Author Topic: How much is the voltage used in Corona Discharge tubes for ozone Generation?  (Read 979 times)

Offline UB_Serpent

  • High Voltage Enthusiast
  • *
  • Posts: 7
  • Karma: +0/-0
    • View Profile
I am planning to make a corona Discharge tube mechanism for ozone generation.
Something like this.

It is basically a High voltage ferrite transformer feeding to the inside and outside of a steel tube, separated by quartz glass tube inside. Corona discharge happens inside the tube and the oxygen passing the tube is converted to ozone.
The electronic circuit I can understand, it is a half bridge driving the transformer from 320VDC at 25 kHz.
So I am looking for information to design and construct this transformer.
Thank you

Offline johnf

  • High Voltage Technician
  • ***
  • Posts: 135
  • Karma: +4/-0
    • View Profile
Corona discharge goes up non linearly with voltage
in normal circumstances there is little corona until you are over 20kV although very sharp points or edges will get significant amounts at 10kV
What I mean by sharp is the point or edge being a few atoms across ie atomically sharp. Electrostatic precipitators which use corona to launch electrons typically use 40 - 140kV
Most of us try to minimise corona by using smooth rounded edges and spheres to control the electric field so as to keep low gradients

Offline UB_Serpent

  • High Voltage Enthusiast
  • *
  • Posts: 7
  • Karma: +0/-0
    • View Profile
in normal circumstances there is little corona until you are over 20kV although very sharp points or edges will get significant amounts at 10kV

Are you suggesting that the tube I have shown above should be operating at above 20 kV? if So how much will be a rough approximation of the current in that voltage needed to maintain a ARC over the area inside the tube. The transformer you can see in the picture.
Even rough approximations will be helpful  for me to reach a starting point.

Thank you

Online davekni

  • High Voltage Senior
  • *****
  • Posts: 556
  • Karma: +31/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
For ozone generation, you want corona discharge, not arcs.  That's why the quartz glass tube exists - to prevent arcs between the inner and outer electrodes.

Voltage depends on geometry.  Field strength (volts/meter) is more important than absolute voltage.  I made an small ozone-generator from a glass test-tube with aluminum-foil coating the inside and metal screen on the outside.  The screen was spaced ~1mm from the glass, and a fan blew air across the tube/screen assembly.  5kV at 30kHz and 2-3mA was plenty for that.  Your tube looks much larger, so will need higher voltage, perhaps 20-30kV, and more current.  To make any reasonable guess, the dimensions of the tube would be needed, especially the cross-section (ID and OD of the glass, OD of the inner electrode, and ID of the outer electrode).
David Knierim

Offline UB_Serpent

  • High Voltage Enthusiast
  • *
  • Posts: 7
  • Karma: +0/-0
    • View Profile
For ozone generation, you want corona discharge, not arcs.  That's why the quartz glass tube exists - to prevent arcs between the inner and outer electrodes.
Thanks for the clarification.
To make any reasonable guess, the dimensions of the tube would be needed, especially the cross-section (ID and OD of the glass, OD of the inner electrode, and ID of the outer electrode).
I am attaching a photograph from the tube I collected from junk market.
The glass tube ID is 37mm. The inner electrode is a  thin Stainless steel sheet rolled inside the glass tube. The glass is 1.5 mm thick and outside the glass there is around 4 mm gap ( cannot measure properly) through which oxygen flows. Then comes the steel outer tube with heat-sink to dissipate heat. One side of the transformer secondary is connected to the outer steel pipe and other end to the inside stainless steel foil/thin sheet.

Thanks
« Last Edit: August 01, 2020, 06:25:22 AM by UB_Serpent »

Online klugesmith

  • High Voltage Engineer
  • ****
  • Posts: 269
  • Karma: +6/-0
    • View Profile
Where are the sharp metal points, edges, or very thin wires in the space for air to be ozonated?
I may have found the answer, with this picture, at https://www.oxidationtech.com/ozone/ozone-production/corona-discharge.html


That page says that it takes 0.82 kWh to generate 1 kg of ozone, times 10 or 20 because of inefficiencies.
Don't put too much trust in that, without independent verification.
The Internet has a lot of misinformation and poorly dumbed-down details about ozone, corona discharges, and corona viruses.

Commercial ozonators usually specify the O3 production rate in grams per hour.  Does anyone here know of a practical chemical assay to measure that?  For example, bubbling the gas through some simple reagent until there is a color change?
« Last Edit: August 01, 2020, 04:04:49 PM by klugesmith »

Offline TMaxElectronics

  • High Voltage Experimenter
  • **
  • Posts: 88
  • Karma: +3/-0
    • View Profile
    • My random (and very empty) electronics blog
Quote
Does anyone here know of a practical chemical assay to measure that?

It is fairly easy to titrate ozone, using a solution of potassium iodide, like in this video:

He tested his 3 g/h ozone tube from ebay, and it only produced like 300 mg/h. So those ratings really can't be trusted, unless it is from a trustworthy brand, purchased through a trustworthy source.

Online davekni

  • High Voltage Senior
  • *****
  • Posts: 556
  • Karma: +31/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Somewhat like the little test-tube version I made, except that my outer electrode was screen with air blowing across it.

Presuming 3.7 dielectric constant for the glass, 1.5mm is roughly the same voltage drop as 0.4mm of air, or about 10% of applied voltage below corona threshold.  Breakdown voltage for dry air is around 3.4MV/m, for about 15kV peak across 4.4mm, just under 11kV RMS.  If there are sharp points inside the outer electrode, corona should start at less than 11kV.  For significant total power, that tube is likely intended for 15-20kV RMS.

As you've probably read, commercial ozone generators are typically fed 90+% oxygen from an oxygen-concentrator.  That's the case for our (Wilsonville, Oregon, USA) water treatment plant.  The ozone generators at the plant output about 12% ozone from a 95% oxygen feed.  The monitoring meters fluctuated from 11-13% when I was there on a tour.

I wonder in the titration video you showed, how much of the ozone is captured as it bubbles through the solution.  I'd think that accurate ozone measurement would require finer bubbles rising through a taller column of liquid.  (His EBay-purchased tube may have been rated for oxygen feed.  Using air may be another reason for his low measurement.)

Hopefully you are aware of the hazards of breathing ozone.  Judging by smell isn't safe, as people vary by 1000:1 in ability to smell ozone.  Smell also gets desensitized from ontoing exposure, so it's easy to get used to slowly increasing concentration without realizing it's becoming damaging to your lung tisue.
David Knierim

Offline UB_Serpent

  • High Voltage Enthusiast
  • *
  • Posts: 7
  • Karma: +0/-0
    • View Profile
Presuming 3.7 dielectric constant for the glass, 1.5mm is roughly the same voltage drop as 0.4mm of air, or about 10% of applied voltage below corona threshold.  Breakdown voltage for dry air is around 3.4MV/m, for about 15kV peak across 4.4mm, just under 11kV RMS.  If there are sharp points inside the outer electrode, corona should start at less than 11kV.  For significant total power, that tube is likely intended for 15-20kV RMS.
Thanks for the approximations. Definitely that is a starting point for me. Since my electrode is not having mesh like structure I may need good voltage. Also because of the tube length, current capacity may also be needed more. Can the TV fly-back core or other Big U cores take power above 2000Watts?

Also thanks for reminding about the health hazard, will take maximum care about that.

Offline UB_Serpent

  • High Voltage Enthusiast
  • *
  • Posts: 7
  • Karma: +0/-0
    • View Profile
For example, bubbling the gas through some simple reagent until there is a color change?
Here in India, we have a Fabric whitener (used to provide light bluish tint to white cloths) Named UJALA http://www.jyothylaboratories.com/ujala-supreme.php. we pour few drops of this into water and if we bubble ozone, depending on the strength of ozone, it will immediately clear the water. That gives us a rough approximation.

Thanks

Offline Uspring

  • High Voltage Technician
  • ***
  • Posts: 182
  • Karma: +14/-0
  • Physicist
    • View Profile
You'll definitely need tubes with sharp points. Smooth tubes in your geometry will generate nearly homogeneous fields, which either arc over or not conduct at all. The sharp points will cause lokal strong fields causing lokal breakdowns around them but no arcs if not driven too hard.

Another thing to watch is, that the glass is an insulator, so conducting current through it requires AC voltages. You can calculate the required frequency from the capacitance of the tube, i.e. area, thickness and dielelectric constant. You'll probably need a few kHz, DC and line frequency won't work.

Online davekni

  • High Voltage Senior
  • *****
  • Posts: 556
  • Karma: +31/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
"You'll definitely need tubes with sharp points. Smooth tubes in your geometry will generate nearly homogeneous fields, which either arc over or not conduct at all."

No, that's the reason for the glass.  It prevents arcs.  It's easy to get corona using AC between smooth electrodes separated by an insulator such as glass or mica that can withstand corona.  I've done that myself experimenting with etching polyimide using corona discharge.

Not sure how long your tube is, but I estimate 130pF for 500mm length.  A resonant inverter will be best for driving that load.  TV flyback cores are too small for 2kW, but there are larger U cores available.  This may be larger than necessary, but larger is often better for home-built transformers:
https://www.ebay.com/itm/LARGE-FERRITE-U-CORE-1-X-SQUARE-BY-2-1-4-BY-4-OUTSIDE-LOT-OF-2-PIECES/264655191280?hash=item3d9eadbcf0:g:kwUAAOSwtANeXmrP#shpCntId
That's a US listing.  If you live in Europe, a smaller (but still more than typical TV flyback) core:
https://www.ebay.com/itm/Ferrit-Trafo-Kern-UR57-28-16-3C30-Ferroxcube-UR-Cores-2x/313035982357?hash=item48e2660615:g:kboAAOSwTXVeeIRc

For making your own HV transformer, explore some of the threads on this forum about vacuum-potting, or immerse the coil in oil when operating.
David Knierim

Offline UB_Serpent

  • High Voltage Enthusiast
  • *
  • Posts: 7
  • Karma: +0/-0
    • View Profile
That's a US listing.  If you live in Europe, a smaller (but still more than typical TV flyback) core:

I am from India. I am looking in our E bay. Big ones are not easily available in small quantities. But will collect soon.
and btw, what about the calculations for ferrite transformer in these voltages? Same as normal SMPS transformer calculations?
Thanks

Offline johnf

  • High Voltage Technician
  • ***
  • Posts: 135
  • Karma: +4/-0
    • View Profile
You could try to find a bust panasonic inverter microwave the transformer core is good for 1200 watts and if you are careful the bobbin already is a high voltage bobbin try to leave the primary intact to rewind back on the core it is the multi-stranded one.

Offline Uspring

  • High Voltage Technician
  • ***
  • Posts: 182
  • Karma: +14/-0
  • Physicist
    • View Profile
Quote
No, that's the reason for the glass.  It prevents arcs.

I had been thinking of this  :)



The electrode on the right is a grounded rod with a test tube stuck on it. But you're right. If voltage and frequency aren't too large, there won't be any arcs.
« Last Edit: August 05, 2020, 12:59:21 PM by Uspring »

Online davekni

  • High Voltage Senior
  • *****
  • Posts: 556
  • Karma: +31/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Uspring, In particular, insulation (glass tube etc.) is useful when the corona discharge distance is short compared to electrode area.  With a long distance, there's room for the charge to collect together into an arc path.  If it's possible to dial back your coil voltage enough to avoid dielectric breakdown, laying the test tube side adjacent the top load should show just corona.  Even better if you can line the inside of the tube with foil, perhaps foil over some rolled-up card-stock or mylar or ... to help keep the foil pressed against the inside of the test tube.

Concerning ferrite core size, it will likely need to be larger than one for an SMPS of the same power.  Two reasons:  First is because the ozone-generator is a highly-capacitive load, so the transformer will have relatively-large resonant current.  Second is that home-wound transformers are generally not as efficient as a design-optimized and machine-wound commercial transformer.  If you can find a inverter microwave oven transformer as John suggested, that might be a good starting point.  However, that may not support enough resonant current to get to 20kV RMS if you are after high-power use.

Edit:  Threw together a quick demo using a 61mm ID 65mm OD (large) test tube with a piece of foil inside and another propped 1mm (at closest point) off the outside.  Applied 6.5kV RMS at 40kHz from a ZVS oscillater and transformer I'd made a couple years ago:


With a 2mm air gap above the 2mm-thick glass, 6.5kV wasn't quite enough.  At 1mm it worked well, looking like the corona extended to at least 1.5mm gap.  For size perspective, the upper foil piece is 40mm wide.  (The foil piece inside is ~50 x 50mm, with a piece of thin black plastic inside pressing it against the glass.)
« Last Edit: August 06, 2020, 04:00:42 AM by davekni »
David Knierim

Offline UB_Serpent

  • High Voltage Enthusiast
  • *
  • Posts: 7
  • Karma: +0/-0
    • View Profile

Concerning ferrite core size, it will likely need to be larger than one for an SMPS of the same power.  Two reasons:  First is because the ozone-generator is a highly-capacitive load, so the transformer will have relatively-large resonant current.  Second is that home-wound transformers are generally not as efficient as a design-optimized and machine-wound commercial transformer.  If you can find a inverter microwave oven transformer as John suggested, that might be a good starting point.  However, that may not support enough resonant current to get to 20kV RMS if you are after high-power use.

With a 2mm air gap above the 2mm-thick glass, 6.5kV wasn't quite enough.  At 1mm it worked well, looking like the corona extended to at least 1.5mm gap.  For size perspective, the upper foil piece is 40mm wide.  (The foil piece inside is ~50 x 50mm, with a piece of thin black plastic inside pressing it against the glass.)

Thank you. Definitely that is a starting point.

High Voltage Forum


 


* Recent Topics and Posts

post Re: Polyphonic Interrupter with STM32F103C8T6 (Bluepill)
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Max
September 28, 2020, 11:48:25 PM
post Re: Half bridge sstc help with troubleshoot
[Solid State Tesla Coils (SSTC)]
costas_p
September 28, 2020, 09:56:15 PM
post Re: Easy to build Polyphonic MIDI Interrupter
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
futurist
September 28, 2020, 09:52:47 PM
post Re: Half bridge sstc help with troubleshoot
[Solid State Tesla Coils (SSTC)]
davekni
September 28, 2020, 09:50:44 PM
post Re: Half bridge sstc help with troubleshoot
[Solid State Tesla Coils (SSTC)]
costas_p
September 28, 2020, 09:16:53 PM
post Re: Half bridge sstc help with troubleshoot
[Solid State Tesla Coils (SSTC)]
Zipdox
September 28, 2020, 09:09:46 PM
post Re: Half bridge sstc help with troubleshoot
[Solid State Tesla Coils (SSTC)]
costas_p
September 28, 2020, 09:03:23 PM
post Re: Half bridge sstc help with troubleshoot
[Solid State Tesla Coils (SSTC)]
Zipdox
September 28, 2020, 08:50:33 PM
post Re: Stroboscopic photography & motors
[Digital Compact Camera]
Zipdox
September 28, 2020, 08:49:09 PM
post Re: Half bridge sstc help with troubleshoot
[Solid State Tesla Coils (SSTC)]
costas_p
September 28, 2020, 08:40:11 PM
post Re: Stroboscopic photography & motors
[Digital Compact Camera]
klugesmith
September 28, 2020, 07:36:45 PM
post Re: DRSSTC Tuning for Music vs Big Sparks?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Uspring
September 28, 2020, 07:07:29 PM
post Re: Stroboscopic photography & motors
[Digital Compact Camera]
Zipdox
September 28, 2020, 06:41:56 PM
post Re: Polyphonic Interrupter with STM32F103C8T6 (Bluepill)
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Zipdox
September 28, 2020, 06:28:50 PM
post Re: Stroboscopic photography & motors
[Digital Compact Camera]
klugesmith
September 28, 2020, 05:14:51 PM
post Re: Polyphonic Interrupter with STM32F103C8T6 (Bluepill)
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Max
September 28, 2020, 04:44:56 PM
post Re: Stroboscopic photography & motors
[Digital Compact Camera]
TMaxElectronics
September 28, 2020, 10:50:31 AM
post Re: Stroboscopic photography & motors
[Digital Compact Camera]
plasma
September 28, 2020, 09:15:09 AM
post Re: Stroboscopic photography & motors
[Digital Compact Camera]
Zipdox
September 28, 2020, 09:09:45 AM
post Re: type of driver
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mads Barnkob
September 28, 2020, 08:39:30 AM
post Re: Stroboscopic photography & motors
[Digital Compact Camera]
klugesmith
September 28, 2020, 08:16:46 AM
post Re: Polyphonic Interrupter with STM32F103C8T6 (Bluepill)
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Zipdox
September 28, 2020, 01:27:07 AM
post Re: Polyphonic Interrupter with STM32F103C8T6 (Bluepill)
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
TMaxElectronics
September 28, 2020, 01:14:46 AM
post Re: Rogowski current probe
[Laboratories, Equipment and Tools]
flyglas
September 28, 2020, 12:30:58 AM
post Re: Rogowski current probe
[Laboratories, Equipment and Tools]
Hydron
September 27, 2020, 11:48:35 PM
post Polyphonic Interrupter with STM32F103C8T6 (Bluepill)
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Zipdox
September 27, 2020, 11:43:14 PM
post Re: Rogowski current probe
[Laboratories, Equipment and Tools]
flyglas
September 27, 2020, 11:01:13 PM
post type of driver
[Dual Resonant Solid State Tesla coils (DRSSTC)]
octopus1
September 27, 2020, 10:51:03 PM
post Re: Easy to build Polyphonic MIDI Interrupter
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Max
September 27, 2020, 10:44:43 PM
post Re: MIDI videos from my DRSSTC1/2/3 demonstrations/shows
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mads Barnkob
September 27, 2020, 09:44:58 PM
post Re: DRSSTC Tuning for Music vs Big Sparks?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
September 27, 2020, 09:20:32 PM
post Re: Easy to build Polyphonic MIDI Interrupter
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
futurist
September 27, 2020, 07:49:49 PM
post Stroboscopic photography & motors
[Digital Compact Camera]
klugesmith
September 27, 2020, 07:08:06 PM
post Re: Led Closeups with Mobile phone and lens .
[Smart Phones]
klugesmith
September 27, 2020, 06:43:46 PM
post Re: DRSSTC Tuning for Music vs Big Sparks?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mads Barnkob
September 27, 2020, 06:42:46 PM
post Re: Finding accelerometer location inside a phone
[Smart Phones]
klugesmith
September 27, 2020, 06:36:41 PM
post Re: steam engine
[Capacitor Banks]
plasma
September 27, 2020, 02:32:14 PM
post Re: Easy to build Polyphonic MIDI Interrupter
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Max
September 27, 2020, 02:02:53 PM
post Re: Easy to build Polyphonic MIDI Interrupter
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
futurist
September 27, 2020, 12:24:20 PM
post Re: Easy to build Polyphonic MIDI Interrupter
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Max
September 27, 2020, 12:52:36 AM
post Re: DRSSTC Tuning for Music vs Big Sparks?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
September 26, 2020, 11:06:04 PM
post Re: One NOS ac flyback is nice. Finding its identical twin 3 months later is awesome
[Transformer (Ferrite Core)]
profdc9
September 26, 2020, 10:48:41 PM
post Re: Rogowski current probe
[Laboratories, Equipment and Tools]
Hydron
September 26, 2020, 05:27:37 PM
post Re: DRSSTC Tuning for Music vs Big Sparks?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Zipdox
September 26, 2020, 04:32:53 PM
post Re: DRSSTC Tuning for Music vs Big Sparks?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mads Barnkob
September 26, 2020, 04:17:51 PM
post Re: DRSSTC Tuning for Music vs Big Sparks?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
GKnapp
September 26, 2020, 02:06:04 PM
post Re: Rogowski current probe
[Laboratories, Equipment and Tools]
flyglas
September 26, 2020, 10:57:58 AM
post Re: DRSSTC Tuning for Music vs Big Sparks?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mads Barnkob
September 26, 2020, 08:55:27 AM
post Re: DRSSTC Tuning for Music vs Big Sparks?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
September 26, 2020, 05:19:45 AM
post Re: Rogowski current probe
[Laboratories, Equipment and Tools]
klugesmith
September 26, 2020, 04:46:12 AM

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