Developing HV DC Pulses...

[whit3rd]

On Saturday, May 27, 2023 at 10:50:14 AM UTC-4, Lamont Cranston wrote:

Does adding an Isolation transformer on the neon sign transformer primary solve the issue?
That would be easy and I have several isolation transformers, that will fit in the HV PS enclosure.

\"An isolation transformer\" is an unlikely phrase to pull up a safe solution..

The only transformer you are likely to find that has isolation up to the secondary voltage of
your neon sign transformer is... the neon sign transformer. Since you WANT DC, and
high voltage but low-ish current, the AC HV solution of that neon sign unit
is just not a good fit. Consider instead a modest (1 or 2 kV) transformer with
a voltage-multiplier output rectifier circuit, diodes and capacitors.

Most laser printers (used to) have something of that sort built in, for drum charging.
CCFL circuits used switchmode power with tiny such transformers to make kilovolts
of AC, which you can voltage-multiplier rectify.

 

[Lamont Cranston]

On Saturday, May 27, 2023 at 10:41:09 AM UTC-5, whit3rd wrote:

On Saturday, May 27, 2023 at 10:50:14 AM UTC-4, Lamont Cranston wrote:

Does adding an Isolation transformer on the neon sign transformer primary solve the issue?
That would be easy and I have several isolation transformers, that will fit in the HV PS enclosure.
\"An isolation transformer\" is an unlikely phrase to pull up a safe solution.

The only transformer you are likely to find that has isolation up to the secondary voltage of
your neon sign transformer is... the neon sign transformer. Since you WANT DC, and
high voltage but low-ish current, the AC HV solution of that neon sign unit
is just not a good fit. Consider instead a modest (1 or 2 kV) transformer with
a voltage-multiplier output rectifier circuit, diodes and capacitors.

Most laser printers (used to) have something of that sort built in, for drum charging.
CCFL circuits used switchmode power with tiny such transformers to make kilovolts
of AC, which you can voltage-multiplier rectify.

 

[Lamont Cranston]

On Saturday, May 27, 2023 at 10:41:09 AM UTC-5, whit3rd wrote:

On Saturday, May 27, 2023 at 10:50:14 AM UTC-4, Lamont Cranston wrote:

Does adding an Isolation transformer on the neon sign transformer primary solve the issue?
That would be easy and I have several isolation transformers, that will fit in the HV PS enclosure.
\"An isolation transformer\" is an unlikely phrase to pull up a safe solution.

The only transformer you are likely to find that has isolation up to the secondary voltage of
your neon sign transformer is... the neon sign transformer. Since you WANT DC, and
high voltage but low-ish current, the AC HV solution of that neon sign unit
is just not a good fit. Consider instead a modest (1 or 2 kV) transformer with
a voltage-multiplier output rectifier circuit, diodes and capacitors.

Most laser printers (used to) have something of that sort built in, for drum charging.
CCFL circuits used switchmode power with tiny such transformers to make kilovolts
of AC, which you can voltage-multiplier rectify.

It looks like this flyback has a lot of isolation, can I tie the output of this to ground?
Ebay link was 9 lines long, so this, https://tinyurl.com/bdhhuwzm
So far not finding any 1kV transformers.
Any reason I can\'t put a mutilpier on a MOT?
Mikek

 

[Glen Walpert]

On Sat, 27 May 2023 08:41:05 -0700 (PDT), whit3rd wrote:

On Saturday, May 27, 2023 at 10:50:14 AM UTC-4, Lamont Cranston wrote:

Does adding an Isolation transformer on the neon sign transformer
primary solve the issue?
That would be easy and I have several isolation transformers, that will
fit in the HV PS enclosure.

\"An isolation transformer\" is an unlikely phrase to pull up a safe
solution.

The only transformer you are likely to find that has isolation up to the
secondary voltage of your neon sign transformer is... the neon sign
transformer. Since you WANT DC, and high voltage but low-ish current,
the AC HV solution of that neon sign unit is just not a good fit.
Consider instead a modest (1 or 2 kV) transformer with a
voltage-multiplier output rectifier circuit, diodes and capacitors.

Most laser printers (used to) have something of that sort built in, for
drum charging.
CCFL circuits used switchmode power with tiny such transformers to make
kilovolts of AC, which you can voltage-multiplier rectify.

If you want to use the neon sign transformer you could use half of the
secondary winding, case to either HV terminal, the other HV terminal not
used. Half voltage but no HV connection to ground. Not optimal but easy.

 

[John Larkin]

On Sat, 27 May 2023 04:55:35 -0700 (PDT), Lamont Cranston
<amdx62@gmail.com> wrote:

I have now hit a point where I think I have a problem where I am putting one side of the neon sign transformer secondary at ground potential.
The scenario is using a sig gen to drive the tube grid, this ties the HV PS at ground (near) potential though the circuit breaker box (sure would be nice to know if the neon transformer could handle that).
I have made a drawing showing in red how the neutral/ground bond causes this to happen.
https://www.dropbox.com/s/kjut0dl8j28q3se/Pulser%20Push-pull%20design.Neutral%20ground%20problem.jpg?dl=0

Can I opto isolate the drive signal from the sig gen to isolate the ground?

Is there another way around this?


Thanks, Mikek

Ground the transformer case, use one side, make a doubler or tripler.

Ground everything to minimize firefighting and funeral expenses.

You need R3 but no other resistors, unless you want to keep R1 to
snoop cathode current.

 

[whit3rd]

On Saturday, May 27, 2023 at 12:12:40 PM UTC-4, Lamont Cranston wrote:

On Saturday, May 27, 2023 at 10:41:09 AM UTC-5, whit3rd wrote:
On Saturday, May 27, 2023 at 10:50:14 AM UTC-4, Lamont Cranston wrote:

Does adding an Isolation transformer on the neon sign transformer primary solve the issue?
That would be easy and I have several isolation transformers, that will fit in the HV PS enclosure.
\"An isolation transformer\" is an unlikely phrase to pull up a safe solution.

The only transformer you are likely to find that has isolation up to the secondary voltage of
your neon sign transformer is... the neon sign transformer. Since you WANT DC, and
high voltage but low-ish current, the AC HV solution of that neon sign unit
is just not a good fit. Consider instead a modest (1 or 2 kV) transformer with
a voltage-multiplier output rectifier circuit, diodes and capacitors.

It looks like this flyback has a lot of isolation, can I tie the output of this to ground?
Ebay link was 9 lines long, so this, https://tinyurl.com/bdhhuwzm

It\'s not clear that this is a positive DC solution; if you\'re chopping DC
with a vacuum tube triode, negative ground has to be where the common connection
point is.

> So far not finding any 1kV transformers.

The old CCTV transformers... still have some stock; this one at Digikey does 2kV out
https://www.digikey.com/en/products/detail/eaton-electronics-division/CTX310407-R/1144893

> Any reason I can\'t put a mutilpier on a MOT?

No reason not to; it is a bit overkill, but as long as the problem dies, and stays dead...

Those MOT units can output a lot of power, which you don\'t need, and
the low (60 Hz?) input frequency types are heavy. But, a variac would probably
control output easily.

 

[Lamont Cranston]

On Saturday, May 27, 2023 at 11:24:07 AM UTC-5, Glen Walpert wrote:

If you want to use the neon sign transformer you could use half of the
secondary winding, case to either HV terminal, the other HV terminal not
used. Half voltage but no HV connection to ground. Not optimal but easy.

I could do that and add a multiplier to get the 8kV we are using at this time.
I like easy. I need more HV diodes, another order.

I\'m making this thing a bit modular, because there is still AC, halfwave DC,
DC and Pulsed DC to look at. DC has not tested well, probably not viable,
but, many different types of emulsions to test.
Power Supply, Pulser Circuit, and then Voltage divider with outputs to vessels.
Here\'s my first iteration of how I\'ll wire the Pulser Circuit.
Box is 10.5\" wide x 8.5\" deep x 4.5\" tall. Plenty of room between conductors.
https://www.dropbox.com/s/drbxghel9sfjmwu/Tube%20wireing.jpg?dl=0
John I did note your comment about the resistors.

Mikek

 

[John Larkin]

On Sat, 27 May 2023 10:52:16 -0700 (PDT), Lamont Cranston
<amdx62@gmail.com> wrote:

On Saturday, May 27, 2023 at 11:24:07?AM UTC-5, Glen Walpert wrote:

If you want to use the neon sign transformer you could use half of the
secondary winding, case to either HV terminal, the other HV terminal not
used. Half voltage but no HV connection to ground. Not optimal but easy.

I could do that and add a multiplier to get the 8kV we are using at this time.
I like easy. I need more HV diodes, another order.

Your bridge uses 4 diodes. A half-wave doubler would use two.

 

[Lamont Cranston]

On Saturday, May 27, 2023 at 1:25:19 PM UTC-5, John Larkin wrote:

On Sat, 27 May 2023 10:52:16 -0700 (PDT), Lamont Cranston
amd...@gmail.com> wrote:
On Saturday, May 27, 2023 at 11:24:07?AM UTC-5, Glen Walpert wrote:

If you want to use the neon sign transformer you could use half of the
secondary winding, case to either HV terminal, the other HV terminal not
used. Half voltage but no HV connection to ground. Not optimal but easy.

I could do that and add a multiplier to get the 8kV we are using at this time.
I like easy. I need more HV diodes, another order.
Your bridge uses 4 diodes. A half-wave doubler would use two.

Hmm, yes. The 8kV diodes I have should work.
I had an excuse to order 12kV diodes, I used it!
Mikek

 

[John Larkin]

On Sat, 27 May 2023 11:37:54 -0700 (PDT), Lamont Cranston
<amdx62@gmail.com> wrote:

On Saturday, May 27, 2023 at 1:25:19?PM UTC-5, John Larkin wrote:
On Sat, 27 May 2023 10:52:16 -0700 (PDT), Lamont Cranston
amd...@gmail.com> wrote:
On Saturday, May 27, 2023 at 11:24:07?AM UTC-5, Glen Walpert wrote:

If you want to use the neon sign transformer you could use half of the
secondary winding, case to either HV terminal, the other HV terminal not
used. Half voltage but no HV connection to ground. Not optimal but easy.

I could do that and add a multiplier to get the 8kV we are using at this time.
I like easy. I need more HV diodes, another order.
Your bridge uses 4 diodes. A half-wave doubler would use two.

Hmm, yes. The 8kV diodes I have should work.
I had an excuse to order 12kV diodes, I used it!
Mikek

At low current, you can make a tripler or more. The tube pulser would
work to crazy voltages, provided you keep the grid negative enough.

A c-w multiplier can make a lot more voltage than the diodes are rated
for.

 

[Lamont Cranston]

On Saturday, May 27, 2023 at 3:35:39 PM UTC-5, John Larkin wrote:

On Sat, 27 May 2023 11:37:54 -0700 (PDT), Lamont Cranston
amd...@gmail.com> wrote:
On Saturday, May 27, 2023 at 1:25:19?PM UTC-5, John Larkin wrote:
On Sat, 27 May 2023 10:52:16 -0700 (PDT), Lamont Cranston
amd...@gmail.com> wrote:
On Saturday, May 27, 2023 at 11:24:07?AM UTC-5, Glen Walpert wrote:

If you want to use the neon sign transformer you could use half of the
secondary winding, case to either HV terminal, the other HV terminal not
used. Half voltage but no HV connection to ground. Not optimal but easy.

I could do that and add a multiplier to get the 8kV we are using at this time.
I like easy. I need more HV diodes, another order.
Your bridge uses 4 diodes. A half-wave doubler would use two.

Hmm, yes. The 8kV diodes I have should work.
I had an excuse to order 12kV diodes, I used it!
Mikek

At low current, you can make a tripler or more. The tube pulser would
work to crazy voltages, provided you keep the grid negative enough.

A c-w multiplier can make a lot more voltage than the diodes are rated
for.

Yes, I see the diodes and the caps only need to be 2X peak input voltage.
I have some 4700pf 20kV diodes coming.
If we try running a 50% duty cycle at 5Hz, that\'s a long time to discharge a tripler.
I\'m not sure if 4700pf is enough capacitance to hold up to a 10MΩ, ie how much will it sag and how much ripple?
Mikek


Mikek

 

[John Larkin]

On Sat, 27 May 2023 14:13:24 -0700 (PDT), Lamont Cranston
<amdx62@gmail.com> wrote:

On Saturday, May 27, 2023 at 3:35:39?PM UTC-5, John Larkin wrote:
On Sat, 27 May 2023 11:37:54 -0700 (PDT), Lamont Cranston
amd...@gmail.com> wrote:
On Saturday, May 27, 2023 at 1:25:19?PM UTC-5, John Larkin wrote:
On Sat, 27 May 2023 10:52:16 -0700 (PDT), Lamont Cranston
amd...@gmail.com> wrote:
On Saturday, May 27, 2023 at 11:24:07?AM UTC-5, Glen Walpert wrote:

If you want to use the neon sign transformer you could use half of the
secondary winding, case to either HV terminal, the other HV terminal not
used. Half voltage but no HV connection to ground. Not optimal but easy.

I could do that and add a multiplier to get the 8kV we are using at this time.
I like easy. I need more HV diodes, another order.
Your bridge uses 4 diodes. A half-wave doubler would use two.

Hmm, yes. The 8kV diodes I have should work.
I had an excuse to order 12kV diodes, I used it!
Mikek

At low current, you can make a tripler or more. The tube pulser would
work to crazy voltages, provided you keep the grid negative enough.

A c-w multiplier can make a lot more voltage than the diodes are rated
for.

Yes, I see the diodes and the caps only need to be 2X peak input voltage.
I have some 4700pf 20kV diodes coming.

4700 pF diodes?

If we try running a 50% duty cycle at 5Hz, that\'s a long time to discharge a tripler.
I\'m not sure if 4700pf is enough capacitance to hold up to a 10M?, ie how much will it sag and how much ripple?
Mikek


Mikek

The time constant is R*C. 4.7n * 10M = 47 milliseconds, kinda fast to
filter 60 Hz ripple.

 

[Lamont Cranston]

On Saturday, May 27, 2023 at 4:25:03 PM UTC-5, John Larkin wrote:

Yes, I see the diodes and the caps only need to be 2X peak input voltage..
I have some 4700pf 20kV diodes coming.
4700 pF diodes?

It should said capacitors.

If we try running a 50% duty cycle at 5Hz, that\'s a long time to discharge a tripler.
I\'m not sure if 4700pf is enough capacitance to hold up to a 10M?, ie how much will it sag and how much ripple?
Mikek


Mikek

The time constant is R*C. 4.7n * 10M = 47 milliseconds, kinda fast to
filter 60 Hz ripple.

Not sure I get it, half wave rectifier is 30 Hz, 33 milliseconds, that\'s a 1.42 time constants.
or about a 75% drop in the voltage. (maybe I get it) I suspect I need a little over 3 Time constants to get under 10% with
part losses. Maybe 0.02uF, I\'ll start looking-- :-(
Looks like I may need to make it at least a quadrupler to lower the capacitor voltage needed to keep the cost down on the caps.
Mikek

 

[Lamont Cranston]

On Saturday, May 27, 2023 at 5:13:46 PM UTC-5, Lamont Cranston wrote:

The time constant is R*C. 4.7n * 10M = 47 milliseconds, kinda fast to
filter 60 Hz ripple.
Not sure I get it, half wave rectifier is 30 Hz, 33 milliseconds, that\'s a 1.42 time constants.
or about a 75% drop in the voltage. (maybe I get it) I suspect I need a little over 3 Time constants to get under 10% with
part losses. Maybe 0.02uF, I\'ll start looking-- :-(
Looks like I may need to make it at least a quadrupler to lower the capacitor voltage needed to keep the cost down on the caps.
Mikek

I found these 0.04uF 3kV caps at a good price.
Have to do at least a tripler to stay under 3kV input voltage.
https://tinyurl.com/48ktmden Ebay link shortened.
What do you think?
Mikek

 

[John Larkin]

On Sat, 27 May 2023 15:13:42 -0700 (PDT), Lamont Cranston
<amdx62@gmail.com> wrote:

On Saturday, May 27, 2023 at 4:25:03?PM UTC-5, John Larkin wrote:

Yes, I see the diodes and the caps only need to be 2X peak input voltage.
I have some 4700pf 20kV diodes coming.
4700 pF diodes?

It should said capacitors.

If we try running a 50% duty cycle at 5Hz, that\'s a long time to discharge a tripler.
I\'m not sure if 4700pf is enough capacitance to hold up to a 10M?, ie how much will it sag and how much ripple?
Mikek


Mikek

The time constant is R*C. 4.7n * 10M = 47 milliseconds, kinda fast to
filter 60 Hz ripple.

Not sure I get it, half wave rectifier is 30 Hz, 33 milliseconds, that\'s a 1.42 time constants.

60 Hz half-wave rectified has one recharge blip every 16 milliseconds.
So does a multiplier.

or about a 75% drop in the voltage. (maybe I get it) I suspect I need a little over 3 Time constants to get under 10% with
part losses. Maybe 0.02uF, I\'ll start looking-- :-(
Looks like I may need to make it at least a quadrupler to lower the capacitor voltage needed to keep the cost down on the caps.
Mikek

A c-w multiplier has the caps effectively in series, so the effective
capacitance is small. Spice it.

A high frequency power supply solves a lot of problems.

 

[Jasen Betts]

On 2023-05-27, Lamont Cranston <amdx62@gmail.com> wrote:

On Saturday, May 27, 2023 at 6:55:39 AM UTC-5, Lamont Cranston wrote:
I have now hit a point where I think I have a problem where I am putting one side of the neon sign transformer secondary at ground potential.
The scenario is using a sig gen to drive the tube grid, this ties the HV PS at ground (near) potential though the circuit breaker box (sure would be nice to know if the neon transformer could handle that).
I have made a drawing showing in red how the neutral/ground bond causes this to happen.
https://www.dropbox.com/s/kjut0dl8j28q3se/Pulser%20Push-pull%20design.Neutral%20ground%20problem.jpg?dl=0

Can I opto isolate the drive signal from the sig gen to isolate the ground?

Is there another way around this?


Thanks, Mikek

Does adding an Isolation transformer on the neon sign transformer primary solve the issue?

Yes, but you would need an isolating trasformer that\'s good to stand
off 10Kv or whatever voltage the neon transformer makes.

--
Jasen.
🇺🇦 Слава Україні

 

[Jasen Betts]

On 2023-05-27, Lamont Cranston <amdx62@gmail.com> wrote:

On Saturday, May 27, 2023 at 10:41:09 AM UTC-5, whit3rd wrote:
On Saturday, May 27, 2023 at 10:50:14 AM UTC-4, Lamont Cranston wrote:

Does adding an Isolation transformer on the neon sign transformer primary solve the issue?
That would be easy and I have several isolation transformers, that will fit in the HV PS enclosure.
\"An isolation transformer\" is an unlikely phrase to pull up a safe solution.

The only transformer you are likely to find that has isolation up to the secondary voltage of
your neon sign transformer is... the neon sign transformer. Since you WANT DC, and
high voltage but low-ish current, the AC HV solution of that neon sign unit
is just not a good fit. Consider instead a modest (1 or 2 kV) transformer with
a voltage-multiplier output rectifier circuit, diodes and capacitors.

Most laser printers (used to) have something of that sort built in, for drum charging.
CCFL circuits used switchmode power with tiny such transformers to make kilovolts
of AC, which you can voltage-multiplier rectify.

It looks like this flyback has a lot of isolation, can I tie the output of this to ground?

What does the datasheet say?

> So far not finding any 1kV transformers.

The age of toobs is past.

> Any reason I can\'t put a mutilpier on a MOT?

If you don\'t mind that one end of the secondary is grounded, and
aren\'t worried by the deadly current capability. A neon transfromer
will tickle you a bit, an MOT will toast you.

--
Jasen.
🇺🇦 Слава Україні

 

[Glen Walpert]

On Sat, 27 May 2023 11:25:00 -0700, John Larkin wrote:

On Sat, 27 May 2023 10:52:16 -0700 (PDT), Lamont Cranston
amdx62@gmail.com> wrote:

On Saturday, May 27, 2023 at 11:24:07?AM UTC-5, Glen Walpert wrote:

If you want to use the neon sign transformer you could use half of the
secondary winding, case to either HV terminal, the other HV terminal
not used. Half voltage but no HV connection to ground. Not optimal but
easy.

I could do that and add a multiplier to get the 8kV we are using at this
time.
I like easy. I need more HV diodes, another order.

Your bridge uses 4 diodes. A half-wave doubler would use two.

The bridge is not usable for producing a ground referenced HVDC from a
transformer with a grounded secondary, but full wave rectification is
possible with 2 diodes from a center tapped transformer with a single
diode in each transformer HV to the filter cap, return to grounded CT.
This gives you full wave rectification at half voltage. Substitute a half
wave doubler for both diodes to get back to full voltage, still ground
referenced.

 

[Lamont Cranston]

On Sunday, May 28, 2023 at 6:10:50 AM UTC-5, Glen Walpert wrote:

On Sat, 27 May 2023 11:25:00 -0700, John Larkin wrote:

On Sat, 27 May 2023 10:52:16 -0700 (PDT), Lamont Cranston
amd...@gmail.com> wrote:

On Saturday, May 27, 2023 at 11:24:07?AM UTC-5, Glen Walpert wrote:

Your bridge uses 4 diodes. A half-wave doubler would use two.
The bridge is not usable for producing a ground referenced HVDC from a
transformer with a grounded secondary, but full wave rectification is
possible with 2 diodes from a center tapped transformer with a single
diode in each transformer HV to the filter cap, return to grounded CT.
This gives you full wave rectification at half voltage. Substitute a half
wave doubler for both diodes to get back to full voltage, still ground
referenced.

Hi Glen, I\'m not sure how to draw what you describe, but I found this, that uses the earth grounded
CT as the negative lead of the HV. I think this solves the Neutral/ground problem.
Is this what you described?
https://www.dropbox.com/s/c4t13kzi62txwqo/Centertapped%20Doubler.jpg?dl=0
Thanks, Mikek

 

[John Larkin]

On Sun, 28 May 2023 05:16:56 -0700 (PDT), Lamont Cranston
<amdx62@gmail.com> wrote:

On Sunday, May 28, 2023 at 6:10:50?AM UTC-5, Glen Walpert wrote:
On Sat, 27 May 2023 11:25:00 -0700, John Larkin wrote:

On Sat, 27 May 2023 10:52:16 -0700 (PDT), Lamont Cranston
amd...@gmail.com> wrote:

On Saturday, May 27, 2023 at 11:24:07?AM UTC-5, Glen Walpert wrote:

Your bridge uses 4 diodes. A half-wave doubler would use two.
The bridge is not usable for producing a ground referenced HVDC from a
transformer with a grounded secondary, but full wave rectification is
possible with 2 diodes from a center tapped transformer with a single
diode in each transformer HV to the filter cap, return to grounded CT.
This gives you full wave rectification at half voltage. Substitute a half
wave doubler for both diodes to get back to full voltage, still ground
referenced.

Hi Glen, I\'m not sure how to draw what you describe, but I found this, that uses the earth grounded
CT as the negative lead of the HV. I think this solves the Neutral/ground problem.
Is this what you described?
https://www.dropbox.com/s/c4t13kzi62txwqo/Centertapped%20Doubler.jpg?dl=0
Thanks, Mikek

You don\'t really need the bottom half; your load current will be low.