Author Topic: Weird snubber design (plasma speaker)  (Read 992 times)

Offline John123

  • High Voltage Technician
  • ***
  • Posts: 155
  • Karma: +2/-0
  • All the gear, no idea. UK
    • View Profile
Weird snubber design (plasma speaker)
« on: May 01, 2020, 07:20:58 PM »
Anyone ever seen a primary MOSFET snubber like this before?

That auxiliary winding is on the main core of a TV flyback transformer, primary and aux winding look to be about 5 turns each in the photos of where I found it here https://hackaday.io/project/160041-plasma-speaker-exploring-properties-of-plasma

Just wondering if any of you have seen this arrangement before? Being a plasma speaker it's probably been designed with sound quality in mind, but still I've never come across this before.

On the other hand the circuit appears to be tripping the current limit on his PSU at 5.2 amps 20v, but the arc looks like 20-30w worth of hv. He appears to be using a big heatsink with a fan too so I'm guessing a lot of power is being wasted in silicon heating.
« Last Edit: May 03, 2020, 09:25:20 AM by John123 »

Offline ritaismyconscience

  • High Voltage Technician
  • ***
  • Posts: 159
  • Karma: +3/-0
    • View Profile
Re: Weird snubber design
« Reply #1 on: May 02, 2020, 02:17:52 AM »
Haven't seen this one yet.

Looks like the second winding on a forward converter though (and probably serves the same purpose)

I'm wondering why it's wasting so much power. The guy basically built a Class D amp, which should be fairly efficient. I'm also wondering why the frequency is set so low (like 100nF + 22k for CT and RT works out to about 500Hz), and even at the high end with 510 ohm and 100nF, the frequency is still only 20kHz, which might be within hearing range.
« Last Edit: May 02, 2020, 02:24:50 AM by ritaismyconscience »

Offline John123

  • High Voltage Technician
  • ***
  • Posts: 155
  • Karma: +2/-0
  • All the gear, no idea. UK
    • View Profile
Re: Weird snubber design
« Reply #2 on: May 02, 2020, 05:09:42 AM »
That timing capacitor value might be a mistake in the schematic as don't you need a minimum of 40khz or so to reproduce the audible range with modulation properly.

So he's basically fighting the flyback part of the waveform by (unintentionally perhaps) implementing forward converter snubbing and core reset techniques? Last time I was playing with a single ended flyback plasma speaker I did notice slightly better sound quality when the winding was phased for forward converter action, so maybe something similar is going on here. But with the arc jumping that size of gap there's definitely flyback action going on there as in the forward direction you're lucky to get it to jump more than a centimeter with those kinds of supply voltages.

As for the high current draw there could be a number of things I guess, that snubber, avalanche breakdown and no peak current control? Even with the first two the current can still shoot up without peak current control.

BTW I think his project and PCB looks good, the VU meter is a great addition.

Edit: Found this on http://danyk.cz/zdroje.html its in Czech but google translate does a good job.

Its from Dans section on forward converters and the one in the plasma speaker looks like "A" without the series resistor, but having said that I've used "C" with a flyback converter without issue (albiet one limited to 50% duty cycle max and peak current control).

Edit: found another example of such a snubber in the wild:

http://danyk.cz/isstc3_en.html

Wonder if the plasma speaker guy just didn't know to add an extra diode in series with the lower coil for flyback snubbing, maybe his current draw would drop considerably with it.
« Last Edit: May 02, 2020, 10:13:14 AM by John123 »

Offline davekni

  • High Voltage Senior
  • *****
  • Posts: 555
  • Karma: +30/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Weird snubber design
« Reply #3 on: May 02, 2020, 06:47:03 PM »
John:  Nice find with that final schematic you added with edit!  It's the same as the first schematic except for reversing polarity of one transformer winding.  I suspect the first schematic has an error in winding dot placement (polarity).  With that fix it all makes sense, including the large capacitor value (3 x 4.7uF).

These schematics are all non-resonant flyback designs (transformer-coupled boost converters).  Leakage inductance is always an issue with such converters.  The set of four topologies in your upper schematic are four variations of snubbing such converters.  The fifth one you added is clever, avoiding most of the wasted power.  (BTW, non-resonant flyback is common in low-power wall-warts such as phone chargers, but with simple RC or RCD snubbing to avoid the cost of an additional winding.)

I can imagine that non-resonant primary drive is easier to control precisely for a plasma speaker.  I have no personal experience with plasma speakers, however.
David Knierim

Offline John123

  • High Voltage Technician
  • ***
  • Posts: 155
  • Karma: +2/-0
  • All the gear, no idea. UK
    • View Profile
Re: Weird snubber design
« Reply #4 on: May 03, 2020, 09:07:22 AM »
Thanks, yeah that danyk site is great and has loads of projects for people like us http://danyk.cz/index_en.html (he's that diodegonewild guy on youtube).

As drawn wouldn't most of these lossless snubbers "steal" the stored flyback energy when used with flyback converters and add extra loading to the transformer, especially with such a large capacitor? It's just on http://danyk.cz/zdroje.html it implies these snubbers are geared toward forward converters with an ungapped core rather than gapped ferrite TV flyback converters (apart from C which I've seen used on another flyback mode TV lopt driver).

Also in the original articles video his power supply limit instantly kicks in at 100w yet the arc doesn't look to be anywhere close to 100w of power (compared to my own experiments), with the heatsink being large and fan cooled it makes me think the snubber could be the cause of it. Only other thing I can think of is lack of peak current control causing it to integrate up over the course of a few cycles and saturate the core, I could be wrong but I doubt there's much fet avalanching going on with such heavy snubbing.

I'm not trying to knock the guy or anything as I think his setup looks great with the VU meter and all, but it makes my brain twitch as to why it consumes so much power and maybe the snubber is behind it.


On the subject of forward mode plasma speakers using TV hv flyback transformers, here's another one designed by the king of flybacks jan martis. This one clamps the primary voltage to the supply level and relies on turns ratio alone to determine the final output voltage (ignoring any diode delay and leakage inductance), unlike bridge designs I couldn't observe any parasitic capacitance and inductance resonant rise cheats to boost the output voltage up back when I tried it  :D Shame his website is no longer up as he spent a lot of time coming up with stuff like this
« Last Edit: May 03, 2020, 01:23:40 PM by John123 »

Offline davekni

  • High Voltage Senior
  • *****
  • Posts: 555
  • Karma: +30/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Weird snubber design (plasma speaker)
« Reply #5 on: May 03, 2020, 07:22:29 PM »
These non-resonant drive circuits aren't for typical TV-style flybacks.  They're for designs where the flyback voltage is at or below the forward voltage, which is common in small phone-chargers and such.  Notice that the two primary windings are only 3 turns each.  So there's apparently enough secondary voltage with only +-60V (+-20V/turn) primary voltage.
David Knierim

Offline T3sl4co1l

  • High Voltage Technician
  • ***
  • Posts: 153
  • Karma: +9/-0
    • View Profile
Re: Weird snubber design (plasma speaker)
« Reply #6 on: May 03, 2020, 07:23:03 PM »
If the winding were reversed, it would be a SEPIC arrangement, backfeeding into the supply to recycle reactive power.  A "quarter bridge" arrangement.  More often(?) seen with a tapped winding as here, left:



If the phasing is in fact correct as shown, it might not be very efficient...

Tim

Offline John123

  • High Voltage Technician
  • ***
  • Posts: 155
  • Karma: +2/-0
  • All the gear, no idea. UK
    • View Profile
Re: Weird snubber design (plasma speaker)
« Reply #7 on: May 04, 2020, 11:10:56 AM »
These non-resonant drive circuits aren't for typical TV-style flybacks.  They're for designs where the flyback voltage is at or below the forward voltage, which is common in small phone-chargers and such.  Notice that the two primary windings are only 3 turns each.  So there's apparently enough secondary voltage with only +-60V (+-20V/turn) primary voltage.

Gotcha, so they may not be the best choice for the guys plasma speaker?

Also the 60v schematic is for a tesla coil so I'm guessing it's not limited by the same constraints as tv flybacks in terms of output voltage and energy transfer being limited.
If the winding were reversed, it would be a SEPIC arrangement, backfeeding into the supply to recycle reactive power.  A "quarter bridge" arrangement.  More often(?) seen with a tapped winding as here, left:



If the phasing is in fact correct as shown, it might not be very efficient...

Tim
Never heard the phrase quarter bridge before! Looks interesting thought, is it some kind of ring voice modulator like they use to make the daleks voice on doctor who?

Offline davekni

  • High Voltage Senior
  • *****
  • Posts: 555
  • Karma: +30/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Weird snubber design (plasma speaker)
« Reply #8 on: May 05, 2020, 02:13:02 AM »
It could be good for the plasma speaker if the primary turn-counts are low enough.  Fine control as needed for a speaker should be easier with only the secondary resonance (winding capacitance) to deal with.  Don't have any personal experience here.
David Knierim

Offline John123

  • High Voltage Technician
  • ***
  • Posts: 155
  • Karma: +2/-0
  • All the gear, no idea. UK
    • View Profile
Re: Weird snubber design (plasma speaker)
« Reply #9 on: May 05, 2020, 11:12:25 PM »
Fair enough then. Yeah I think you're right, the more linear the output stage the better the sound.

I just set a single ended design up and oddly it sounds better when I AC couple the audio signal into the timing resistor (frequency modulation) as opposed to the comparators of the TL494 for PWM. I've just got the bog standard capacitor across the fet right now and I think you're right it introduces lots of weird effects and nonlinearities, going to give that x2 diodes and capacitor and inductor lossless snubber a try as that won't allow the primary capacitor to ring back and fourth.

Offline John123

  • High Voltage Technician
  • ***
  • Posts: 155
  • Karma: +2/-0
  • All the gear, no idea. UK
    • View Profile
Re: Weird snubber design (plasma speaker)
« Reply #10 on: May 08, 2020, 04:14:43 PM »
More progress: With my flyback topology plasma speaker set up, removing the resonant capacitor style primary snubber (like in a CRT circuit) allowed PWM audio modulation mode to sound much better, pretty clear actually. I've swapped it for that two diode, capacitor and extra air cored inductor approach and most resonant action is gone apart from a small notch around 100khz, must be on the secondary side or parasitics of the snubber itself.

Arcs are a bit smaller with no resonant rise (as expected) but that doesn't seem to matter much, another plus is the fet and primary coil don't heat up as much. I'm guessing that's down to less reactive currents and the need for accurate switching synchronization is no longer needed. When I say fet heating it wasn't getting too hot to touch or anything and it's protected from avalanching, the heatsink is currently a small flat piece of metal laying face up on the work surface  ;D

For frequency modulation mode however a primary resonant capacitor is beneficial as I think it works by slope detection, needs to be tuned to the right frequency to make it sound good and that frequency will change depending on primary turns and size of capacitor.

As for loudness, I think FM was a tad louder but not as good sounding, can't quite explain when I mean by good sounding but PWM with no resonant action is easier on the ears. I can only describe it as listening fatigue.

Going to try buffering the audio input with an opamp as looking on the scope there looks to be a bit more wiggle room for PWM which my phone headphone output can't quite reach.

Found these interesting preamps some people used with their plasma speakers, the noise filtering looks interesting maybe they're onto something with toning the signal (these are not my works and mine is set up differently, only focusing on the preamps).

« Last Edit: May 08, 2020, 04:21:04 PM by John123 »

High Voltage Forum

Re: Weird snubber design (plasma speaker)
« Reply #10 on: May 08, 2020, 04:14:43 PM »

 


* Recent Topics and Posts

post Re: Rogowski current probe
[Laboratories, Equipment and Tools]
flyglas
Today at 12:30:58 AM
post Re: Rogowski current probe
[Laboratories, Equipment and Tools]
Hydron
September 27, 2020, 11:48:35 PM
post Polyphonic Interrupter with STM32F103C8T6 (Bluepill)
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Zipdox
September 27, 2020, 11:43:14 PM
post Re: Rogowski current probe
[Laboratories, Equipment and Tools]
flyglas
September 27, 2020, 11:01:13 PM
post type of driver
[Dual Resonant Solid State Tesla coils (DRSSTC)]
octopus1
September 27, 2020, 10:51:03 PM
post Re: Easy to build Polyphonic MIDI Interrupter
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Max
September 27, 2020, 10:44:43 PM
post Re: MIDI videos from my DRSSTC1/2/3 demonstrations/shows
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mads Barnkob
September 27, 2020, 09:44:58 PM
post Re: DRSSTC Tuning for Music vs Big Sparks?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
September 27, 2020, 09:20:32 PM
post Re: Easy to build Polyphonic MIDI Interrupter
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
futurist
September 27, 2020, 07:49:49 PM
post Stroboscopic photography & motors
[Digital Compact Camera]
klugesmith
September 27, 2020, 07:08:06 PM
post Re: Led Closeups with Mobile phone and lens .
[Smart Phones]
klugesmith
September 27, 2020, 06:43:46 PM
post Re: DRSSTC Tuning for Music vs Big Sparks?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mads Barnkob
September 27, 2020, 06:42:46 PM
post Re: Finding accelerometer location inside a phone
[Smart Phones]
klugesmith
September 27, 2020, 06:36:41 PM
post Re: steam engine
[Capacitor Banks]
plasma
September 27, 2020, 02:32:14 PM
post Re: Easy to build Polyphonic MIDI Interrupter
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Max
September 27, 2020, 02:02:53 PM
post Re: Easy to build Polyphonic MIDI Interrupter
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
futurist
September 27, 2020, 12:24:20 PM
post Re: Easy to build Polyphonic MIDI Interrupter
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Max
September 27, 2020, 12:52:36 AM
post Re: DRSSTC Tuning for Music vs Big Sparks?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
September 26, 2020, 11:06:04 PM
post Re: One NOS ac flyback is nice. Finding its identical twin 3 months later is awesome
[Transformer (Ferrite Core)]
profdc9
September 26, 2020, 10:48:41 PM
post Re: Rogowski current probe
[Laboratories, Equipment and Tools]
Hydron
September 26, 2020, 05:27:37 PM
post Re: DRSSTC Tuning for Music vs Big Sparks?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Zipdox
September 26, 2020, 04:32:53 PM
post Re: DRSSTC Tuning for Music vs Big Sparks?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mads Barnkob
September 26, 2020, 04:17:51 PM
post Re: DRSSTC Tuning for Music vs Big Sparks?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
GKnapp
September 26, 2020, 02:06:04 PM
post Re: Rogowski current probe
[Laboratories, Equipment and Tools]
flyglas
September 26, 2020, 10:57:58 AM
post Re: DRSSTC Tuning for Music vs Big Sparks?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mads Barnkob
September 26, 2020, 08:55:27 AM
post Re: DRSSTC Tuning for Music vs Big Sparks?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
September 26, 2020, 05:19:45 AM
post Re: Rogowski current probe
[Laboratories, Equipment and Tools]
klugesmith
September 26, 2020, 04:46:12 AM
post Re: Easy to build Polyphonic MIDI Interrupter
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Max
September 26, 2020, 01:14:47 AM
post Re: Rogowski current probe
[Laboratories, Equipment and Tools]
flyglas
September 25, 2020, 09:41:10 PM
post Re: DRSSTC Tuning for Music vs Big Sparks?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mads Barnkob
September 25, 2020, 08:40:39 PM
post Re: Easy to build Polyphonic MIDI Interrupter
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Mads Barnkob
September 25, 2020, 08:26:52 PM
post Re: FPGA speeds these days?
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
SteveN87
September 25, 2020, 06:56:41 PM
post Re: FPGA speeds these days?
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
klugesmith
September 25, 2020, 06:15:47 PM
post Re: FPGA speeds these days?
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
SteveN87
September 25, 2020, 01:56:27 PM
post Re: FPGA speeds these days?
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
TMaxElectronics
September 25, 2020, 11:55:05 AM
post Re: FPGA speeds these days?
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Weston
September 25, 2020, 09:18:51 AM
post One NOS ac flyback is nice. Finding its identical twin 3 months later is awesome
[Transformer (Ferrite Core)]
extremeodd
September 25, 2020, 03:43:26 AM
post Re: FPGA speeds these days?
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
davekni
September 25, 2020, 03:42:56 AM
post FPGA speeds these days?
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
klugesmith
September 25, 2020, 02:13:40 AM
post Re: Server PSU OPV Hack - Fujitsu DPS-800GB-5 A
[Electronic Circuits]
klugesmith
September 24, 2020, 03:02:00 PM
post Re: Server PSU OPV Hack - Fujitsu DPS-800GB-5 A
[Electronic Circuits]
v.coelho
September 24, 2020, 01:11:27 PM
post Re: Info needed: probably old military High Voltage Multiplier?
[Voltage Multipliers]
Zipdox
September 24, 2020, 10:25:42 AM
post Re: Info needed: probably old military High Voltage Multiplier?
[Voltage Multipliers]
johnnyzoo
September 24, 2020, 07:32:43 AM
post Re: Current transformer for narrow spaces
[Laboratories, Equipment and Tools]
klugesmith
September 23, 2020, 11:20:53 PM
post WTS [US]:"Terry filter" varistors for sale
[Sell / Buy / Trade]
MRMILSTAR
September 23, 2020, 10:41:54 PM
post Re: Info needed: probably old military High Voltage Multiplier?
[Voltage Multipliers]
Zipdox
September 23, 2020, 10:25:07 PM
post Info needed: probably old military High Voltage Multiplier?
[Voltage Multipliers]
HighVoltageRulezz
September 23, 2020, 10:06:30 PM
post Re: Server PSU OPV Hack - Fujitsu DPS-800GB-5 A
[Electronic Circuits]
petespaco
September 23, 2020, 04:26:34 PM
post Re: 6" Coil Update now 72" and better lighting
[Spark Gap Tesla Coils (SGTC)]
klugesmith
September 23, 2020, 03:06:28 PM
post Re: Server PSU OPV Hack - Fujitsu DPS-800GB-5 A
[Electronic Circuits]
Zipdox
September 23, 2020, 01:34:42 PM

Sitemap 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 
SimplePortal 2.3.6 © 2008-2014, SimplePortal