First Thread : A challenge and an opportunity for all of you here! / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

First Thread : A challenge and an opportunity for all of you here!

previous 1 2 3 4 next

Move Thread

LAN_403

sashua

Tue Jul 13 2010, 03:30PM

Registered Member #2984 Joined: Fri Jul 09 2010, 02:55PM
Location:
Posts: 4

We look forward to hearing your collective advice!

Sulaiman

Tue Jul 13 2010, 05:34PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3141

A bit more on specifications (cost/size/weight)

A) Producing hv audio using a switcher

The upper part of the audio spectrum is only harmonics (& distortion)
this is why "CD Quality" only samples at 44 kHz.
With 40 kHz switching frequency you can still claim 20 kHz bandwidth.
The difference between a 40 kHz switcher and a 200 kHz switcher with +/-5 kV ac pk output
will be rather significant due to insulation problems.
________________________________________ ______________________________
B) Producing hv audio via a linear audio transformer.

Imagine the weight/size/cost of a 50/60 Hz transformer rated 3.5 kV 35 mA (5kV 50mA pk)
Now at least quadruple the cost for audio use to ??kHz
Now double it again for 25/30 Hz lower frequency bandwidth.
If the direct audio modulation via a transformer method is used the audio bandwidth specified will significantly alter the cost/size/ weight.
__________________________________________ _____________________________

In either case 1% thd is going to be a challenge.
I'm not a 'lifter' constructor but I thought 20 kV is too low for a decent effect.

cdanjo

Thu Jul 15 2010, 11:20PM

Registered Member #2995 Joined: Mon Jul 12 2010, 10:12PM
Location:
Posts: 4

Good points. People seem to worry about aliasing at low-ish switching frequencies like 40kHz, and I've been told to shoot for about 10x the highest frequency to reduce hash/jitter noise, but I could certainly be happy with ~40kHz for a first-run device just to get the experiment proved. THD under 1% is difficult for any switching device for mid/high frequencies; they work quite well for subs though.

Yes, the linear transformer is quite a terrible idea given the DC current it would have to withstand!

I just love that commercial where the guys are looking for excuses to indulge in extravagance, and the one guy, sitting next to a nice shiney tube amplifier, says "it can reproduce frequencies only dogs can hear" !!

Pinky's Brain

Mon Jul 19 2010, 05:27PM

Registered Member #2901 Joined: Thu Jun 03 2010, 01:25PM
Location:
Posts: 837

I'm just wondering, why couldn't you do this with a stack of MOSFETs with floating opamp drivers which control the drain source voltage? Floating closed loop drivers have been used for high voltage switching with MOSFET stacks ... why wouldn't it work for modulation?

It's obviously not a trivial circuit, but it's certainly going to be cheaper/smaller than a transformer based solution IMO.

Crunchy Frog

Wed Jul 21 2010, 02:28AM

Registered Member #2422 Joined: Tue Oct 06 2009, 02:41AM
Location:
Posts: 85

Are you making one of those ion "wind" speaker things? I know someone who made one of those a while ago who I may be able to put you in contact with or get a schematic from.

teravolt

Wed Jul 21 2010, 03:26AM

Registered Member #195 Joined: Fri Feb 17 2006, 08:27PM
Location: Berkeley, ca.
Posts: 1111

Is the output supposed to be sinusoidal and what kind of harmonic distortion can you live with?
Is the load static or dynamic?
what is the end purpose, maybe there is a better way to accomplish your task

Carl Pugh

Wed Jul 21 2010, 04:49PM

Registered Member #1064 Joined: Tue Oct 16 2007, 05:04PM
Location:
Posts: 42

teravolt opens an interesting approach.

Can a slow rise and fall square wave be used?

A modulator with a slow rise and fall square wave output would be much smaller and lower cost than a modulator with a sine wave output.

Pinky's Brain

Wed Jul 21 2010, 10:04PM

Registered Member #2901 Joined: Thu Jun 03 2010, 01:25PM
Location:
Posts: 837

Still wondering ... why wouldn't a real implementation of this pseudo circuit work :

Of course you would need 7 1500V MOSFETs, with independent drivers with their own power supply (probably buck converters, seems easier than lots of transformers) and circuitry for getting it to power up without blowing up. Each of those MOSFET needs overvoltage protection, you need resettable overcurrent protection for the main high voltage power supply. Perhaps some filtering in the drivers if there is danger of spontaneous oscillation, etc etc.

Still, I don't see why this wouldn't work.

PS. used low supply voltage because the simplified models wouldn't work at high voltage.

[Edit: Size]

Sulaiman

Thu Jul 22 2010, 12:12PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3141

It is almost impossible to match the turn on delay- and rise-times for each transistor,
or turn-off delay- and fall-times.
the last one to turn on (or the first one to turn off) would get full voltage.

Pinky's Brain

Thu Jul 22 2010, 03:17PM

Registered Member #2901 Joined: Thu Jun 03 2010, 01:25PM
Location:
Posts: 837

You don't really turn them on or off as such (they are always conducting, they have to ... because regardless of voltage sharing they can not handle the full supply voltage, most of it is always across the load). Each stage has a closed loop control following an input signal with limited slew rate (audio frequency). The drain-source voltages of the faster (higher bandwidth) stages will have some oscillation, but I very much doubt the amplitude will be significant.

previous 1 2 3 4 next

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