Not enough gate drive / General Science and Electronics / Forums

Forums

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

« Previous topic | Next topic »

Not enough gate drive

Move Thread

LAN_403

IamSmooth

Fri Sept 03 2010, 10:28PM

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

I am striving for a more stable drive for my induction heater. Sometimes it is very stable. Other times it is not. For this question, I am MANUALLY tuning the tank. There is no uP or PLL. The frequency is locked.

So, I have two pairs of mosfets for a half-bridge. Each mosfet is driven through a 10R resistor with a ferrite bead at the junction. The gate drive lead is shielded. When all four are running the signal gets a little unstable at higher currents. If I disconnect any one of the four gate drive leads, the signal is totally stable. I scoped the gate signal and can see that the slope is a little steeper when there is a smaller load.

I am using the TC4421 and TC4422 which drives a gate-drive transformer. The gate drive chips have 100uf capacitors on the supply lead to ground as well as 1uf and 0.1uf. I have 15v regulated going to the chips using a 3A voltage regulator.

I ran into a problem when I was using 5R on the gate drive. The lower resistance caused some ringing.

So, what do I do? It seems I need to get to a maximum on-state more quickly. I can use a different voltage regulator like 18 or 20v. The TC4421 are rated for 20v. The mosfet can handle 30v on the gate. Any other suggestions? If I go with a higher voltage regulator, is there one that can supply more than 3A of current? If not, what can I do to supply power to my gate-drive chips?

Should I have a separte regulator powering each chip?

Is this a reasonable solution for more current?

Marko

Fri Sept 03 2010, 11:04PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Hi Jon,

First of all - what makes you think your gate crive is inadequate? If your mosfets are popping there must be some more radical reason why it's happening.
You have not told what frequency are you using - I wouldn't expect a too high one for an induction heater, and 10ohm gate resistors are probably ok anyway.

If you get few volts of ringing with smaller resistors, I don't see what's wrong with that either, especially if your mosfets are allowed 30V on gates? (I'm curious, what are the devices anyway?) If ringing is too much, you can always simply clamp it with zeners.

If you're still keen on faster waveforms, this is what I would do:

Firstly, there's no point of using more than 15V supply. Instead, I would go on for a better GDT. My way of creating performance GDT's is wrapping 10-20 turns of RG174 coax around a ferrite core. Then I use the shield as a primary, and the core as a secondary. Such a transformer is used to drive only a single device - providing superbly low leakage inductance. With such transformers, you could probably even get away without the resistors, although i wouldn't recommend that.
GDT's designed this way will also have no trouble driving even largest bricks.

Also, if you're having trouble with regulators, you could simply not use them at all - power your gate drive chips from an unregulated, but well filtered supply. A volt or two or ripple is really completely irrelevant to them. (Unless the'yre UCC's and you need to care about 16V max limit).

I wouldn't expect power draw as high as 3A at 15V with an induction heater that is not using bricks - what frequency and what devices are you using, anyway?

Marko

IamSmooth

Sat Sept 04 2010, 03:01AM

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

Every time I disconnect any of the four mosfet gate drive leads the waveform is perfect. If all four are connected it is a little jittery at higher currents. I can see that that slope of the gate drive is affected when I disconnect one of the leads. This leads me to think I don't have enough current supplying the gate drive for the initial demand.

I plan I building a small 18v regulated supply that can deliver more amps and see what happens.

Marko

Sat Sept 04 2010, 03:29PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

This leads me to think I don't have enough current supplying the gate drive for the initial demand.

What makes you conclude so? Have you atually measured the current drawn by gate drive chips?

If I had gate drivers that required so much current (which I never did for <500kHz drives anyway) I'd probably just use a 12V transformer of adequate power, rectified and filtered with a ~10000uF cap. This would always yield over 15V to the drivers and assure they are stable even with high loads. I've actually used such a supply for ages - and when I recently switched back to a regulated supply I just ran into trouble with their dropout.

IamSmooth

Sat Sept 04 2010, 07:30PM

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

I took out my power supply for the inverter. I had a 470uf cap after the regulator. I replaced it with a 39000uf capacitor and it seemed to clean things up.

I plan on adding the outboard transistor now. I am not sure if I will change out the 15v regulator with an 18v regulator. The tc4421/2 can go as high as 20v. Any recommendations?

Marko

Sat Sept 04 2010, 08:15PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Hi,

If you ask me, I don't think an 18V power supply will make any difference compared to a 15V one. Actually 18V seems kinda too close to rating of your gate drive chips.

Marko

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