Author Topic: Neon power supply + voltage multiplier producing too much voltage  (Read 420 times)

Offline Davide

  • High Voltage Enthusiast
  • *
  • Posts: 5
  • Karma: +0/-0
    • View Profile
Hi everyone!
I'm an high voltage noob so I hope my question is not too obvious. What I'm trying to make is a power supply to power an x-ray tube, and to do that I decided to connect the output of a Neon power supply (this one: https://it.aliexpress.com/item/4000151726656.html?spm=a2g0s.9042311.0.0.6a9d4c4dNXmRFC) to a CW voltage multiplier made of 3 stages. The output of the Neon power supply is 10kV AC, so, with three stages of multiplication I should have around 84kV DC, as the CW calculator says (link: https://www.extremeelectronics.co.uk/calculators/cw-voltage-calculator/).
However, when I try to strike an arc between the two terminals of the voltage multiplier, no matter how much I increase their distance, the arc will always be able to jump, reaching over 12cm of lenght! Moreover, if I move the terminals too far apart, the arc will strike inside the oil where the voltage multiplier is!  :o
Considering this, I concluded that this setup is actually producing way more than 84kV: the problem is that I don't know why.
My concern is that if I will connect this source to the x-ray tube (which is rated to 70kV) it will damage it.

I know I'm probably doing something wrong in the peak voltage calculation, so if someone could adress me on the right way it would be great! :) Also if you think this setup is not the right way to build an x-ray tube power supply feel free to comment as I'm also considering other alternatives (e.g. flybacks).

I attached the images of the Neon power supply, of the long arc, and of the Cockcroft Walton calculator.

Thank you very much in advance!  :)




Offline davekni

  • High Voltage Expert
  • ******
  • Posts: 1180
  • Karma: +57/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Neon power supply + voltage multiplier producing too much voltage
« Reply #1 on: May 10, 2021, 03:21:24 AM »
I'm not sure how accurate spark length is as a measure of voltage, especially when electrode shapes aren't controlled and aren't spaced far from any other surfaces.  However, the voltage may still be too high.

Neon sign supplies are designed to provide constant current, not constant voltage.  The 10kV rating is starting-voltage.  Must supplies will shut down if the output voltage remains high for more than 1 second or so.  (My smaller "8kV 20mA" supply shuts down if the output RMS voltage isn't loaded down below 3.5kV.)

I'm a bit surprised that your supply would output more than 10kV starting voltage.  Most such cheap parts are over-rated, performing under rated specifications.  However, the 10kV starting voltage may not be at all accurately controlled.  Even if you fix the 10kV accuracy issue, I doubt your supply will output 10kV for long.  (The supply is rated at 30mA and 20-120W.  The high end of 120W is only 4kV * 30mA operating conditions.)

David Knierim

Offline Davide

  • High Voltage Enthusiast
  • *
  • Posts: 5
  • Karma: +0/-0
    • View Profile
Re: Neon power supply + voltage multiplier producing too much voltage
« Reply #2 on: May 16, 2021, 08:08:29 AM »
Hi Davekni, thank you for your prompt response :)
I didn't know that those power supplies were designed to output constant current and dropped the voltage right after starting. So maybe the 12cm long arcs that it produced were due to the fact that the starting voltage may not be carefully controlled. Moreover, as you said, spark length isn't the best way to measure voltage.

Considering these new informations, I'm thinking that this setup is not the right way to build the x-ray power supply that I need. Would you recommend onther way to generate the high voltage needed to power a 70kV x-ray tube? (which is this one by the way: https://www.alibaba.com/product-detail/50-60-70kv-portable-xray-machine_62189076244.html?spm=a2700.9114905.0.0.oHHIuM). I'm currently thinking of two TV flybacks with the secondaries in series.

Thanks again for your help!

Offline davekni

  • High Voltage Expert
  • ******
  • Posts: 1180
  • Karma: +57/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Neon power supply + voltage multiplier producing too much voltage
« Reply #3 on: May 16, 2021, 08:56:57 PM »
Series-connected TV flyback transformers might work for a short time, but insulation between primary and base of secondary winding on the second one will fail before long.

There are several XRay projects described on this forum.  I suggest looking at those.  I believe all use ferrite-core transformers feeding CW multipliers.  In other words, like your initial design, but with AC sources of more fixed-voltage than a neon-sign supply.
« Last Edit: May 16, 2021, 09:00:17 PM by davekni »
David Knierim

Offline Anders Mikkelsen

  • Global Moderator
  • High Voltage Enthusiast
  • *****
  • Posts: 17
  • Karma: +0/-0
    • View Profile
Re: Neon power supply + voltage multiplier producing too much voltage
« Reply #4 on: May 19, 2021, 10:45:53 AM »
12 cm sparks between sharp points should be somewhere in the 80 - 90 kV range, so that makes sense. I'd strongly recommend building a basic resistive divider to measure the actual voltage, both to get reproducible results when taking x-rays, and also for the sake of safety for the tube and operator (X-ray shielding efficiency goes down drastically with increasing voltage). Note that uncompensated dividers are only useful for DC and LF measurements, to get accurate AC voltage measurements is more difficult as stray capacitance in the divider usually starts to dominate over a few hundred hertz.

Probably the loading from the x-ray tube will pull the voltage down to a more reasonable value. Make sure to power the filament before you fire up the HV supply, and regulate the filament current to get the correct anode voltage. This works because the tube will pull more anode current as you increase the filament temperature, causing the output voltage of the multiplier to drop. The current will depend on the output impedance of the SSNST and your multiplier, but this can be compensated for by adjusting the exposure time. Having a consistent voltage is the most important anyways, as this will affect the x-ray penetration and therefore image contrast.
« Last Edit: May 19, 2021, 10:47:29 AM by Anders Mikkelsen »

Offline Davide

  • High Voltage Enthusiast
  • *
  • Posts: 5
  • Karma: +0/-0
    • View Profile
Re: Neon power supply + voltage multiplier producing too much voltage
« Reply #5 on: May 24, 2021, 10:30:52 PM »
Thanks both Anders and Davekni for your answers.

Series-connected TV flyback transformers might work for a short time, but insulation between primary and base of secondary winding on the second one will fail before long.

There are several XRay projects described on this forum.  I suggest looking at those.  I believe all use ferrite-core transformers feeding CW multipliers.  In other words, like your initial design, but with AC sources of more fixed-voltage than a neon-sign supply.

Ok, I understand that the second flyback would work at a much higher voltage than what it is designed for, so probably it won't last long.
I took a look at some other projects on this forum, and I saw that "neukyhm" designed it's own transformer by winding the wire himself. Unfortunately I've never done that and I don't have the equipment to do that, so I was thinking to use a microwave oven transformer instead of the neon power supply. It sohuld generate around 2000V AC, so I would just need to add some stages to the multiplier that I already have. My only concern in this case is that a MOT could provide even too much power.

12 cm sparks between sharp points should be somewhere in the 80 - 90 kV range, so that makes sense. I'd strongly recommend building a basic resistive divider to measure the actual voltage, both to get reproducible results when taking x-rays, and also for the sake of safety for the tube and operator (X-ray shielding efficiency goes down drastically with increasing voltage). Note that uncompensated dividers are only useful for DC and LF measurements, to get accurate AC voltage measurements is more difficult as stray capacitance in the divider usually starts to dominate over a few hundred hertz.

Probably the loading from the x-ray tube will pull the voltage down to a more reasonable value. Make sure to power the filament before you fire up the HV supply, and regulate the filament current to get the correct anode voltage. This works because the tube will pull more anode current as you increase the filament temperature, causing the output voltage of the multiplier to drop. The current will depend on the output impedance of the SSNST and your multiplier, but this can be compensated for by adjusting the exposure time. Having a consistent voltage is the most important anyways, as this will affect the x-ray penetration and therefore image contrast.

Yes I bought a cheap high voltage meter from ebay, but I doesn't give realistic results. I will try building the resistive divider.
Yes probably the tube load would pull down the voltage. So, from what I understand, even using the neon power supply, if I start the filament before the HV circuit, it should give a lower voltage from the beginning. In this case I wouldn't risk to "burn" the tube. Right?

I'm trying to understand what's the most stable, reliable and safe way to generate the "moderately high" AC voltage to feed into the multiplier.

Thanks




Offline davekni

  • High Voltage Expert
  • ******
  • Posts: 1180
  • Karma: +57/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Neon power supply + voltage multiplier producing too much voltage
« Reply #6 on: May 25, 2021, 04:14:50 AM »
Quote
I've never done that and I don't have the equipment to do that, so I was thinking to use a microwave oven transformer instead of the neon power supply.
An MOT could work, or a pair in center-grounded series.  However, their line-frequency (50Hz or 60Hz) creates two problems.  First is shock hazard.  Line frequency is much more lethal than 20-25kHz.  People die with MOT mistakes, including one man here in my little town (Wilsonville, OR, USA) about 10 years ago.  Second issue is that much larger capacitors are required in the voltage multiplier.

I don't have any experience with XRay tubes.  Unless they present a relatively constant-voltage load like a neon sign does, it will be tricky to get a correct and stable output voltage from a modern 20-25kHz neon sign supply.

There are a few AC-output flyback transformers available on EBay etc.  Search here for links.  You would still need to drive it yourself, perhaps with a ZVS such as the common small induction heaters.
David Knierim

Offline Anders Mikkelsen

  • Global Moderator
  • High Voltage Enthusiast
  • *****
  • Posts: 17
  • Karma: +0/-0
    • View Profile
Re: Neon power supply + voltage multiplier producing too much voltage
« Reply #7 on: May 25, 2021, 01:58:19 PM »

Yes probably the tube load would pull down the voltage. So, from what I understand, even using the neon power supply, if I start the filament before the HV circuit, it should give a lower voltage from the beginning. In this case I wouldn't risk to "burn" the tube. Right?

I'm trying to understand what's the most stable, reliable and safe way to generate the "moderately high" AC voltage to feed into the multiplier.


Correct, as long as the filament is warm, the tube will draw current when the anode voltage is applied, keeping the voltage from rising too far.

I think you are on the right track with these solid state neon supplies. Out of curiosity I ordered some myself and I'm really impressed with the performance. They seem to output more than the marked voltage under light loading, probably to help ignite stubborn neon tubes. Even the small 3 kV variants can start an arc at around 1 cm, and the 10 kV unit easily starts an arc across a 3 cm gap.

I did a quick teardown of one of the neon supplies, the much smaller HB-C02TE. It's based on the classic self-oscillating BJT half bridge with a saturating base drive transformer, with a diac for start-up and a SCR sensing the primary current for output overvoltage protection. The circuit starts oscillating at around 45 V DC in, and runs down to 15 V before dropping out of oscillation, which is pretty impressive for such a simple circuit. Output current tracks input voltage nicely, so by feeding these from a variable input voltage you can also limit the output voltage.

High Voltage Forum

Re: Neon power supply + voltage multiplier producing too much voltage
« Reply #7 on: May 25, 2021, 01:58:19 PM »

 


* Recent Topics and Posts

post Re: The dramatic difference between wet and dry sphere splashes
[Science, Research and News In Other Fields Than Electronics]
haversin
Today at 02:49:00 AM
post Re: The dramatic difference between wet and dry sphere splashes
[Science, Research and News In Other Fields Than Electronics]
Mads Barnkob
June 16, 2021, 08:54:32 PM
post Re: The dramatic difference between wet and dry sphere splashes
[Science, Research and News In Other Fields Than Electronics]
davekni
June 16, 2021, 08:00:50 PM
post Re: The dramatic difference between wet and dry sphere splashes
[Science, Research and News In Other Fields Than Electronics]
haversin
June 16, 2021, 05:40:40 PM
post Demo Coil Protection
[Dual Resonant Solid State Tesla coils (DRSSTC)]
alan sailer
June 16, 2021, 05:39:14 PM
post Re: What is the guidence on SpaceX rockets based on ? PID's?
[General Chat]
Twospoons
June 16, 2021, 10:20:30 AM
post Re: What is the guidence on SpaceX rockets based on ? PID's?
[General Chat]
Patrick
June 16, 2021, 07:33:15 AM
post Re: Through-hole layout of UD2.7C universal driver
[Dual Resonant Solid State Tesla coils (DRSSTC)]
dbach
June 15, 2021, 01:21:38 PM
post Re: First middle sized DRSSTC - Phase lead tuning
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Rapy2
June 15, 2021, 12:02:58 PM
post Re: Through-hole layout of UD2.7C universal driver
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
June 15, 2021, 06:52:56 AM
post Re: First middle sized DRSSTC - Phase lead tuning
[Dual Resonant Solid State Tesla coils (DRSSTC)]
thedoc298
June 15, 2021, 05:43:07 AM
post Re: Through-hole layout of UD2.7C universal driver
[Dual Resonant Solid State Tesla coils (DRSSTC)]
dbach
June 15, 2021, 05:26:34 AM
post The dramatic difference between wet and dry sphere splashes
[Science, Research and News In Other Fields Than Electronics]
haversin
June 15, 2021, 02:27:03 AM
post Re: First middle sized DRSSTC - Phase lead tuning
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Rapy2
June 14, 2021, 01:06:02 AM
post Re: First middle sized DRSSTC - Phase lead tuning
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
June 14, 2021, 12:44:31 AM
post Re: First middle sized DRSSTC - Phase lead tuning
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Rapy2
June 14, 2021, 12:41:33 AM
post Re: First middle sized DRSSTC - Phase lead tuning
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Rapy2
June 14, 2021, 12:11:13 AM
post Re: First middle sized DRSSTC - Phase lead tuning
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
June 13, 2021, 11:52:11 PM
post Re: First middle sized DRSSTC - Phase lead tuning
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Rapy2
June 13, 2021, 11:38:44 PM
post Re: First middle sized DRSSTC - Phase lead tuning
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Rapy2
June 13, 2021, 10:26:51 PM
post Re: First middle sized DRSSTC - Phase lead tuning
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
June 13, 2021, 09:57:56 PM
post Re: First middle sized DRSSTC - Phase lead tuning
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Rapy2
June 13, 2021, 09:29:08 PM
post Re: First middle sized DRSSTC - Phase lead tuning
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
June 13, 2021, 08:53:25 PM
post Re: First middle sized DRSSTC - Phase lead tuning
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Rapy2
June 13, 2021, 07:22:05 PM
post Re: First middle sized DRSSTC - Phase lead tuning
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
June 13, 2021, 07:02:28 PM
post Re: HFSSTC with fixed crystal-controlled frequency
[Solid State Tesla Coils (SSTC)]
Steve Ward
June 13, 2021, 06:25:20 PM
post First middle sized DRSSTC - Phase lead tuning
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Rapy2
June 13, 2021, 02:48:59 PM
post OpenTheremin v3 loosing calibration after minutes
[Electronic Circuits]
gianpaolo
June 13, 2021, 12:00:09 PM
post Re: What is the guidence on SpaceX rockets based on ? PID's?
[General Chat]
Mads Barnkob
June 12, 2021, 07:28:03 PM
post Re: Rocket engine questions
[General Chat]
klugesmith
June 12, 2021, 07:23:12 PM
post Re: Rocket engine questions
[General Chat]
rikkitikkitavi
June 12, 2021, 12:24:06 PM
post Re: Nice PC Board Tesla
[Solid State Tesla Coils (SSTC)]
Da_Stier
June 12, 2021, 10:06:18 AM
post Re: Rocket engine questions
[General Chat]
plasma
June 12, 2021, 10:01:29 AM
post Re: Nice PC Board Tesla
[Solid State Tesla Coils (SSTC)]
Twospoons
June 12, 2021, 08:47:08 AM
post Re: What is the guidence on SpaceX rockets based on ? PID's?
[General Chat]
Patrick
June 12, 2021, 07:03:10 AM
post Re: Complications in the design of the first DRSSTC
[Dual Resonant Solid State Tesla coils (DRSSTC)]
PEDRO HENRIQUE
June 12, 2021, 06:42:56 AM
post Re: Nice PC Board Tesla
[Solid State Tesla Coils (SSTC)]
davekni
June 12, 2021, 04:20:53 AM
post Re: SKM100 Mini DRSSTC
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Willsmith001
June 11, 2021, 11:08:24 PM
post Re: How much propagation delay is too much propagation delay
[General Chat]
Willsmith001
June 11, 2021, 11:07:00 PM
post Re: Nice PC Board Tesla
[Solid State Tesla Coils (SSTC)]
Da_Stier
June 11, 2021, 12:33:55 PM
post Re: Nice PC Board Tesla
[Solid State Tesla Coils (SSTC)]
futurist
June 11, 2021, 12:03:24 PM
post Re: How much propagation delay is too much propagation delay
[General Chat]
davekni
June 11, 2021, 06:47:52 AM
post Re: How much propagation delay is too much propagation delay
[General Chat]
Power-Max
June 11, 2021, 06:32:02 AM
post Re: How much propagation delay is too much propagation delay
[General Chat]
davekni
June 11, 2021, 05:16:03 AM
post Re: SKM100 Mini DRSSTC
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
June 11, 2021, 04:53:50 AM
post Re: Why are SGT50T65FD1PN soo cheap on lcsc?
[General Chat]
Hydron
June 10, 2021, 11:51:17 PM
post Re: Rocket engine questions
[General Chat]
MRMILSTAR
June 10, 2021, 09:52:59 PM
post Re: Rocket engine questions
[General Chat]
Patrick
June 10, 2021, 09:16:25 PM
post Re: Why are SGT50T65FD1PN soo cheap on lcsc?
[General Chat]
Weston
June 10, 2021, 08:19:40 PM
post Fujifilm PMT module PMT12A, reverse engineering
[Detection, Counting and Measurements]
Mads Barnkob
June 10, 2021, 08:05:27 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