Boost Converter Mosfet heating / General Science and Electronics / Forums

Forums

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

« Previous topic | Next topic »

Boost Converter Mosfet heating

1 2 next

Move Thread

LAN_403

uzzors2k

Wed May 09 2007, 07:53PM

Registered Member #95 Joined: Thu Feb 09 2006, 04:57PM
Location: Norway
Posts: 1308

I whipped up a new and improved boost converter for my coil gun last weekend, something I've been meaning to do for over a year now. Funny what polyprojectenlitus (spelled right?) can do. Anyway I think it's designed pretty good, as there is no guesswork involved with component values, and everything should be within safe operating ranges. However the IRFP450 heats excessively when the bank voltage passes about 300V. I'm not sure what is causing it. The gate drive signal is perfect, and at 12V it should be over the saturation voltage. The duty cycle, frequency and inductance are selected so that the peak inductor current is 12A. Resistive losses should be 6W max, and the gate drive looks good, so its not operating in the linear range. The main rectifier diode doesn't heat, only the mosfet. Any idea what is causing this?

1178740380 95 FT0 Boost Converter 12v Uzzors

Dr. Dark Current

Wed May 09 2007, 07:58PM

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

Have you tried connecting the fet's gate directly to the output of the 555? I don't know why but I don't quite trust these bipolar gate-drive "amplifiers"... (and the 555 should provide enough peak current just by itself).
I also found that 12V is sometimes not quite enough to fully turn on a fet, if you up the 555 voltage a bit, does the fet run cooler?

thedatastream

Wed May 09 2007, 08:30PM

Registered Member #505 Joined: Sun Nov 19 2006, 06:42PM
Location: Yorkshire!
Posts: 329

12V should be sufficient drive for those current levels.

6W is still a fair amount of power to dissipate - what size heatsink do you have on the MOSFET?

Another possibility is that the diode is not fast enough to catch the inductor's dv/dt spikes which may result in avalanche breakdown and heating in the device. You could measure the drain voltage with a scope and see if it exceeds the avalanche rating of the MOSFET. This is a problem on PFC boost converters IIRC.

What is the part number of the 500V, 1.5A diode?

Sulaiman

Wed May 09 2007, 08:45PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

Try connecting the op-amp output to pin4 of the 555 via 1kOhm with 100nF from pin4 to 0V to prevent rapid oscillation of the loop at the preset output voltage (300V), no resistance from pin4 to 0V.

P.S. I too drive mosfets DIRECTLY from the 555 output.
I usually add diodes from 0V to pin3 and pin3 to +supply, just in case

uzzors2k

Wed May 09 2007, 08:53PM

Registered Member #95 Joined: Thu Feb 09 2006, 04:57PM
Location: Norway
Posts: 1308

I tried running the mosfet directly from the 555 first which worked fine, but I upgraded to a totempole just to be sure. I forgot the specs; It runs at 5.7kHz, 78% duty and the inductor is 140ÂµH (measured, might be larger).

I've slapped a large heatsink on it, which should tackle 6W fine. After all, I only run it for 30 seconds at a time. The heatsink gets pretty hot by the time the bank is done charging, and it seems to heat more as the voltage increases. The diode is a SFA1608G, 600V, 16A, superfast. If I can find the right probes I'll take a look at the drain voltage and see what is happening. Would a higher voltage mosfet should solve this problem, assuming it's voltage spikes?

EDIT: I just saw your post Sulaiman. I used a voltage divider because the crappy op-amp only swings down to 2 volts or so. I've set the max voltage for 425V, and at 410V the frequency increases until the 555 turns off at 425V.

Steve Ward

Wed May 09 2007, 08:56PM

Registered Member #146 Joined: Sun Feb 12 2006, 04:21AM
Location: Austin Tx
Posts: 1055

Im voting avalanching of the MOSFET, especially since it heats more as the output voltage increases. Your boost diode needs to be *right* at the mosfet connections, and you should have a low inductance ceramic or polypropylene cap connected right to the diode and the MOSFET source. Only then can you allow several inches of wiring to your main lytics being charged. The idea is that you dont want any stray inductance in the circuit, as this results in voltage transients. The more stray inductance you have, the more wasted energy. I went through all of this when designing my 8kw PFC booster.

uzzors2k

Thu May 10 2007, 06:32PM

Registered Member #95 Joined: Thu Feb 09 2006, 04:57PM
Location: Norway
Posts: 1308

How big should the capacitor be? The only suitable one I could find was 4.7nF, which seems to have helped a little. I scoped the waveform, although I didn't get much time to look at it. (Quickly rising output voltage, and scope only rated for 600V (DC+AC)) There were voltage spikes of course, however their amplitude was continuously rising. So I think avalanching is causing the heating. I guess a mosfet rated for higher voltage would've been better suited.

Steve Ward

Fri May 11 2007, 01:11AM

Registered Member #146 Joined: Sun Feb 12 2006, 04:21AM
Location: Austin Tx
Posts: 1055

Well, how hot does the MOSFET really get? How about just better heatinking? Of course, ideally you would fix it up with a better snubber. The 4.7nF is going to be far too small, but paralleling it with something like a .1uF or larger cap will work well.

WaveRider

Fri May 11 2007, 07:50PM

Registered Member #29 Joined: Fri Feb 03 2006, 09:00AM
Location: Hasselt, Belgium
Posts: 500

Another suggestion:
Check to make sure you are not on the verge of saturating the inductor core...

Try reducing the duty cycle and see if the problem goes away...

lamazoid

Sat May 12 2007, 06:41PM

Registered Member #573 Joined: Fri Mar 09 2007, 11:26AM
Location: Russia
Posts: 14

Making a boost converter with so high step-up is generally bad idea... Anyway, you can try TRANSIL supressors across the mosfet. It worked for me in similar circuit. What efficency did you get? Also, can you provide us with scope waweforms of this?

1 2 next

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