Win's next 10kV project, a 1us ramp

[Winfield Hill]

Spehro Pefhany wrote...

How about a HY11 or similar hydrogen thyratron. ~20kV anode
voltage, 1600A peak anode current, and fits in the palm of
your hand? Plus a passive shaping network, of course.

How fast can they switch, can they do 10kV/us or more?

 

[Winfield Hill]

Ken Smith wrote...

Winfield Hill wrote:
[....]

Now I'm challenged with a creating a moderately-precise (5%),
programmable, fairly-fast 10kV ramp, of 1 to 50us in duration.
It's floating on 3kV. The 1us-long, 3kV to 13kV ramp spec is
tough, but may have a relaxed 5%-precision spec, provided it's
repeatable.

ASCII Art:



pivot 0================0===============0
! # !
Copper plate ================ ### ! ! ! ! ! Piston
Gap ### ! ======= ! assembly
Copper plate ================ ### ===== =====
Non conductive ! ! ### ! !
mounting ! ! ### X-X Steam valve
. ......### / \
. . ### ! ! Boiler
. . ### -------
. . Conductive
. . mounting
. .
. .
cam operated . . Lead wires
switch ! . GND
--- .
3KV -----O O-----+--------------- To load

ROFLOLMGO!

 

[Joerg]

Hello Michael,

The 6BK4? I have two NOS 6BK4C/6EL4A Westinghouse tubes on hand for
$5.00 (US) each. they are available from a lot of surplus dealers, and
the bunch over on could probably fill a
dump truck with them.

I just don't know whether they could do the pulse current in Win's
application. You could probably extrapolate and drive them a little into
the grid current region but there comes a point where particles may be
ripped off the cathode.

Regards, Joerg

 

[Joerg]

Hello Win,

Hah, that's more current than they're up for anyway! Certainly if
I'm switching anywhere from 0.5 to 2A, a short-lived FET avalanche
isn't something to fear. But if we did fear this (or, ahem, if it
failed to pass a know-nothing design review) ...

What if you'd package the avalanche part of the presentation with some
really scientific lingo, lots of complicated math going all the way back
to Maxwell, tons of integrals in there? Maybe folks won't ask out of
fear to embarrass themselves in front of everybody.

Regards, Joerg

 

[John Larkin]

John Larkin wrote...

That could work.

OK, now we're talking a fet switch followed by an R-L-C ramp forming
network with a few relay-switchable elements. The switch could be a
stack of fets with transformer-coupled gate drive. That doesn't sound
nearly as bad as trying to control the stacked fets in linear mode.

What's the actual breakdown voltage of RG-174 sized coax? I get it's
way above 10KV. So a nice little transmission-line transformer would
be easy to make.

I used to run surplus RG-59 at 15KV when I was a kid, playing with
flashtubes and spark gaps and stuff.

John

 

[John Larkin]

Spehro Pefhany wrote...

How fast can they switch, can they do 10kV/us or more?

Yes. Expect under a ns of jitter if you drive them right.

Hey, how about a krytron? Size of a walnut. They have a keep-alive and
a radioactive thingie, so jitter is very low.

John

 

[Winfield Hill]

John Larkin wrote...

What's the actual breakdown voltage of RG-174 sized coax?
I get it's way above 10kV.

Where'd you get that? IIRC, it's about 1kV for extended use.
And I've had RG-58 breakdown at 2kV. We use special grades
of RG-9 at 10kV, but not with any of the usual connectors.

So a nice little transmission-line transformer would be
easy to make.

To make a nice 20us ramp the transformer needs a good low-
frequency response, which implies a fair-sized core. It
may not need transmission-line transformer HF bandwidth.

I used to run surplus RG-59 at 15KV when I was a kid,
playing with flashtubes and spark gaps and stuff.

How long did it last? Most anything will last a few minutes
or hours, even, before the breakdown paths start killing you.

 

[Jim Thompson]

OK, now we're talking a fet switch followed by an R-L-C ramp forming
network with a few relay-switchable elements. The switch could be a
stack of fets with transformer-coupled gate drive. That doesn't sound
nearly as bad as trying to control the stacked fets in linear mode.

What's the actual breakdown voltage of RG-174 sized coax? I get it's
way above 10KV. So a nice little transmission-line transformer would
be easy to make.

I used to run surplus RG-59 at 15KV when I was a kid, playing with
flashtubes and spark gaps and stuff.

John

Doesn't RG-8 have a really good HV rating?

For my 30KV stuff at MIT we used parallel-plate (aluminum)
transmission lines with Teflon as the dielectric ;-)

...Jim Thompson

 

[John Larkin]

Robert Baer wrote...

I like the constant-current into a capacitive load that a
flyback coil / transformer can deliver. If the transformer
has a 10:1 step-up ratio I could use 1kV MOSFETs to switch
5 to 10A on the primary... But I keep worrying about the
transformer's leakage inductance. If the secondary + load
was say 60pF total, then that'd have to be less than 17uH
for a say 5MHz bandwidth. Hmm, 17uH doesn't look so hard.

A coax transmission-line transformer is easy to make and has pretty
much zero leakage. I'd just be concerned about getting enough core to
support your longer sweeps without saturating. Sometimes people drive
the core the other way before or after each shot to demagnetize (or
reverse magnetize) it and double the available volt-amps.

Hey, could you drive your electrode with a really high voltage sine
wave and use the sorta linear middle region? Synchronize your particle
beam shots (or whatever you're doing) with the sines. That would be a
simple resonant hv oscillator, dead simple.

Or use the first swing of a damped sinewave?

John

 

[John Larkin]

ASCII Art:



pivot 0================0===============0
! # !
Copper plate ================ ### ! ! ! ! ! Piston
Gap ### ! ======= ! assembly
Copper plate ================ ### ===== =====
Non conductive ! ! ### ! !
mounting ! ! ### X-X Steam valve
. ......### / \
. . ### ! ! Boiler
. . ### -------
. . Conductive
. . mounting
. .
. .
cam operated . . Lead wires
switch ! . GND
--- .
3KV -----O O-----+--------------- To load



Or, perhaps:

Some ceramic capacitors have decreasing capacitance for increasing
voltage. You may be able to use these to straighten out an RC ramp
reducing the problem to just one of switching.

--

Gunpowder/supermagnet bullet/shaped coil.

John

 

[Joerg]

Hello John,

Hey, cool:

Max plate voltage 4000, but if you look at the amount of vacuum in
there, I'd go for lots more.

4 amps Id (er... Ip) with about +350 on the grid *and* the plate.

Cout probably in the 15 pF range.

And needs a mere 10 volts, 10 amps for the filament.

I have a couple, and they're beautiful hunks of glass.

My favorite is the QB5/1750. The plate connection alone is an impressive
steel rod of about 1/3" diameter. This tube has no problem whatsoever to
sink a few amps. Even with the grid at 0V it already sits around 500mA.

http://geocities.com/nikolic_lj_dragan/Ham/PhilipsQB5.htm

Regards, Joerg

 

[Joerg]

Hello Kevin,

EHT shunt regulators as used in old colour televisions may still be
available and meet your needs. e.g. 6BD4 They would only have low
heater cathode insulation so would need a floating supply for that.

http://tubedata.tigahost.com/tubedata/sheets62.html

Do you know what their peak current rating is for short duty cycles? The
specs only cover the mA range because that's all they needed in TV designs.

Regards, Joerg

 

[John Larkin]

Doesn't RG-8 have a really good HV rating?

Sure, but who wants to wind that on a toroid?

For my 30KV stuff at MIT we used parallel-plate (aluminum)
transmission lines with Teflon as the dielectric ;-)

I'll have to test some small teflon coax just for fun. I have an old
Spellman hv supply that goes to 30 KV.

John

 

[Joerg]

Hello Win,

Hmm.

Sven is right. Toroids don't leak much if the AL value of the core is
high enough. Designing highly isolated supplies and signal coupling is
my bread and butter and I never had leakage become a problem. #77
ferrite is great, or #43 for fast stuff. Getting certified wire for the
primary is a huge job but only because my designs go into production in
a heavily regulated industry. For a lab test rig you might get away with
spark plug wire from the auto shop if you don't need many turns.

I don't mind that part - my technician will do it.

It's not difficult. Especially not in a lab situation, as long as your
tech fully understands how not to compromise a 13kV barrier. And as long
as the techs never, ever fire it up on their own.

Take a large enough core to make it easy for the techs. Up to 2" OD is
often in stock and should be fine, larger could be special order.

Regards, Joerg

 

[John Larkin]

Hello John,


My favorite is the QB5/1750. The plate connection alone is an impressive
steel rod of about 1/3" diameter. This tube has no problem whatsoever to
sink a few amps. Even with the grid at 0V it already sits around 500mA.

http://geocities.com/nikolic_lj_dragan/Ham/PhilipsQB5.htm

Regards, Joerg

Very pretty, except for the radiant heat peeling the stickers off the
fans.

I think Win should do it with a big tube, just for the sheer novelty
of it.

See abse for my 833A.

John

 

[Joerg]

Hello Jim,

It's been a lo-o-o-ong time since I've lurked around a TV repair shop.
Wonder how fast HV regulator tubes like 6BD4/6BK4 can be pushed?

They are fast but there is no peak current data because it didn't matter
for the HV regulator. If you overdo the peaks they might not live long.

Regards, Joerg

 

[Tim Williams]

I just don't know whether they could do the pulse current in Win's
application. You could probably extrapolate and drive them a little into
the grid current region but there comes a point where particles may be
ripped off the cathode.

As I recall, the heater requirements are on par with a 12AX7, so the
emission is on par. Same as 12AU7 yes, which is a bit beefier, so the
electrons may just be hiding behind the grid; but still nothing anywhere
near needed. You'll probably get the grid hot enough to boil electrons off,
and still not have enough emission ;-)

Although... 10 or 20 in parallel might work. That's a lot of dead weight
capacitance though, isn't it?

Tim

 

[Winfield Hill]

John Larkin wrote...

I think Win should do it with a big tube, just for the
sheer novelty of it. See abse for my 833A.

AWESOME!! I'd have to use three or four in series?

 

[Winfield Hill]

John Larkin wrote...

Hey, how about a krytron? Size of a walnut. They have a
keep-alive and a radioactive thingie, so jitter is very low.

Aren't those those things where the FBI comes and arrests you?

 

[Winfield Hill]

John Larkin wrote...

A coax transmission-line transformer is easy to make and has pretty
much zero leakage. I'd just be concerned about getting enough core to
support your longer sweeps without saturating. Sometimes people drive
the core the other way before or after each shot to demagnetize (or
reverse magnetize) it and double the available volt-amps.

Hey, could you drive your electrode with a really high voltage sine
wave and use the sorta linear middle region? Synchronize your particle
beam shots (or whatever you're doing) with the sines. That would be a
simple resonant hv oscillator, dead simple.

Or use the first swing of a damped sinewave?

Perhaps that could work, as a fallback. I'll have to check.