USB Powered Tesla Coil - Design Study / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

USB Powered Tesla Coil - Design Study

1 2 next

Move Thread

LAN_403

thedatastream

Thu Nov 30 2006, 08:40AM

Registered Member #505 Joined: Sun Nov 19 2006, 06:42PM
Location: Yorkshire!
Posts: 329

It makes sense... you know it does!

The aim here is to flesh out the basic design of such a device. Thoughts, comments and constructive criticism welcome.

Criteria:

It is small (lets say no more than 6" high (including drive electronics) and a 4" by 4" footprint. However, the smaller the better!
It should produce a spark that is visible of at least 1" in length and appears visibly continuous
Should ideally only use one USB port
It must not fry the computer under any circumstances!

Problems:

Assuming you have a 500mA capable USB port, this means that we can at best get 5V @ 500mA = 2.5W - not enough to run a decent sized coil (I assume). So charging a capacitor from a constant current regulator and then periodically discharging it. Problem also if only lower current ports are available for whatever reason.
Heavy filtering required on the USB power and data lines to prevent the computer from being affected. Good shielding around the USB cable is also a must!
Running in close proximity to the computer could affect it - perhaps a Faraday cage of some kind around the coil?

Initial questions:

What topology of coil would be most appropriate - SSTC, DRSSTC, OLTC? I would imagine that a SGTC would be out of the question for such low powers.
Run primary from 5V or step up to a higher voltage? Resistive losses get less with a higher voltage, but charging of the necessary large storage capacitor would get longer - affecting spark output?

Rgds,
James

Steve Conner

Thu Nov 30 2006, 10:16AM

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

I think some variation on this theme

would be the way to go. One of our members JimG built a tiny DRSSTC that ran off a 12V wall wart, you could scale his design down even further to 5V and connect it to the above mentioned 30 USB ports

The first thing he did, IIRC, was to step the 12V up to 350V and use it to charge a big electrolytic cap, that allowed a 1 second burst every few seconds. Derek Woodroffe's various Tiny/Nano/Eny-WeenyTeslas could be a good source of ideas too.

I think due to the high frequencies involved, at least a few MHz, an OLTC is probably out of the question. SGTCs are probably out too on the grounds of the insane levels of EMI they put out.

Once you get it working, you should make it serve a Java applet so it can be fired remotely over the net and e-mail you spark pix of itself

P.S. For making USB stuff, the first thing I do is use the FT232 chip from Future Technologies to convert it to RS232. If you're embedding the chip in a product, they give you your own PID and all the stuff you need to make a custom driver, so it can actually enumerate as "USB Tesla Coil" if you like

thedatastream

Thu Nov 30 2006, 11:20AM

Registered Member #505 Joined: Sun Nov 19 2006, 06:42PM
Location: Yorkshire!
Posts: 329

Steve Conner wrote ...
One of our members JimG built a tiny DRSSTC that ran off a 12V wall wart, you could scale his design down even further to 5V and connect it to the above mentioned 30 USB ports

The first thing he did, IIRC, was to step the 12V up to 350V and use it to charge a big electrolytic cap, that allowed a 1 second burst every few seconds.

Is this the project that you are referring to?

Steve Conner wrote ...
Derek Woodroffe's various Tiny/Nano/Eny-WeenyTeslas could be a good source of ideas too.

I'd forgotten about Derek's stuff, certainly his Nano Tess coil looks very promising.

Steve Conner wrote ...
...an OLTC is probably out of the question. SGTCs are probably out too

Yes, solid state would definitely be the way forward. I guess I need to look at boost converters and a big electrolytic as well as working out my power budgets.

Adding a micro and a converter IC to do enumeration and control would be super cool, definitely something to sort out for the future!

Thanks for the tips
James

Marko

Thu Nov 30 2006, 02:03PM

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

Considering your obivous programming skills, software control of that coil would be great idea!

Maybe with an external soundcard inside, audio-mod it too?

/my imagination runs away

For '1inch spark wich appears continuous' small SSTC would be the best idea - DRSSTC would require much more power unless you are statisfied with single bangs separated with few seconds of off-time.

SSTC is also going to be a challenge since you have very high operating frequency, and pretty high losses in secondary with that.

For design of the coil itself, I can imagine it might look like my recent class E coil from here:

That coil needed about 100Watts to produce 10 centimeter spark, with interrupter.
So 1 inch from 2,5Watts is going to be quite a challenge!

That's why I think you will desperately need lower resonant frequency, and I fear you might end with some kind multi-layer secondary, dug out from old DC flyback or something.

If you end in Megahertz range I fear you might get in trouble with so small power input!

Another try would be to copy derek woodrofe and use IXDD gate drivers on the coil, directly.

This would enable you to use higher frequencies, and maybe even direct 5V on primary; but 1 inch is still a lot.

Your choice now

thedatastream

Thu Nov 30 2006, 08:02PM

Registered Member #505 Joined: Sun Nov 19 2006, 06:42PM
Location: Yorkshire!
Posts: 329

It all depends on the lowest possible interrupter repetition rate that gives an apparently continous spark - I don't know if anyone has studied this yet. Low enough and the cap charging rate from a single USB port could almost result ia "continuously" visible spark.

Also, it is a fine balance between a small cap size for rapid charging and a large cap size for longer operation. There may well not be a balance given the input power.

The alternative is to charge up a big capacitor and only have short bursts of operation, that would be acceptable considering this is mainly a feasbility study.

Either approach will require a very efficient coil. What steps could be used to achieve this? Tighter primary coupling? Lower frequency secondary for reduced switching losses in the primary (minimal effect)? Answers on a postcard...

If we used a constant current source to charge the cap and triggered the coil operation from the voltage on the cap, additional USB ports could be connected and used to charge the cap quicker and get a better output. This wouldn't be against the rules, so long as it works on at least one port some of the time.

Software control (short operation vs continuous bursts, audio modulation) would be super cool

James

Marko

Thu Nov 30 2006, 09:35PM

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

You, lower secondary frequency is a good idea.
It would be ideal to keep Fres under 500kHz but that would require pretty big secondary wound with ridiculously thin wire, or some kind of multilayer secondary.

In order for streamer to be 'visibly continuous' you are going to need about some 25bps, and over 100 bps is needed for it to really become 'continuous' (air remains ionised beetween bursts); but even coming to 25 with so little power will be a challenge.

If I were to do it, I would build such a coil this way:

> find a way to lower resonant frequency. That would include using multilayer secondaries, using ferrite cores, maybe using ferrite transformer-driver (it would be a small magnifier in that case).

> I would use some gate driver chips, UCC's or XDD414's in a fullbridge directly - driving mosfets or IGBT's would just be a waste of power. (derek's nano tess uses such topology).
Keeping design simple also helps debugging the coil later.

> boost converter could be used to max the input voltage and minimize switching losses; but it must not be too lossy itself.

> coil is very unlikely to crash or fry your PC, since it's usually well shielded and has ground conductors all over it.
But, it is very likely to inject noise and ultimately crash your DSL or whatever internet connection, wich I expeirenced myself alot.

From external electric field, shielding is fairly easy; simple ''cage'' consisting of four grounded metal wires vertically mounted on coil's edges will stop it well.

Couplint through USB cable itself is major problem, and no filters or anything will stop it unless there is good RF ground nearby.

Hopefully, coil will be weak enough to spare you of such problems.

JimG

Sun Dec 03 2006, 03:07AM

Registered Member #52 Joined: Thu Feb 09 2006, 04:22AM
Location: Austin TX
Posts: 57

A USB powered powered tesla coil sounds interesting. It would a cool feat if it was also USB controlled.

When I started working on my tiny tesla coils I read a lot about Derek's coils. I really liked the simplicity of the Nano Tess, but I wanted to generate long streamers, not just corona. I had built a number of DRSSTCs so that was the reason I decided to use a boost converter and a full bridge DRSSTC.

Something that I've always wondered about is if it was possible to run a DRSSTC with a low input voltage and then just use longer on times to build up the current. I know I've had larger DRSSTCs break out at 30 VDC, though I don't know how much losses would be affected by a lower voltage.

I also got some disposable cameras a short time ago to see if it was possible to use the flash circuit to generate a high enough voltage to power a tesla coil using the neon charge lamp as a spark gap. I have yet to find out if the neon lamp would have a fast enough turn on time.

I haven't found much interference from my tiny tesla coils, but if your getting too much interference you may be able to fix most of it by filtering the power supply.

...

Sun Dec 03 2006, 04:30AM

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

This definitely sounds fun

As to power, good luck getting a continuous spark 1" long off of 2.5w. While that gives a 0.66in/w^.5, I have found small coils to be much less efficient that their larger counterparts (either that or I just suck at building coils...)

I am going to say your best bet would be to have the 5v go through a bigass inductor (to keep hf nasties out of the computer) then into farad or 2 worth of super caps for energy storage. Then off to a boost converter (or possibly a half bridge or my favorite a Mazzilli converter) and into a the HV storage cap. Then go for a normal drsstc, winding the coil with the smallest wire possible to get the lowest possible Fres.

--OR---
You COULD try making the hv source regulatable, then build a sisg. If you use a tl494 based fulbridge, set for constant current operation (set by a digital pot or something) into a single SISG stage you should be able to get something usefulll... Just choose a tank cap big enough to give the disired spark and then set the bps by the current you give it. Certinly simpler than a drsstc

I might have to try that sometime.

Then you can set the on time to give 1" worth of spark and then see how high you can get the BPS. Then as you tune the coil you can keep turning up the bps until the usb port shuts down.

You could also add either a small LiPo battery in the bottom that could charge off the usb port and then be used for 'high power' operation... Or better yet have an external 5v in.

Also, I wouldn't plan on selling them to the public... Just doesn't seem like a good idea to me, regardless of how many safety warnings you put on it... Ebay maybe...

flannelhead

Sat Feb 23 2008, 11:21AM

Registered Member #952 Joined: Mon Aug 13 2007, 11:07AM
Location: Finland
Posts: 388

Just ran across this old thread and wanted to drop a pair of my own ideas:
You could power a disposable camera transformer from the usb port and charge some of those photoflash caps with it. You might want to regulate the USB 5V to somewhere near 2Volts or something. The photoflash caps could supply power to the bridge of a SSTC or DRSSTC in short bursts (propably 1 sec) and then the caps charge again... (hopefully you understood what I'm trying to say). This is much like JimG's DRSSTC idea. The disposable camera transformers don't use much power, so the 2.5W from the USB port could be enough...
I've also got good experences with those transformers in an SGTC, so I think an USB tesla coil IS possible.

Gotta try it some day!

Alexandre

Sat Feb 23 2008, 12:47PM

Registered Member #1228 Joined: Tue Jan 15 2008, 01:08AM
Location: Curitiba, Brazil
Posts: 162

Interesant...

1 2 next

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