1 MHz GDT problems / General Science and Electronics / Forums

Forums

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

« Previous topic | Next topic »

1 MHz GDT problems

first previous 2 3 4 5 next

Move Thread

LAN_403

ShawnLG

Mon Oct 15 2007, 03:46PM

Registered Member #286 Joined: Mon Mar 06 2006, 04:52AM
Location:
Posts: 399

You should read my thread on my micro SSTC build. I eventually settled on the 2N7002 mosfets because they are very fast and have a lower on state resistance compared to small signal transistors. 2N7000 are the same as the 2N7002 except that they are not SMT.

Marko

Mon Oct 15 2007, 04:11PM

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

I tried a GDT too, with success.

First, I tried a GDT on small core wound with CAT5 wire. It had terrible amount of leakage inductance and with 1nF load it gave a nice resonance with nearly pure sinewave ringing up to like 3x supply voltage. Series resistor only affected the amplitude, as expected.

When I rebuilt the GDT with some of that thin microphone mini-coax, I got practically identical copy of primary waveform at 2nF load.

Thinner coax may be better since it's has lower impedance, but I think you should be great with some of that 2-cored stereo coax used in CD-rom audio connection.

With good GDT you should have no problems.

One practical thing first; can I run a 2N2222/3906 pair directly from the VCO out on a 4046? I think the buffer makes the drive signal uneven.

I don't think 4046 will be able to drive that, at least not that well - I would suggest you to use a single class A amplifier stage with BS170 and 100 ohm resistor as I did.

This is not to scale:

I think you could hardly get better than this:

uzzors2k

Mon Oct 15 2007, 09:02PM

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

ShawnLG: I've read your thread many times. How did you make the GDT, did you use coax or just normal wire?

Marko: Thanks for the inspiration. The sine wave you got was much like mine!

I tried a 2N2222/2N3906 pair but they didn't seem to have the current capability. The waveform was almost triangle and had a low amplitude, so I went back to BD's for testing. I cut the mosfets out of the GDT drive all together. I also found a small coax cable from a magnet reader head and stole it for the GDT.

Good or bad news first?

The bad news, the waveform is more or less the same with an IRFP450. Just less noisy and somewhat more square.

The good news, with an IRF630 the waveform is much better, so I'll just use them instead.

Running the coil offline would be a bit excessive anyway.

I'll experiment some more later but I'm fairly satisfied with the gate drive right now. I should be able to make the 2N2222/2N3906 pair work since you were able to. Thanks for the help so far everyone!

Marko

Mon Oct 15 2007, 09:17PM

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

What are you driving the totem pole from? Use BS170 to drive them as I explained... I don't know if logic gates are powerful enough.
Emitter follower has voltage gain <1 so make sure you don't attempt to drive it from 5V logic or something.

Absolutely get rid of that 560 ohm resistor if you are still using it.

Then make sure you haven't swapped anything (2N3906 has different pin layout), or etc.
Always RTFDS throughly.

You should be able to get nice clear square wave with fall and rise times <40ns at least.

You won't get anywhere with BD's, they are too slow.

Marko

Dalus

Wed Oct 17 2007, 05:19PM

Registered Member #639 Joined: Wed Apr 11 2007, 09:09PM
Location: The Netherlands, Herkenbosch
Posts: 512

I had a similar problem with my drsstc which runs @ 1.4 MHz. I solved it by building a small half bridge controlled by a UCC37321 & UCC37322 combo. This half bridge drives a full bridge of IRFP450 mosfets trough a 2cm ferrite core. It sure gives a nice square wave at the gates. The only downside is that it's consuming 30 watts.

Marko

Wed Oct 17 2007, 09:50PM

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

Dalus, I really don't think your driver would consume 30W without dying instantly.

UCC's can switch few watts and dissipate a watt at most, and you don't need but like 3W in overall to drive a H bridge IRFP450's.

If it's a DRSSTC (with mosfets?) it should need only a fraction of that.

Dr. Dark Current

Thu Oct 18 2007, 12:54PM

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

Marko wrote ...

Dalus, I really don't think your driver would consume 30W without dying instantly.

UCC's can switch few watts and dissipate a watt at most, and you don't need but like 3W in overall to drive a H bridge IRFP450's.

If it's a DRSSTC (with mosfets?) it should need only a fraction of that.

I think it's quite possible, my gate driver powering two IRF740's at some 300-400khz consumes some 3-5W or so (including volt.stabiliser etc.), now multiply it by 4 for 4x higher frequency and by 2 for fullbridge (probably even more due to larger gate charge) and you're at 30W...

Marko

Thu Oct 18 2007, 01:11PM

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

It is all about stored energy, (CU^2)/2 times frequency. Driver can't dissipate much more than this in any case, unless something is seriously wrong with it.
To drive 2nF to 12V in this thread at 4Mhz I use only about a watt which is consistent with observed heating of components. (5V regulator and logic not included).

Dalus: I seem to have misunderstood your post, you use something in between UCCs and mosfets? Can you post the circuit you are talking about? Is the 30W dissipation of your bridge in operation or the driver only?

Dalus

Thu Oct 18 2007, 06:19PM

Registered Member #639 Joined: Wed Apr 11 2007, 09:09PM
Location: The Netherlands, Herkenbosch
Posts: 512

Marko wrote ...

It is all about stored energy, (CU^2)/2 times frequency. Driver can't dissipate much more than this in any case, unless something is seriously wrong with it.
To drive 2nF to 12V in this thread at 4Mhz I use only about a watt which is consistent with observed heating of components. (5V regulator and logic not included).

Dalus: I seem to have misunderstood your post, you use something in between UCCs and mosfets? Can you post the circuit you are talking about? Is the 30W dissipation of your bridge in operation or the driver only?

Here it is

(t1 is one ferrite core)
It's basically steve's mini sstc circuit driving a small half-bridge through a small gdt. This half-bridge drives the full bridge (of IRFP450mosfet's) trough a rather large ferrite core.
So I think it's all about stored energy and the large magnetizing currents caused by the large ferrite core. But it could also be that my digital meters are effected by the electromagnetic fields of my DRSSTC.

uzzors2k

Thu Oct 18 2007, 08:36PM

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

I measured the phase shift of the driver and found it to be near zero. Are BD's really that slow? I haven't found any mention of speed in their datasheets which isn't promising, but my oscilloscope measurements prove it's negligible at 1MHz. I took one shot of bases VS emitters, and I can see no noticeable phase shift. The other shot of the 4049 input VS emitters doesn't show much phase shift either. Did I measure the delays correctly?

first previous 2 3 4 5 next

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