1000pf/24R snubber, but I expect it will get less warm, but still warmer than without the snubber.

Does anyone with a lot of experience with mosfets and snubbers understand what is happening? Am I changing how fast the switch is turning on? Is there anything I can do, other than not using the RC snubber? I already have a FRED across the drain-source.

Dr. Dark Current

Sat Mar 26 2011, 11:06PM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

Depends on what load are you driving. If the output voltage does not commutate to the other switch afther the first one switches off AND the FET's or gate signal turn on transitions are slow, the other device will "eat away" most of the energy stored in the snubber cap.

Experimentonomen

Sun Mar 27 2011, 09:49AM

Registered Member #941 Joined: Sun Aug 05 2007, 10:09AM
Location: in a swedish junk pile
Posts: 497

A guy on IRC pointed out a fatal flaw in your schematic that may be the culprit. The diodes you have in series with the drains show to be 15V zeners when googled, if this actually is the case, the FREDS across D and S may not do anything and that you get freewheeling through the fets internal slow diode.

I have neither diodes, nor snubbers in my induction heater inverter and my switches whatever they are gets warm too despite a fan, though i've measured a inverter current up to 65A.

IamSmooth

Sun Mar 27 2011, 03:57PM

Registered Member #190 Joined: Fri Feb 17 2006, 12:00AM
Location:
Posts: 1567

Are you referring to the 65PQ015 diodes? They are schottky barrier diodes rated for 15v and 65A. If he thinks this is a zener I'd like to see the link.

However, I am not using these. There are more schematics at the end of the tutorial. I am using the MBR7030

So, the question still remains as to why they are getting warm? I don't think the chip resistor transferring that much heat to the sink. The mica capacitors only hold 11W. Also, the heatsink was getting warm when I was just using a capacitor without the resistor.

Could the snubber somehow be rendering the freewheeling diodes ineffective?

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob