PlasmaSonic II based SSTC, with poor performance / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

PlasmaSonic II based SSTC, with poor performance

first previous 4 5 6 7 next

Move Thread

LAN_403

Cotlone

Tue Oct 07 2008, 03:28AM

Registered Member #1383 Joined: Mon Mar 10 2008, 11:09AM
Location: Wollongong, NSW
Posts: 27

Yeah I was think of trying that, it's just around the corner. Unfortunately the core there with the highest permeability is L8 material which is only 1500 (initial) . The core I'm currently using is supposed to be about 2000 (initial).

I measured the inductance of a winding of 15 turns to be 540uH, this gives an Al value of 2400 using L(mH) = Al * (N/1000)^2. So this seems to match the supposed Al = 2250 of material 77 within 7%.

Cotlone

Thu Oct 09 2008, 02:30PM

Registered Member #1383 Joined: Mon Mar 10 2008, 11:09AM
Location: Wollongong, NSW
Posts: 27

Ok so when I connect a smaller power MOSFET up to the gate drive transformer with a gate resistor of 75 ohms I get a beautiful signal across my secondary, textbook waveform. So could the much larger input capacitance of the original MOSFET be doing this? The original MOSFET is called the IXFN48N50 and has a C_iss of 8400pF while the P16NF06 (smaller mosfet) has a C_iss of 315pF.
What do you think?

HV Enthusiast

Fri Oct 10 2008, 02:49AM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

Its your transformer. I use the same circuit as you and drive large IXFN44N50's and IXFN48N50's, even IXFN20N100's which are the 1000V variant with higher gate capacitance.

The waveform doesn't have to look "textbook", but it should look much better than the waveforms you provided. It should still resemble a very good looking square wave.

Cotlone

Fri Oct 10 2008, 04:16AM

Registered Member #1383 Joined: Mon Mar 10 2008, 11:09AM
Location: Wollongong, NSW
Posts: 27

Hmm, I think my tracks will let me down to some extent. I wound the trifilar winding even tighter and shortened the leads right up. This signal is pretty overdamped because I am using a 10ohms resistor on the gate - but its ten fold better than before. Can I further improve my results by getting a core with a higher permeability (currently 2000)?

1223612143 1383 FT54770 Gdt Sec 10r Mosfet Tight

- cheers

HV Enthusiast

Sun Oct 12 2008, 12:46AM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

that doesn't look too bad.

what are the rise times on those waveforms?

LithiumLord

Mon Oct 13 2008, 10:07AM

Registered Member #1739 Joined: Fri Oct 03 2008, 10:05AM
Location: Moscow, Russia
Posts: 261

The bad thing I see here is a small peak already showing up on the zero line. If with a lower gate resistor (I normally use 2.35 Ohms per gate, but I'm driving my fets directly) the rise and fall times will be less then 1000nS I'd say it may go, however that peak forming up may cause a minor shoot-through, which is definitely bad, especially with such a long rise/fall times as the heat losses will increase dramatically. Keep trying, but if everything with the GDT fails you can try a direct drive, this will add you three small low-voltage transformers but will definitely work well.

Tom540

Mon Oct 13 2008, 05:15PM

Banned on 3/17/2009.
Registered Member #487 Joined: Sun Jul 09 2006, 01:22AM
Location:
Posts: 617

Dr. Spark wrote ...

Tom540 wrote ...

I find it weird everyone is so anti magnet wire.

because it failed on me................

Tom lighten up, if want to use magnet wire go for it, no one is stopping you or is saying your coils will fail. I am just stating what I have found to be true of my coils (and peers) and want to ensure I share it with the great 4HV community!

Cheers,
Ch

Oh i know but if you think about it its odd saying don't use magnet wire on a transformer? You know what I mean just because thats it sole purpose. You know that ferrite cores are meant to be coated or painted before you use them, but of course use whatever your heart desires.

Cotlone

Tue Oct 14 2008, 09:16PM

Registered Member #1383 Joined: Mon Mar 10 2008, 11:09AM
Location: Wollongong, NSW
Posts: 27

Sorry about the delay here, I've been a bit caught up of late.

I'd just measured the rise time to be 460ns and a voltage swing of +- 8.2V.

I'm using the a 1.2ohm resistor and those small, low freq overshoots have returned after the falling edge. They are about +- 3.0V. The cores currently only have 9 turns on them and the trifilar winding is fairly tight. This max voltage isn't good enough, these FETs are just chillin' (or warming) in the linear region most of the time. I'll try going for a few more turn on the cores if that will help the voltage.

HV Enthusiast

Wed Oct 15 2008, 01:59AM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

Cotlone wrote ...

Sorry about the delay here, I've been a bit caught up of late.

I'd just measured the rise time to be 460ns and a voltage swing of +- 8.2V.

I'm using the a 1.2ohm resistor and those small, low freq overshoots have returned after the falling edge. They are about +- 3.0V. The cores currently only have 9 turns on them and the trifilar winding is fairly tight. This max voltage isn't good enough, these FETs are just chillin' (or warming) in the linear region most of the time. I'll try going for a few more turn on the cores if that will help the voltage.

Its your core. Those cores are no good.

Get a new core. No reason you should be gettng such horrendous waveforms. You should be getting risetimes of 25-50ns easily and almost perfect pulses.
Don't even bother fixing what you have.

Again, get new cores.

ragnar

Wed Oct 15 2008, 03:07AM

Registered Member #63 Joined: Thu Feb 09 2006, 06:18AM
Location:
Posts: 1425

Looking at the way your cores split in the first post, something is terribly wrong. With the largest of Jaycar's ferrite cores, fifteen turns should be about perfect at 300kHz. Don't bother wrapping it with tape, just go crazy with the Cat5E.

first previous 4 5 6 7 next

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