classic 4046 PLL tesla coil tips / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

classic 4046 PLL tesla coil tips

1 2 next

Move Thread

LAN_403

Shrad

Sun Jun 12 2011, 11:21PM

Registered Member #3215 Joined: Sun Sept 19 2010, 08:42PM
Location:
Posts: 780

hi all

I finally tied together my first SSTC with a classic PLL driver

it has given its first light and I'm happy, but I'm asking your advice here

30VDC, single IGBT (27n120), everything is running fine but there is a bit of heat in the IGBT

what can I do before feeding it 170VDC 2A from my voltage quadrupler ?

I tested it and it popped after perhaps 2 seconds (but I had streamers!)

what do I need to keep it safe?

the IGBT switches the primary (20 turns 1mm enameled wire) to ground and I tried with an external diode

Daniel Kramnik

Mon Jun 13 2011, 02:52AM

Registered Member #3885 Joined: Sun May 15 2011, 12:47AM
Location: Newton, Massachusetts, United States
Posts: 94

First, I would absolutely recommend upgrading to a half bridge of IGBTs - I have seen a few single transistor coils achieve respectable performance, but given that even the simplest half bridge essentially guarantees good results (take Steve's SSTC-5, for example), it seems like a rather obvious choice to me.

Second, how large is the IGBT heatsink? If you want to avoid overheating, the simplest answer is usually involves a larger piece of aluminum.

Third, are you using an interrupter? If not, then you need one. Running your coil in CW will almost certainly lead to fast IGBT failure. On the other hand, if you are using one, then try testing your coil with the pulse rate/pulse width set lower to decrease the stress on the transistor.

Fourth, you need to upgrade your primary with significantly thicker wire, especially if you want to run it at full power.

Fifth, are you sure that your gate driver is giving your IGBT a clean square wave at the proper gate voltage? Excessive ringing can destroy the IGBT's gate, or, if the voltage of the gate drive signal is too low, then the IGBT will not switch on all the way and this can also lead to failure. Try testing the signal on the IGBT's gate with an oscilloscope to make sure this isn't the case.

Finally, try "soft starting" the circuit with a variac - I blew several MOSFETs in my first SSTC by applying full power to it without a variac. Later, you can add a transistor circuit that will soft start without a variac once you are satisfied with the performance. Steve Ward's SSTC-1 uses this type of circuit.

Good luck!

Dr. ISOTOP

Mon Jun 13 2011, 05:02AM

Registered Member #2919 Joined: Fri Jun 11 2010, 06:30PM
Location: Cambridge, MA
Posts: 652

What frequency are you running at? Slow IGBTs (e.g. the Fairchild B3D series) don't like to run over 100 khz, fast ones (e.g. the A4D series) don't like to be pushed over 200.

Shrad

Mon Jun 13 2011, 09:05AM

Registered Member #3215 Joined: Sun Sept 19 2010, 08:42PM
Location:
Posts: 780

the frequency is 500Khz-ish, and that IGBT is in the 20ns on-time, 120ns off-time range

I'll try putting an IRF450 there and give it a try before trying to investigate further

switching waveform is nice, there's not much ringing and the driver pair is just warm to the touch after 10 minutes

heatsink is a nice one (athlon xp) and blown by a small fan

I'll get some pics of the IGBT waveforms at 30VDC in and decide what to do, and will use an interrupter btw

thanks guys! I'll post pictures as soon as I have nice streamers

Shrad

Mon Jun 13 2011, 01:11PM

Registered Member #3215 Joined: Sun Sept 19 2010, 08:42PM
Location:
Posts: 780

hmm lets enumerate my few problems

I switched to an IRF840 half bridge with PP caps of 1.5Âµ and fast diodes

I also rewound my GDT to feed the two mosfets but I have much ringing at my GDT output and no voltage (perhaps 2 volts)

Previous winding was excellent and waveform was perfect with the IGBT

The GDT is 12 turns 1mmÂ² primary, 2x12 turns bifilar 0.8mmÂ² secondary on a grey core, 25mm outer diameter

I have tried paralelling gate drivers with no more results

the GDT is powered through a 10 ohm + 0.47Âµ

Maybe someone could point me to the good direction here? I'm losing patience because I still have to learn in that domain ^^ and I dont understand where the problem could be

Daniel Kramnik

Mon Jun 13 2011, 01:46PM

Registered Member #3885 Joined: Sun May 15 2011, 12:47AM
Location: Newton, Massachusetts, United States
Posts: 94

Are you using bypass capacitors across all of your ICs? Sometimes, insufficient bypass capacitance can cause the voltage to drop on the MOSFET side of the GDT. Then again, you don't seem to be pushing too much power with your coil, so it might not be a problem.

I think you might want to rewind the GDT again - try using a core that you can be absolutely sure is made of ferrite (also try to make sure that AL >= 4000) and made one trifilar winding instead of a bifilar secondary and separate primary - this will greatly reduce leakage inductance and probably solve your ringing problem and help your voltage problem.

Can you post pictures of your driver and bridge, taking a look at the physical setup may help.

Once again, good luck with your project!

Shrad

Mon Jun 13 2011, 03:06PM

Registered Member #3215 Joined: Sun Sept 19 2010, 08:42PM
Location:
Posts: 780

Thanks Daniel,

I'll have a rewinding session ASAP

The coil is for sure ferrite, it was used in a feedback transformer from a big SMPS (grey ferrite toroid)

I'll have a try with another one, blue this time, which already has a trifilar winding

Shrad

Mon Jun 13 2011, 04:02PM

Registered Member #3215 Joined: Sun Sept 19 2010, 08:42PM
Location:
Posts: 780

here are the shots

I replaced the core with another one (the one switching my igbt easily) and wound 30 turns instead of 12 turns as the signal was spikey

now the waveform is pristine :

Shrad

Mon Jun 13 2011, 05:14PM

Registered Member #3215 Joined: Sun Sept 19 2010, 08:42PM
Location:
Posts: 780

still nothing with the half bridge

when powering it with 30VDC I have no light, even if it is in tune (I have 40Vpp across the coil)

when powering it with more than 100VDC the schottkys keep shorting (they are 400V 16A double schottkys with paralelled diode)

I'll wait and see if I can make a better bridge or try with a single IRF450

I have no clue why it keeps behaving badly, so if anyone has an idea I'll take

(I had breakout with 30VDC and a single IGBT switching the coil to ground)

Dr. ISOTOP

Mon Jun 13 2011, 08:10PM

Registered Member #2919 Joined: Fri Jun 11 2010, 06:30PM
Location: Cambridge, MA
Posts: 652

Are you getting anything out of your bridge?

1 2 next

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