Hi Calvin,
Looks like you have not tuned the primary and secondary of your AC flyback, hence the secondary voltage may be low.
You want to tune these two, but not too much - with a high Q you may have high reflected voltage on your primary side that will fry the ZVS mosfets. This will also depend on the air gap in your ac flyback.
If you have tools to measure the inductance & capacitance of your flyback primary and secondary inductance and capacitance, then try to match the secondary resonance frequence (not exactly, but say 3-5 khz off) with your primary winding - number of turns.
In my case the electrode was 250mm long with about 2-2.5mm gap and capacitance was about 40pF.
The smaller ZVS drivers usually have a total of 0.6 or 0.66uF (two 0.3 or 0.33uF capacitors). I had to add another 1uF capacitor (big white Cornell Dubilier) to get within 5kHz of the secondary resonace (with 8 primary turns of 18AWG).
Provide a 0.3-0.5mm air gap between the hot and grounded electrodes to seen the “purple corona glow”
Be careful and good luck.
Regards
Zvon
Hi,
thanks alot for those valuable tips. 😉
Since the treatment itself worked so well already it might not be worth the effort of tuning.
I'll see if I investigate any further.
jauu
Calvin
thanks alot for those valuable tips. 😉
Since the treatment itself worked so well already it might not be worth the effort of tuning.
I'll see if I investigate any further.
jauu
Calvin
Happy New Year and all the Best in 2023 to all DIY-ers.
I have finally managed to finalise this project and in this message, I'll provide the (previously missing) data and construction details.
The previous iteration worked quite well but (in my opinion) had a couple of issues :
1) Borosilicate glass is a great insulator, however, it is fragile
2) The HV electrode inside the borosilicate tube gets quite hot after a couple of minutes of continuous use
To address the above problems and finish this project, the following changes have been made:
A) HV electrode is now a 1/2" copper pipe (plumbing) sleeved inside a silicone tube with 3mm thick walls
B) 5015 blower fan is introduced in the corona stick head to cool inside of the copper pipe(s)
The additional air cooling has eliminated the HV electrode-heating problem.
Even after several minutes of operation, the electrode temperature is lower than 45 degC.
The FLIR image below shows the PCB sheet being hotter than the HV electrode below.
With 3mm silicone insulation, I need 24DC and approximately 5A for a stable and uniform corona discharge.
The HV output is 10.8kV at approximately 20.8kHz. (finally managed to build that HV scope probe 🙂 )
The corona driver details are provided in the paragraph and image below.
The final version of the corona stick
Final testing
Also, the corona HV electrode is now replaceable and I will make another head with a blower and Jacobs ladder-type spark generator (previously mentioned by Calvin). That should be quite useful for the activation of irregularly shaped plastic objects that require glueing (e.g. ESL63 PCB to eggcrate).
I am very happy with the corona treater operation, functionality and the ease of ESL resistive coating application now.
Regards and good luck if you decide to build this.
In my final prototype the "hot wiring" is fully enclosed, and cannot be touched, however, bear in mind that this is a dangerous high-voltage device that can easily push more than 15mA @ 7.8 kV RMS and this is lethal ☠️ ⚰️💀
I have finally managed to finalise this project and in this message, I'll provide the (previously missing) data and construction details.
The previous iteration worked quite well but (in my opinion) had a couple of issues :
1) Borosilicate glass is a great insulator, however, it is fragile
2) The HV electrode inside the borosilicate tube gets quite hot after a couple of minutes of continuous use
To address the above problems and finish this project, the following changes have been made:
A) HV electrode is now a 1/2" copper pipe (plumbing) sleeved inside a silicone tube with 3mm thick walls
B) 5015 blower fan is introduced in the corona stick head to cool inside of the copper pipe(s)
The additional air cooling has eliminated the HV electrode-heating problem.
Even after several minutes of operation, the electrode temperature is lower than 45 degC.
The FLIR image below shows the PCB sheet being hotter than the HV electrode below.
With 3mm silicone insulation, I need 24DC and approximately 5A for a stable and uniform corona discharge.
The HV output is 10.8kV at approximately 20.8kHz. (finally managed to build that HV scope probe 🙂 )
The corona driver details are provided in the paragraph and image below.
- Input 24 V DC @ 5A
- ZVS Driver from eBay or Amazon (use the one without the inductor centre tap)
- Additional 1uF film polypropylene capacitor across the inductor to lower the resonant frequency
- AC flyback transformer (must not use DC flybacks with built-in diodes) FLYTCL250 from https://www.amazing1.com/transformers-high-voltage-high-frequency.html should work
The final version of the corona stick
Final testing
Also, the corona HV electrode is now replaceable and I will make another head with a blower and Jacobs ladder-type spark generator (previously mentioned by Calvin). That should be quite useful for the activation of irregularly shaped plastic objects that require glueing (e.g. ESL63 PCB to eggcrate).
I am very happy with the corona treater operation, functionality and the ease of ESL resistive coating application now.
Regards and good luck if you decide to build this.
In my final prototype the "hot wiring" is fully enclosed, and cannot be touched, however, bear in mind that this is a dangerous high-voltage device that can easily push more than 15mA @ 7.8 kV RMS and this is lethal ☠️ ⚰️💀
Hi,
great job 👍
I haven´t proceeded with my implementation, as I rather tend towards the blown cold plasma that one can use on any shape and with already mounted diaphragm. I hope that the cleaning action of the plasma will be helpful after handling and glueing the membrane and just ahead of the coating procedure.
jauu
Calvin
great job 👍
I haven´t proceeded with my implementation, as I rather tend towards the blown cold plasma that one can use on any shape and with already mounted diaphragm. I hope that the cleaning action of the plasma will be helpful after handling and glueing the membrane and just ahead of the coating procedure.
jauu
Calvin
very nice ! does it make the coating stick so much better >? creating a lower surface tension ? Nice pictures and construction !! very interesting !!!
Hi Calvin and Joppe,
Thank you for the kind comments.
Calvin, I will make a corona stick head with cold-blown plasma, which is just Jacob's ladder structure in a box with forced cooling/blower, but don't
hold your breath for it.
Joppe, I'm following your YouTube channel about MFB woofer/subwoofer with great interest. A long time ago I had an opportunity to listen to a pair of the original Philips MFB speakers (MH???) and they produced impressively clean and extended bass. I'm hoping that I'll have time to build a speaker with a segmented ESL and MFB woofer one day.
Regards
Zvon
Thank you for the kind comments.
Calvin, I will make a corona stick head with cold-blown plasma, which is just Jacob's ladder structure in a box with forced cooling/blower, but don't
hold your breath for it.
Joppe, I'm following your YouTube channel about MFB woofer/subwoofer with great interest. A long time ago I had an opportunity to listen to a pair of the original Philips MFB speakers (MH???) and they produced impressively clean and extended bass. I'm hoping that I'll have time to build a speaker with a segmented ESL and MFB woofer one day.
Regards
Zvon
Сongrats. Not sure if you need to cool electrode per say. It probably helps electronics rather than piece of copper. Nice solution though.
Hi Alex, thanks for the nice words.
The pipe has a plug inside with a 3mm hole for the HV cable and all fan airflow is directed through the copper pipe.
The copper electrode and silicone sleeve have an elliptic hole in the middle intersection to accommodate the airflow.
The copper electrode is fixed to 3D printed head by 2 x M4 machine bolts that are recessed into the print head.
I have 3mm thick press-fit hole plugs to seal the holes and bolt heads that are @ HV.
This cooling is indeed quite effective but I have to rant a little about the ZVS.
It appears the majority of electronic circuits/boards sold on Amazon, eBay & AliExpress are over-specified.
The ZVS driver I purchased from eBay is specified at max 30V and 1000W which (at least in my books) is unrealistic.
ZVS has two IRFP064N MOSFETs rated 200W each. I know that we use zero switching, but 30A @ 30V is outside SOA IR datasheet.
Furthermore, the toroidal inductor copper wire is 1mm dia which is only good for 16-19A short equipment wiring.
This inductor wiring is now the hottest component on the PCB.
But then, when you look at the eBay seller's other sale items there are plastic orange peelers, nylon soap net, etc 😉
What I'm trying to say is that I purchase these items because they are cheap.
For the eBay purchase price, I cannot buy even passive components from the local RS, let alone MOSFETS, inductors and PCB.
But then on the other side, I do not rely on the advertised specs and use common sense and basic knowledge to de-rate things.
The pipe has a plug inside with a 3mm hole for the HV cable and all fan airflow is directed through the copper pipe.
The copper electrode and silicone sleeve have an elliptic hole in the middle intersection to accommodate the airflow.
The copper electrode is fixed to 3D printed head by 2 x M4 machine bolts that are recessed into the print head.
I have 3mm thick press-fit hole plugs to seal the holes and bolt heads that are @ HV.
This cooling is indeed quite effective but I have to rant a little about the ZVS.
It appears the majority of electronic circuits/boards sold on Amazon, eBay & AliExpress are over-specified.
The ZVS driver I purchased from eBay is specified at max 30V and 1000W which (at least in my books) is unrealistic.
ZVS has two IRFP064N MOSFETs rated 200W each. I know that we use zero switching, but 30A @ 30V is outside SOA IR datasheet.
Furthermore, the toroidal inductor copper wire is 1mm dia which is only good for 16-19A short equipment wiring.
This inductor wiring is now the hottest component on the PCB.
But then, when you look at the eBay seller's other sale items there are plastic orange peelers, nylon soap net, etc 😉
What I'm trying to say is that I purchase these items because they are cheap.
For the eBay purchase price, I cannot buy even passive components from the local RS, let alone MOSFETS, inductors and PCB.
But then on the other side, I do not rely on the advertised specs and use common sense and basic knowledge to de-rate things.
It's not like that. You need to know trajectory in terms U-I and duration of a transition per say. Assuming the latter is about 50 may be 100 nS you are more than fine.
Anyhow glad it works for you, assuming you're walking upside down. Cheers!
Hi Alex, actually you are right!
I have probably looked at the datasheet upside down 😁. I have not considered ZVS nature of the circuit, and looked only at the SOA diagram using 30V and 30A (which is wrong 😂).
It’s really good to know that we have a pair of knowledgeable eyes reviewing our posts👍.
Regards
Zvon
I have probably looked at the datasheet upside down 😁. I have not considered ZVS nature of the circuit, and looked only at the SOA diagram using 30V and 30A (which is wrong 😂).
It’s really good to know that we have a pair of knowledgeable eyes reviewing our posts👍.
Regards
Zvon
I’m reviving this long-time closed thread to respond to Calvin’s post number #45.
Yes, cold-blown plasma should be more convenient, as it can be used with almost any substrate shape.
To confirm it is doable, I have 3d printed a different cold plasma head - quick and dirty, just to prove the concept. In a nutshell, it works quite well. Refer to my attached video clip - I used the driver for my mylar film treater. The prototype is built based on the concept described on this site:
Cold Plasma Blower
Now I don’t have (any additional) excuse(s) not to repair my old ESL63 😂
Regards,
Zvon
Yes, cold-blown plasma should be more convenient, as it can be used with almost any substrate shape.
To confirm it is doable, I have 3d printed a different cold plasma head - quick and dirty, just to prove the concept. In a nutshell, it works quite well. Refer to my attached video clip - I used the driver for my mylar film treater. The prototype is built based on the concept described on this site:
Cold Plasma Blower
Now I don’t have (any additional) excuse(s) not to repair my old ESL63 😂
Regards,
Zvon
Abrasive prep is good. Sanding doesn’t activate the plastic surface.
I lightly abrade plastic stator stiffeners before treatment and glueing to improve adhesion.
Chemical cleaning is good.
I clean films with IPA before treatment.
Chemical surface activation is usually associated with nasty chemicals.
Plasma cleans and activates (plastic) surfaces to improve adhesion and wettability, but you are dealing with dangerous HV.
Take your pick. 😃
I lightly abrade plastic stator stiffeners before treatment and glueing to improve adhesion.
Chemical cleaning is good.
I clean films with IPA before treatment.
Chemical surface activation is usually associated with nasty chemicals.
Plasma cleans and activates (plastic) surfaces to improve adhesion and wettability, but you are dealing with dangerous HV.
Take your pick. 😃