The Cherry Cokinator / Projects / Forums

Forums

4hv.org :: Forums :: Projects

« Previous topic | Next topic »

The Cherry Cokinator

Move Thread

LAN_403

...

Fri Feb 08 2013, 04:01PM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

A clone of an older coil of mine,

, built using junk laying around miters

Every design decision was made by using the most convenient piece of abandoned gear I could find, which resulted in a very odd looking coil, which draws tons of power and bursts into flames every once in a while.

General specs-
*160v buss voltage, 1600uf filter cap
*Currently running from a function generator (audio fed strait into the VCF input, no conditioning whatsoever except for a 1k pulldown resistor to keep the noise down), but I have already breaboarded a driver based on the CD4046, just using it as a VCO (pll components unused), inspiration from Steve Connor's Tesla 5 which uses a bluetooth headset receiver (currently using one designed for phone calls, but due to poor availability of programs to send audio to a mono headset, I have ordered a A2DP compatible one) as the audio source.
*Pair of microchip TC4451 gate drivers, and a some simple logic to allow interrupted mode (no enable input on the TC442x or TC445x parts), feeding a 4 turn times 7 wires (one for each gate, 3 in parallel to the drivers, since there was only 3 holes free on the breadboard

) GDT. No dampening resistors, the 0.5ohm output impedance of the mosfet drivers seems to be plenty to give a nice clean waveform.
*Using some massive minibricks for the bridge, one of the halves is an APT6015JN 600v/38A, the other half is a similar brick that I could not find any information on, labeled apt546-227-6238 So far no problems with the mismatched bricks...
*6 turn primary out of 14awg wire, which gets pretty warm
*Secondary was originally designed for DRSSTC duty (primary/secondary were originally from the first post of

), wound with 36awg wire which was smoking hot after the 4 minute run in the video
*Soda can topload

Video of audio modulation-

Videos with a real topload and a lower frequency secondary
Running with no audio (didn't want to risk damaging the nice secondary with the stresses from audio modulation), pulling about 2kW

And with the large DC filter cap removed for halfwave operation (power consumption down to about 1300w)

Interestingly enough, when the youtube DMCA violation scripts kicked in, it identified the song as a cover

by 15 year old Canadian girl o_0

Clearly I have some more work to do on the sound quality, because a tesla coil which makes audio sounds like a 15 year old girl won't do!

Steve Conner

Fri Feb 08 2013, 04:25PM

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

Wooo... I've never seen Bluetooth audio done before. 8)

Marko

Fri Feb 08 2013, 08:57PM

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

Hey Peter, nice work! I've actually been contemplating the possibility of controlling a TC over bluetooth for a while now for my future smart-TC prospects - glad to see someone having tested it! With the latest spam of portable android based devices I thought it'd be really nice is they could be used as great Tesla coil interrupters, controllers and debuggers... However, I thought to use some simple serial bluetooth modules so I can transfer commands and data in both directions, along with keep-alive signals so the coil turns off if connection is interrupted.

Despite the commonly accepted boo-boo's with radio communication around Tesla coils, I think this one has a good chance of actually working well, considering bluetooth is a really robust spread spectrum protocol.

Just some other thoughts on your setup...

1. Open loopn VCO drive??? I thought you guys at MIT already went for virtex 5's or something :)

2. I seem to hear your fans slow down when the coil is energized. I had a similar problem which can actually be heard here:

At first I was convinced it was due to the fact that I used live heatsinks, and alternating electric field was screwing up the fans. However, a fan of same type ran normally on a heatsink outside the coil when I connected the heatsink to one of bridge heatsinks by an alligator clip...

Then I believed it had to be the power supply voltage sagging, but the fans did the same even when powered directly from a lead acid battery, and attempts to shield the fans by grounded foil also failed completely!

I had to conclude it was the magnetic field from TC primary screwing the hall sensors inside the fans, despite they were mounted under thick chunks of aluminum. It was great frustration in overall - later I replaced the fans with AC ones which proved rather pathetic at moving air. From now on I think things through well before putting a DC fan into a Tesla coil!

I wonder if your problem is the same, or perhaps it's just your lab power supply going mad? It'd be interesting to see what happens if you moved the bridge assembly away from the coil.
e
3. It seems that experiments established that TC plasma turns quiet and flame-like at around 4-5Mhz, so if you want really high fidelity audio modulation, wouldn't it be a better idea to pursue a HFSSTC instead... of course soup it up to some serious power level! Perhaps with a phase-shift class DE or double-ended class E driver which Richie Burnett had good success with.

4. To combat frequency distortion you could record a frequency characteristic of the coil using a microphone and some of the badass scopes you must have around there, and then use it to design an input filter to compensate for it.

Cheers,

Marko

...

Fri Feb 08 2013, 10:22PM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

The fan does in fact slow down under operation, I 'solved' the issue by switching to tandem 10k rpm fans

, with no voltage into the bridge they sound like jet engines, but after the coil gets going they slow down to a reasonable speed and everything is peachy. I may end up mounting a thermistor on the bridge and using a msp430 to put the fans in feedback (they are 4 wire fans, so there is a pwm in and tachometer out for each one), and adjust the fan speed to keep up with the heatsink temperature. I had considered switching to a normal AC powered fan to avoid the issue altogether, but the only fan I could find laying around was a 220v one and didn't have enough airflow to compensate for the terribly undersized heatsink the bridge is made on (originally out of a 1u server, the fins are about 1/2" tall, and it is only 60mm square, not designed for cooling a bridge handling 2kw...). With the current arrangement the secondary starts smoking before the bridge so I declared it a success

Per the VCO, I switched to open loop VCO a while ago, and haven't had much of a desire to change it (total of 3 coils built using this topology now, the one in the linked thread, one I built with the UCSB electronics club

and this coil). I started our with the PLL running, which worked fine, but I ended up mounting the phase adjust pot on the outside of each coil so that I could change the 'bias' level of the sparks (detuning at no input, the best audio is had running above the resonant frequency, about half way between large sparks and small sparks) for best audio or best sparks, or somewhere in between. Then I realized that when running with heavy streamer loading the Q of the resonator was so low that the frequency really did not depend on the surroundings, for example in the linked video the frequency is being tuned from ~330-360kHz) so I figured I may as well run open loop. The one thing I do is make sure that the tuning range presented by the external knob is small enough that the coil does not explode if someone turns it too far, same for the audio input.

I don't think we have any FPGA based coils, although there is someone working on a QCW coil based around an mbed, and the oneTesla which has an atmega328 for the interrupter (but the normal self resonant primary feedback for the actual controller).

One of these days I may build a nice high frequency coil, I have found that a nice big CW coil running at 500KHz is 'quiet enough' that as long as you have a decent amount of modulation depth you don't really care about the hiss since it is so loud! This one is currently running a bit low since the secondary was designed for DRSSTC duty, so I hope to get a bit better performance out of it when I get around to making a better secondary. It would be neat to add in a preamp that takes into account the frequency response, although I think even cooler would be to make one that takes out the nonlinearities of the modulation so that I can go to higher modulation depths without distortion. There is also a lot that can be done with fancier breakout points...

I was surprised when wifi/bluetooth worked so well around the coil, when I made the first coil (2008) I had a sony Clie' palm pilot which would stream an icecast stream from my laptop, (this was pre-a2dp, even the mono headsets were pretty pricey back then IIRC)

1360361842 56 FT150443 Streaming Streamers

I figured that the bluetooth receiver would be easier (most phones support them out of the box), so I found one on amazon for about $10 which should get here shortly after the blizzard passes.

Marko

Fri Feb 08 2013, 10:31PM

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

Hi Peter

Have you actually determined what causes your fans to slow down? Have you measured their supply voltage, or tried running them from a battery? If they are getting goofed by the primary field the feedback may not work as you expect.

How do you intend to interface the serial bluetooth module to the coil? You'd need some sort of a microcontroller, at least, to make it work with your current design. If you plan keeping the coil audiomodulation only, then perhaps the bluetooth headphone module might be better suited?

Marko

...

Sun Feb 10 2013, 03:46PM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

I am pretty sure the fans are slowing down due to the RF, I know it is not the power supply because it runs the gate drivers just fine (confirmed by looking on the scope). It is possible that it is noise coming in on the power lines (I observed almost a volt of pickup on the gate driver power lines before I added in filtering, although some of it may have been due to the scope probe), but in any case I figure that if I put a 80% or so PWM signal the fan should run 'slower-ish' and be quieter than the max speed (10krpm -> 70dB) and keep the bridge cool.

Per the bluetooth, I may have incorrectly described my new plan, I am switching from a mono (HSP) to stereo (A2DP) receiver, but it will still be used for audio (no control signals of any kind planned). Currently my plan is to try the iWave

which seems to be a perfect fit, it automatically pairs at power up, and according to the reviews it sounds like it remembers several phones that were paired in the past. This behavior is a lot nicer than the normal behavior (press the button to pair, must re-pair any time you change sources, designed to be battery powered and forget the pairing data if the battery is removed) for tesla coil applications. Hopefully it works

...

Fri Feb 15 2013, 07:20AM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

A few updates:
Some steel wire

Running with the new bluetooth module -

valentines day edition

A shot running halfwave rectified mains for that nice streamer propagation

The bluetooth module ended up working perfect, internal construction is very clean and it was trivial to trace down where to apply the 5v and where to tap off the audio. Pairing lived up to the manual, just turn on the coil and it will pair to any phone in range

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