Hipot Design Ideas?

[Michael D. Harmon]

I recently picked up an old Allen "high tension tester" at at the local
electronics emporium. It was originally designed for checking
automotive HV electrical systems, and is basically a hipot using a
vibrator(!), a 6V ignition coil, and a 1B3 rectifier tube to generate
0-15,000 V at a few uA. It was powered by a 6V. 12V, or 24V battery.
The output voltage was controlled by a rheostat in the coil primary.
The caps in the thing were old wax-paper aluminum-foil 50's vintage
units, and the HV leads were old stiff rubber spark-plug wire. I did
get a couple of excellent-condition panel meters as well as the HV glass
multiplier resistors

I'm planning to redesign this thing for 117VAC, using silicone spark
plug wire for the HV leads. I still need to keep the variable output
capability. I'd also like to incorporate some sort of current limiting
capability to reduce the death-factor.

Anyone have any ideas for an effective, yet relatively simple way to
generate 0-15,000 KV DC at say 50 uA or less? I've thought of several
ways of doing this. The first is simply using the same basic design,
but replacing the vibrator with a solid-state triggering circuit and the
1B3 TV rectifier tube with oscilloscope HV diodes. Another was to use a
HV transformer(neon?) and a small Variac. Still another was to use a
surplus flyback, but those things have a bunch of unknown leads sticking
out the bottom (and no diagram!). I'd still have to come up with a
trigger oscillator/driver however.

I'm open for suggestions!

Thanks,
Mike Harmon, WB0LDJ
[email protected]

 

[Fred Bloggs]

I recently picked up an old Allen "high tension tester" at at the local
electronics emporium. It was originally designed for checking
automotive HV electrical systems, and is basically a hipot using a
vibrator(!), a 6V ignition coil, and a 1B3 rectifier tube to generate
0-15,000 V at a few uA. It was powered by a 6V. 12V, or 24V battery.
The output voltage was controlled by a rheostat in the coil primary. The
caps in the thing were old wax-paper aluminum-foil 50's vintage units,
and the HV leads were old stiff rubber spark-plug wire. I did get a
couple of excellent-condition panel meters as well as the HV glass
multiplier resistors

I'm planning to redesign this thing for 117VAC, using silicone spark
plug wire for the HV leads. I still need to keep the variable output
capability. I'd also like to incorporate some sort of current limiting
capability to reduce the death-factor.

Anyone have any ideas for an effective, yet relatively simple way to
generate 0-15,000 KV DC at say 50 uA or less? I've thought of several
ways of doing this. The first is simply using the same basic design,
but replacing the vibrator with a solid-state triggering circuit and the
1B3 TV rectifier tube with oscilloscope HV diodes. Another was to use a
HV transformer(neon?) and a small Variac. Still another was to use a
surplus flyback, but those things have a bunch of unknown leads sticking
out the bottom (and no diagram!). I'd still have to come up with a
trigger oscillator/driver however.

I'm open for suggestions!

Thanks,
Mike Harmon, WB0LDJ
[email protected]

I have always been partial to the unijunction (2N6027) oscillator
triggering a HV SCR to discharge a cap into the step transformer with
secondary to the HV diode and filter capacitor. So you can rectify and
double the line for 340Vpk to start. The HV transformer provides
isolation. Simple thyristor phase control on the line input and
unijunction trigger frequency can adjust the output level. There should
be plenty of reference designs around for this:
View in a fixed-width font such as Courier.

..
..
.. +------+
.. LINE---+-------+PHASE-+---+-----------+ +---|>|---+----+->+
.. | | CTL | | | | | |
.. | |RECT | |+ o| |o | |
.. | +------+ === )||( === Rbleed
.. | | )||( | |
.. | | )||( | |
.. +-+---+ | )||( | |
.. |XFRMR|------+ | | | | |
.. +-+---+ +-+--+ | | +---------+----+-->-
.. | | | | ---
.. | |OSC |----|----------\ /
.. | | | | ---
.. | +-+--+ | |
.. | | | |
.. NEUT--+----------+-------+-----------+
..

 

[Joerg]

Hello Michael,

I'm open for suggestions!

Ok, here is one: Take an old flyback transformer from a scrapped color
TV. Take the transistor from the horizontal final, too, along with the
rectifier stack (often a cascade). The only special part you'd need to
procure would be a wee high voltage capacitor. In a TV the capacitance
between tube metallization and outer shield serves as a capacitor but
that would be too large a contraption for your purposes.

Be careful with this stuff. It can generate dangerous and even lethal
voltage and energy levels. Even when supplied from a battery.

Regards, Joerg

 

[Chris Jones]

I recently picked up an old Allen "high tension tester" at at the local
electronics emporium. It was originally designed for checking
automotive HV electrical systems, and is basically a hipot using a
vibrator(!), a 6V ignition coil, and a 1B3 rectifier tube to generate
0-15,000 V at a few uA. It was powered by a 6V. 12V, or 24V battery.
The output voltage was controlled by a rheostat in the coil primary.
The caps in the thing were old wax-paper aluminum-foil 50's vintage
units, and the HV leads were old stiff rubber spark-plug wire. I did
get a couple of excellent-condition panel meters as well as the HV glass
multiplier resistors

I'm planning to redesign this thing for 117VAC, using silicone spark
plug wire for the HV leads. I still need to keep the variable output
capability. I'd also like to incorporate some sort of current limiting
capability to reduce the death-factor.

Anyone have any ideas for an effective, yet relatively simple way to
generate 0-15,000 KV DC at say 50 uA or less? I've thought of several
ways of doing this. The first is simply using the same basic design,
but replacing the vibrator with a solid-state triggering circuit and the
1B3 TV rectifier tube with oscilloscope HV diodes. Another was to use a
HV transformer(neon?) and a small Variac. Still another was to use a
surplus flyback, but those things have a bunch of unknown leads sticking
out the bottom (and no diagram!). I'd still have to come up with a
trigger oscillator/driver however.

I'm open for suggestions!

Thanks,
Mike Harmon, WB0LDJ
[email protected]

Watch out for X-rays from those old rectifier valves (tubes).

A TV flyback is probably the easiest way to get 15kV. There are many
designs on the net which involve winding your own primary onto a TV
flyback, and driving it with a push-pull oscillator built from two 2N3055s.
I think that these designs stress the diodes more than would be the case
when producing the same output voltage in a TV, since in the TV, the
secondary produces a very large voltage in one polarity but a much smaller
voltage in the other polarity, so in the TV, the total peak voltage across
the HV diodes is only slightly higher than the output voltage. In the
push-pull version, the secondary produces high voltage in both polarities,
which could almost double the stress on the diodes. Note that in reality,
most flybacks have several EHT secondary windings which each has its own
rectifier diode, and the whole lot are in series. You probably have some
margin to abuse the diodes however, since you are only asking for 15kV
whereas the flyback from a colour TV would be designed to produce a higher
voltage.

You could drive the flyback much the same way that it is driven in a TV set,
and you could even reuse the transistor and HV capacitor that are used on
the primary of the flyback. Identifying the windings is possible with an
ohmmeter, but it is even better if you have an oscilloscope and a signal
generator: wrap a few turns around the core and drive these with a few tens
of millivolts of AC from the signal generator, and then measure the voltage
an polarity of what is induced in all of the other windings.

There are lots of web pages about driving flyback transformers, some of
which are useful.

Chris

 

[Ken Smith]

[...]
Note this:

capability to reduce the death-factor.

[....]
I have always been partial to the unijunction (2N6027) oscillator
triggering a HV SCR to discharge a cap into the step transformer with
secondary to the HV diode and filter capacitor.

When I was a kid, I nearly killed myself with a circuit like this. The
step up xformer I used was a car ignition transformer.


Modifying drawing to closer to what I did:

. Variac
. LINE------->!----\/\/--+--+-----------+ +---|>|---+----+->+
. ! | | | | |
. -------------- |+ o| |o | |
. ! === )||( === Rbleed
. \ | )||( | |
. / | )||( | |
. \ | )||( | |
. / | | | | |
. ! +-+--+ | | +---------+----+-->-
. +-------| | | ---
. ! |OSC |----|----------\ /
. 24V --- | | | ---
. Zener ^ +-+--+ | |
. ! | | |
. NEUT---+---------+-------+-----------+

In fact, my circuit used a diac as the oscillator not a unijunction, and
the positions of the SCR and primary were reversed.

If you want to semi-regulate the output, you can: Add a rectifier on the
anode of the SCR and monitor it. When the SCR stops conducting, the
transformer is ringing. The ringing couples back into the primary and its
amplitude is somewhat controlled by the output voltage.


If I was doing it today, I would look at using a IGBT transistor or two.
An average of the supply current would allow you to limit the output
current to something less deadly.

 

[Tony]

Watch out for X-rays from those old rectifier valves (tubes).

A TV flyback is probably the easiest way to get 15kV. There are many
designs on the net which involve winding your own primary onto a TV
flyback, and driving it with a push-pull oscillator built from two 2N3055s.
I think that these designs stress the diodes more than would be the case
when producing the same output voltage in a TV, since in the TV, the
secondary produces a very large voltage in one polarity but a much smaller
voltage in the other polarity, so in the TV, the total peak voltage across
the HV diodes is only slightly higher than the output voltage. In the
push-pull version, the secondary produces high voltage in both polarities,
which could almost double the stress on the diodes.

Chris - maybe you could clarify this with some specifics (all this
makes no sense to me at all so far)?

TV flyback transformers are usually driven by high voltage transistors
(eg 1200V), as you only get EHT on the secondary when there is also HV
on the primary (usually the flyback pulse when the single-ended
transistor switches off). The drive COULD be push-pull, but to what
end? To get the same EHT out you'd still need HV on the primary
(perhaps as much as 1000V), which would be a lot more difficult to
handle than a simple single-ended flyback design running off low
voltage DC. The secondary rectifier would then be a bridge, else the
push-pull drive would "go nowhere"; the bridge doesn't suffer the
voltage problems you mention above.

It all seems hard to justify unless you really need to extract much
more power.

And no matter what the circuit, 2N3055s wouldn't even come close.

Tony
Tony (remove the "_" to reply by email)

 

[Joerg]

Hello Tony,

And no matter what the circuit, 2N3055s wouldn't even come close.

They sure don't. If you apply the normal line frequency pulses they'd
just go "Huh?".

If you want to make life easier use a HV FET. Much easier to drive.

Regards, Joerg