Some thoughts about Plasma Speaker designs / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Some thoughts about Plasma Speaker designs

1 2 next

Move Thread

LAN_403

Xray

Tue Mar 29 2011, 06:04AM

Registered Member #3429 Joined: Sun Nov 21 2010, 02:04AM
Location: Minnesota, USA
Posts: 288

I want to build a plasma speaker, but I don't want to make one that sounds as bad as most of the ones I've heard that are on YouTube. They sound tinny and/or distorted. There are a few fairly good sounding ones, like this one:

but even that one is not good enough for my taste.

The problems:
1. They all lack low frequency reproduction due mostly to the physics of the small size of the arc.
2. The sound quality is poor mainly due to the design of the electronic circuit, but also due to an unstable arc.

During my research, I've found numerous design approaches, and two types of spark devices. The two types of spark devices are:
1. A two-terminal spark gap.
2. A single pointed terminal or sharp wire end that produces corrona into the open air, similar to a sharp point on the top of a Tesla Coil.

There are also various driver circuits for the flyback transformer. Some use a single MOSFET driven by a 555 timer chip. Some use 2 MOSFETS in a half-bridge configuration driven by a 3525 PWM chip. I've even seen a few circuits that use a vacuum tube (aka, "valve") to power the hv transformer. Many of the older Plasma tweeters, such as the IONOVAC, use a vacuum tube to power the unit. And some Plasma speakers use gass such as Argon or Helium, or a mix of gases which is ionized and modulated to produce the sound.

Many of the circuits that use the 3525 PWM chip do not actually use it in PWM mode! The designer simply feeds the audio signal into the pin that determines the operating frequency via a resistor/capacitor combo. In that mode, the audio is actually modulating the FREQUENCY of the drive signal, NOT the pulse width.

So, the bottom line is, I would like to know if there is any design information for plasma speakers on the web? I've searched quite a lot already over the past couple of days, and found numerous schematics and many articles on "how to build a plasma speaker" but nothing on the design philosphy. I was hoping to save myself a lot of wasted time and money if someone knows of any plasma speaker design tutorials or tips on the web. I would like to learn what the best type of audio modulationg scheme is for plasma speakers (for example, is PWM better than FM? Is AM better than PWM?, etc.). Also, are there better arc designs than others? Is there a prefered arc design that produces the best low frequency response, etc?

Any info would be greatly appreciated.

Tightmopedman9

Tue Mar 29 2011, 07:07AM

Registered Member #3197 Joined: Tue Sept 14 2010, 04:56PM
Location:
Posts: 19

I could say a lot on this, but I just make a few comments. The signal produced by FM and PWM modulation is going to be the same. The method of producing said modulation is different so there is going to be a difference in quality. Because a PWM chip is meant to pulse width modulate you can bet that it is going to be better quality than a ghetto FM scheme. AM is out of the question, the audio quality would be quite bad.

You can use any output stage you like, it just depends on what your goals are for voltage and amperage. You'll need a MOSFET gate driver (UCC37322) to power the MOSFET gate since the oscillator chip won't have enough power to do so it self. The MOSFET won't switch fast enough and it'll remain in its liner range for too long, will produce a lot of heat and will operate inefficiently. A gate drive transformer would be a good idea to make sure no back EMI gets into the oscillator/driver stage. Also, remember to protect the MOSFET gates with zeners and the drain/source with fast diodes.

A single MOSFET output would be the simplest, however power output will be relatively low. A dual MOSFET setup would be preferable and will allow you run the a transformer in non-flyback mode. This will allow you to take out the air gap and hopefully minimize leakage inductance. You'll need a snubber (quasi resonant or RCD are the simplest) to absorb voltage spikes and the leakage inductance. For dual MOSFET setups you could go with a half-bridge or a two switch forward topology.

Low frequency is going to be hard, you'll have to increase the length/width of the arc which means more power. To do this you'll need more powerful MOSFETs and a bigger transformer (read: self made). Due to the nature of the question I think such high power might be a little out of your range (no offense).

Point to hemispherical head is the best spark breakout I've seen; it wanders very little.

Xray

Tue Mar 29 2011, 03:38PM

Registered Member #3429 Joined: Sun Nov 21 2010, 02:04AM
Location: Minnesota, USA
Posts: 288

Tightmopedman -- Hey, thanks for the great info! I have a few brand new IRFP460 MOSFETS which I plan on using in half-bridge mode driving a high frequency/high voltage transformer from an X-ray tube head. It can handle a lot of current, and I plan to keep it cool in oil.

You stated that AM is out of the question, but that's the method that some of the commercially built plasma tweeters use. The vacuum tube (valve) operates in self-oscillaing mode in the MHz range, and the audio is fed into one of its grids to amplitude-modulate the output. So, I don't think that AM is out of the question, it just needs to be implemented properly.

I do like your suggestion about the point to hemisherical head spark-gap design. Thanks again!

Dr. ISOTOP

Tue Mar 29 2011, 03:44PM

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

Class E SSTC's make great plasma speakers, I hear. The high arc frequency eliminates any residual noise, and the coil streamer is very stable.

Tightmopedman9

Tue Mar 29 2011, 08:19PM

Registered Member #3197 Joined: Tue Sept 14 2010, 04:56PM
Location:
Posts: 19

On amplitude modulation, I was really just going off the stated differences between FM and AM radio, i know that AM has a lot less potential in that arena.
It would be hard to implement AM in the normal MOSFET driving scheme since a change in amplitude at the MOSFET gate won't do anything to the output signal, unless you over voltage the gate and blow it up.

ScotchTapeLord

Tue Mar 29 2011, 08:34PM

Registered Member #1875 Joined: Sun Dec 21 2008, 06:36PM
Location:
Posts: 635

I wouldn't say FM has more potential than AM for a plasma speaker. It would require modulation of the (very smooth) supply voltage and it would be your highest quality option. Class A would be perfect for the job. Then you just need to build a HFHV transformer (can be ferrite or air cored) and feed it the modulated supply.

Steve Conner

Tue Mar 29 2011, 08:51PM

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

The FM ones work by "slope detection". The output voltage from the resonator depends on how close the drive frequency is to the resonant frequency, so the resonator converts the FM back to AM. You can receive FM on an AM radio in the same way, using the slope of the IF filters, hence the name.

For maximum undistorted audio output, the idle frequency is halfway between resonance, and the spark going out. If you tune it exactly to resonance, the output is pure 2nd harmonic distortion.

Because you drive off resonance, the power electronics have to deal with an out-of-phase current, but that isn't nearly as bad as the hard switching that occurs in PWM plasma tweeters. Just make sure to detune upwards for an inductive load and happy MOSFETs.

I don't have any hard data on the Ionovac, but I'm pretty sure it was essentially the same as the tube plasma tweeters built by hobbyists: a screen-modulated oscillator made with an 807 or the like. (Edit: found the schematic at ionovac.com

- I was almost right except the tube was a 6DQ6)

The Ionovac actually had worse distortion performance than a regular moving-coil tweeter. The plasma transducer itself (a ball of plasma in a quartz cell mounted to the throat of a horn) was quite linear, probably more so than any of the spark gap gadgets used by hobbyists now, but the tube modulator wasn't, and the uneven frequency response of the horn probably didn't help matters either.

High-level AM modulation of a MOSFET RF amp with a linear audio amp is probably the best quality/complexity tradeoff you'll get. I think EastVoltResearch's latest plasma tweeter kit works that way: I helped him design some of the audio circuitry, but I'm not sure if he ever used it, or in what particular model.

Dr. Dark Current

Tue Mar 29 2011, 09:03PM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

After experimenting with numerous driver designs and topologies for a TV transformer plasma speaker, a half-switched fullbridge with a RC damping circuit across the primary and PWM modulation worked the best. The cleanest sound was produced from a gapped non-rectified transformer.

Xray

Tue Mar 29 2011, 09:05PM

Registered Member #3429 Joined: Sun Nov 21 2010, 02:04AM
Location: Minnesota, USA
Posts: 288

Steve McConner wrote ...

The FM ones work by "slope detection". The output voltage from the resonator depends on how close the drive frequency is to the resonant frequency, so the resonator converts the FM back to AM. You can receive FM on an AM radio in the same way, using the slope of the IF filters, hence the name.

Steve -- Yes, you are totally correct about the FM units producing sound via slope detection. I used to be a Ham radio operator (many moons ago) and so I am very familiar with different modes of audio transmission and reception. I believe that Amplitude Modulation would produce the cleanest sound, but at the expense of power inefficiency. I am more concerned with fidelity than with saving a few cents on my electric bill. And I want to build a plasma speaker that I can be proud of.

If/when I actually do get around to building such a project, I will of course share my work with you guys here on 4hv.

Xray

Tue Mar 29 2011, 09:16PM

Registered Member #3429 Joined: Sun Nov 21 2010, 02:04AM
Location: Minnesota, USA
Posts: 288

Dr. Kilovolt wrote ...

After experimenting with numerous driver designs and topologies for a TV transformer plasma speaker, a half-switched fullbridge with a RC damping circuit across the primary and PWM modulation worked the best. The cleanest sound was produced from a gapped non-rectified transformer.

Could you show us a schematic diagram of it? Also it would be great if you have a picture of the whole unit, or at least the spark gap assembly. Thanks!

1 2 next

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