Why does my ZVS driver run fine under 9V but the MOSFETs get warm right after that?

Nov 08, 2026 Last reply: 6 days ago 3 Replies

First reply

CopperPilot

Both things you mentioned are real, but the copper plane is the bigger culprit. Copper clad looks like a heatsink but the tab has to get heat thru an insulating pad (or through the solder joint, if you mounted them bare) into a thin layer of copper with almost no mass behind…

Read the full reply

I built one of those cheap ZVS driver circuits on a piece of bare copper clad, point to point style with the transistors screwed straight to the board so the copper acts as a heatsink. Under about 9V it works great, the work coil heats a bolt glowing red and the transistors stay cool to the touch. The moment I push past that, the little TO-220 tabs start getting noticeably warm even with the coil spinning the same load. Is that just becasue the current climbs with voltage and the copper plane can only pull so much heat away, or is something else going on? My gate drive is coming off a simple zener and resistor setup, nothing fancy. I checked the resistors and they read fine. Could the gates be switching too slowly at higher power and cooking the transistors that way? Also curious if the hand wound coil matters here. I wrapped it from bare copper wire with the turns not quite touching, and it sits parallel with a row of film caps. Wondering if the spacing being uneven makes the switching less clean as the voltage goes up. Any pointers appreciated, I'd rather fix it than keep cooking parts.


Both things you mentioned are real, but the copper plane is the bigger culprit. Copper clad looks like a heatsink but the tab has to get heat thru an insulating pad (or through the solder joint, if you mounted them bare) into a thin layer of copper with almost no mass behind it. Below 9V the transistors barely dissipate anything so it hides the problem. Push the voltage up and the losses climb fast, roughly with the square of the current, so the little warm spot you feel at 12V is way more heat than it sounds like. A proper finned heatsink bolted behind the tabs, even a small one, makes a huge difference. I mount mine on standoff screws with mica or silicone pads and a real sink, and they stay cold at voltages where the copper version got painful to touch.

On the gate drive side, a zener and resistor setup usally works fine untill the supply sags or the resistor runs hot and the zener voltage drifts, then switching gets sloppy right when you're pushing hardest. I put a small 12V supply on the gates through a gate transformer on mine, around 30 turns of hookup wire on a ferrite ring, and it cleaned things up a lot. Your uneven coil spacing probably matters less than the caps, though. Are those film caps matched, or just whatever was in the drawer? Mismatched values make the two sides switch at slightly diferent points and one transistor ends up doing more work.

Not buying the zener drift theory myself. I ran a plain resistor and zener gate setup for ages and it never once drifted at high voltage. What actualy got me was the inductance from having the gates wired too far apart, so one transistor would switch a hair before the other and they'd fight each other through the supply. Shortening those leads and adding a small resistor right at each gate base sorted it out. Might be worth checking how symmetric your layout is before blaming the zener.

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required