Primary Coil Driver ('Pinger') / Projects / Forums

Forums

4hv.org :: Forums :: Projects

« Previous topic | Next topic »

Primary Coil Driver ('Pinger')

1 2 next

Move Thread

LAN_403

Hazmatt_(The Underdog)

Thu Aug 02 2007, 01:35AM

Registered Member #135 Joined: Sat Feb 11 2006, 12:06AM
Location: Anywhere is fine
Posts: 1735

I've been in need of some kind of pulsed driver for Tesla coil testing for a long time now, so I decided to make something that would 'excite' the primary circuit under pulsed conditions. It's a simple setup, using a hacked ATX PSU, a 555 driver and a 'scope.

The purpose of the driver is to 'ping' the primary tank so I can see how its coupling and ringing under low voltage conditions. I want to do more testing of the secondary at low voltages so I can see it on the scope without making a lot of noise. I am hopeful that I can find not only the optimal coupling, but to see how the Q and wire gauge relationships are involved here as well.

Sulaiman

Sun Aug 05 2007, 12:02PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3141

I tend to put mosfet/igbt source/emitter to 0V so that the gate drive is predictable,

is there any special reason why you did it as in your circuit diagram?

Antonio

Sun Aug 05 2007, 11:31PM

Registered Member #834 Joined: Tue Jun 12 2007, 10:57PM
Location: Brazil
Posts: 644

To see the correct waveforms, similar to what appears in a spark gap coil, you need very low output impedance in this kind of driver. I use something like this:

mod edit - oversized images - click to see original
It's similar to one of your schematics, but without a series resistor. This is enough to test my coils, that use quite low power. A high-power coil would require a more powerful buffer.
Example of observed waveform over the primary coil:
Tefpreg1

Hazmatt_(The Underdog)

Mon Aug 06 2007, 06:07AM

Registered Member #135 Joined: Sat Feb 11 2006, 12:06AM
Location: Anywhere is fine
Posts: 1735

My driver was just a quickie based on a high current switch I was designing for another application. It didn't really matter where the mosfet was just as long as I got a ringdown. I may have to go in and switch it to get rid of some nastyness I noticed on the initial pulse. I have more time to 'clean it up a bit' now. I was trying to get it done in a hurry so I could test, but other things came up so I may fix that tonight.

Antonio's driver is only correct if you plan on Continuous Wave excitation, which is not the case by the way. You can't get accurate measurements of the SGTC performance in CW operation because its shock excited with a pulse, which is what makes the measurements more difficult.
He is demonstrating a series resonant circuit as well, and the system I am working with is parallel resonant. His LC network is acting as a voltage divider, whereas my parallel resonant system acts as two parallel impedances that 'short' the supply, hence the 12A ATX and 50W of ballast resistance.

Steve Conner

Mon Aug 06 2007, 01:11PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

Antonio's pinger circuit is the one that I've used myself, and it worked perfectly for me. It does provide shock excitation, not CW. The scope screen shot Antonio posted is the step response of the system, exactly as if you had fired a spark gap.

To give it more wallop, I used the bootstrap IGBT gate driver that I developed in place of the complementary pair of transistors. I imagine a gate drive chip like the UCC37322 would work fine too.

Hazmatt: I guess if you're testing a parallel resonant circuit, you need to ping it with a current source instead of a voltage source.

Antonio

Mon Aug 06 2007, 01:35PM

Registered Member #834 Joined: Tue Jun 12 2007, 10:57PM
Location: Brazil
Posts: 644

Exactly. My pinger oscillates at 500 Hz, while the coil resonates at 300 kHz. A voltage step behind the discharged C1 produces the same transient that a charged C1 connected to L1 would produce. Any resistance in series adds damping.
Something similar can be produced with a regular 50 Ohm square wave signal generator, with a 1 Ohm resistor across its output to lower its output impedance. The problem is that the output voltage becomes quite small and noisy.

Marko

Mon Aug 06 2007, 01:38PM

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

I tought if it would be cool to build a simple self-resonant 'driver' and test the secondary just by throwing the circuit over it. If somebody is cool enough he could stuff in a mcu-based frequency counter to display resonant frequency in situ

Still, what it general benefit of having it pulsed instead of CW? You fear primary circuit will detune the secondary a bit? I used such simple 'ping' circuits all until now, they work more than well enough.

Hazmatt_(The Underdog)

Mon Aug 06 2007, 06:13PM

Registered Member #135 Joined: Sat Feb 11 2006, 12:06AM
Location: Anywhere is fine
Posts: 1735

You're still driving your circuit with a 50% duty cycle square wave though from the 555, whereas mine is adjustable from a very short pulse to about 50%.

If I change my timing capacitor from 0.47uF to 33nF I bet I could get the same ringdown that you're showing.

I was planning on building a 'Terry Tuner' with a wide range square wave, frequency counter, output driver, current meter, and internal L and C as a super tuner but haven't gotten to it yet. The current meter will deflect to the best tuning, and then you can sweep it to 70% on either side to get the Q (if you desire), and the frequency at the same time, then you can also use the internal L and C to check other components such as unknown inductors and bottle caps. So that is coming.

*****************
I just fixed the mosfet wiring and it cleaned up the garbage I had earlier on the waveform prior to the ring-up. The Vgs difference was keeping the mos on, of course, and its just a poor design decision. It's one of those things that happens when you're in a hurry, or at least for me anyway. When I was prototyping I had separate supplies for the generator and the high current side, so I didn't have common source problems until I changed it to a single supply.

So if you're going to use the driver make sure you swap the mosfet around for common source, and just put the ballast on the V+ side.

Here's the cleaned up step response:
The step is used as the sync. pulse for external triggering.
006f

And I tried Antonio's setup with my function generator as the output stage and connected/disconnected the RC circuit to see if it was going to change the response, but all it did was attenuate the signal somewhat as a parallel impedance.
The output though was basically the same as my driver, no pretty beating oscillations like the picture is showing, just an LC ringdown.
Driving at 500Hz, LC resonant at 10KHz, Picture showing V(L1)

007f

Antonio

Mon Aug 06 2007, 10:20PM

Registered Member #834 Joined: Tue Jun 12 2007, 10:57PM
Location: Brazil
Posts: 644

10 kHz? Is this the primary of a Tesla coil? The beats appear only when the secondary coil is added, and both resonators are at the same frequency (when observed independently). The damping is substantially reduced (in a normal Tesla coil primary circuit, with low series resistance) if a 1 Ohm resistor (or lower) is connected across the output of the signal generator.

Hazmatt_(The Underdog)

Mon Aug 06 2007, 10:51PM

Registered Member #135 Joined: Sat Feb 11 2006, 12:06AM
Location: Anywhere is fine
Posts: 1735

Right now I'm just testing on a coil coupling setup that I made for coupling evaluation, so its set to be resonant at a low frequency, one the little audio power amp can amplify at.
The thicker 2 section coil I am using ( one section is in use) as the primary driver and the longer coil I am using as the secondary, both resonant at the same frequency ~9560 Hz or so.

I'm still trying to recreate your test results with no success. You're suggesting that the output impedance of the generator be lowered by using a 1 ohm resistor, so I may try lowering the output impedance.

Something else comes to mind though, when you were testing your system with both primary and secondary, was the secondary loaded at all or just open circuit?

1 2 next

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