My single switch with feedback - 70cm sparks! / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

My single switch with feedback - 70cm sparks!

previous 1 2

Move Thread

LAN_403

Mates

Thu Nov 12 2009, 10:02PM

Registered Member #1025 Joined: Sun Sept 23 2007, 07:53PM
Location: Czech Rep.
Posts: 566

Ok, first of all thank you for the positive feedback. It is always nice to see somebody likes my workâ€¦

As I promised I prepared a complete documentation to the coil. I made it to initiate some scientific discussion about the design and mainly to inspire potential followers. My coil has few advantages over classical bridge designs which are also mentioned.

The driver is classical Steve Wardâ€™s design using the current transformer, hex inverter and pair of UCC27322(1). Enable pin of the gate drivers is connected to PWM based on TL494. It means nothing unusual.

The power part employs only 3 semiconductors: The valve (transistor), diode and TVS. Apart from semiconductors three capacitors are employed. First cap is across the mains and serves for lowering the impedance of the mains. It is made of six 3.3uF/630V foil caps. Second cap is resonant and needs to be set according to the primary. It is based on MMC (foil caps 5KV/66nF). Third capacitor makes an important trick. It serves as current limiter and is made of relatively big foil caps in parallel - total value 310uF/400V AC. Such caps are originally used in asynchronous motors (smaller one you can find in every washing machine). They are not very expensive. These caps are also directly involved in regulation of the total power consumption of the coil and by paralleling them you can easily manipulate the coil power (length of the sparks).

The valve:
I use BUP314 transistors three in parallel. Iâ€™ve never tried any other type, but generally the transistors needs to be HV rated (minimum 1000V) and rather robust. On the other hand they donâ€™t need to be super fast. Mainly the turn-off time is not an issue. Thatâ€™s because of the limiter cap which ensures that maximum amps are sucked from the mains within first few hundred nanoseconds. I could read some opinions on this forum, that paralleling IGBTâ€™s makes no sense but I donâ€™t agree. My personal experience tells me the opposite. I successfully tested these transistors in this setup up to 300 KHz, but 166 KHz (the operation freq of the coil at the moment) is much safer.

The principle of operation is still a bit mystery. I claim the system works in Class E, but there are many here on the forum who disagree. Iâ€™m not an expert so I keep this issue open. The fact is that the resonant cap has a big impact on coil performance and thus making the coil different from regular SSTC. Once the valve is opened the big cap is charged over the coil primary. During the negative AC half-wave the cap is discharged over the diode. It means that this coil would never work on DC! The TVS battery is necessary part of the design. It protects the transistors from HV peaks which can be easily induced on the primary. Once the coil is tuned properly they do not heat up at all. On the other hand in case of ground strikes and improper tuning these little semiconductor miracles keep the transistors alive!

Advantages of the design:
The coil uses only one valve. It means that despite multiple transistors are employed (in parallel) you never destroy more than one. Thatâ€™s a big advantage over the bridge setup which always eats transistors in pairs.
Next advantage is that you can push the transistors far behind their recommended values.

Future plans:
To redesign the gate driver to employ two more UCC to get more driving amps
To rewind the GDT to get higher gates voltage
To use more than three transistors in parallel
To increase the power of the coil to reach 1m long streamers (letâ€™s see)

Cheers Mates

GeordieBoy

Thu Nov 12 2009, 10:42PM

Registered Member #1232 Joined: Wed Jan 16 2008, 10:53PM
Location: Doon tha Toon!
Posts: 881

Very impressive.

> I claim the system works in Class E, but there are many here on the forum who disagree.

I'd say that it must work at least close to Class-E. If there was any significant voltage across the IGBTs at turn-on the energy stored in your 6x100nF capacitors connected across it would burn it out quickly.

If you have an isolation transformer and a scope and high-voltage probes you could always post some waveforms and start a discussion on it's method of operation!

-Richie,

Mates

Fri Nov 20 2009, 08:40PM

Registered Member #1025 Joined: Sun Sept 23 2007, 07:53PM
Location: Czech Rep.
Posts: 566

Hi guys,
Nothing so special this timeâ€¦

I finished new coil which is planned for my next VTTC project. It runs on 230Khz and I tried my single switch driver for basic testing of this little fat boy. I was a bit afraid about the frequency, because BUP314 is not very fast. The coil runs without any problems and transistors are cold after 15-20s runs which proves the single switch design as relatively robust. There is one special feature about the new coil. It is winded on a tube made of ceramics so it is pretty heavy stuff. It was originally huge 16A variable resistor which I had no use for

Cheers Mates

And here one more pic of the Red Traffic Cone Coil

Tesladownunder

Wed Nov 25 2009, 04:55PM

Registered Member #10 Joined: Thu Feb 02 2006, 09:45AM
Location: Bunbury, Australia
Posts: 1424

How about adding a brief explanationt with 2 pictures to the ** Add your Tesla Coil here ** thread?

vasil

Wed Nov 25 2009, 07:27PM

Registered Member #229 Joined: Tue Feb 21 2006, 07:33PM
Location: Romania
Posts: 506

Awsome!
This single switch concept is very attractive...
Do the iGBTs warm when you draw arcs to ground?

Mates

Thu Nov 26 2009, 10:33AM

Registered Member #1025 Joined: Sun Sept 23 2007, 07:53PM
Location: Czech Rep.
Posts: 566

Tesladownunder wrote ...

How about adding a brief explanationt with 2 pictures to the ** Add your Tesla Coil here ** thread?

The Red Traffic Cone coil is already there and this small coil is planned to be powered by vacuum tube GU81. Once it is finished I will add it there.

vasil wrote ...

Awsome!
This single switch concept is very attractive...
Do the iGBTs warm when you draw arcs to ground?

Actually I do not have good experience with ground arcs at full power (low voltage low power is not a problem at all). It works also at full power (2000W), but sometimes it also kills the transistor. The problem as I see it is that the ground arcs are such a big disturbance for the resonant frequency that it shifts it out of the class E operation completely. I think it could be solved by some over-current protection (like the one in the mini brute design). There is still a lot of work on that design to make it completely bulletproof.

previous 1 2

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