Author Topic: 9kV CW multiplier potting process  (Read 1445 times)

0 Members and 1 Guest are viewing this topic.

Offline doktor pytaTopic starter

  • Frequent Contributor
  • **
  • Posts: 492
  • Country: pl
9kV CW multiplier potting process
« on: June 09, 2020, 10:00:51 pm »
I've been asked to share some details of potting (moderately) high voltage assembly.
So below You will find some subsequent stages and related photos.
Sorry for not documenting all of the steps.
If someone have any suggestions please share.

Procedure:

1. reflow soldering of components
2. attaching wires
3. ultrasound cleaning
4. drying
5. HV testing before potting (minimize corona discharge!)
6. preparing special mould (I'm proud especially of this design from 2012)
7. covering internals of the mould with 'release agent' (seems to be more less stearine in naphta)
8. putting HV assemblies into the mould
9. mixing two-component silicone
10. degassing under vacuum (approx 30s)
11. careful potting
12. second degassing (approx 30s) to get rid of air trapped between components and PCB
13. 24h curing at room temperature under atmospheric pressure
14. HV testing after potting
« Last Edit: June 09, 2020, 11:09:23 pm by doktor pyta »
 

Offline doktor pytaTopic starter

  • Frequent Contributor
  • **
  • Posts: 492
  • Country: pl
Re: 9kV CW multiplier potting process
« Reply #1 on: June 09, 2020, 10:02:25 pm »
p
« Last Edit: June 09, 2020, 10:07:36 pm by doktor pyta »
 

Offline doktor pytaTopic starter

  • Frequent Contributor
  • **
  • Posts: 492
  • Country: pl
Re: 9kV CW multiplier potting process
« Reply #2 on: June 09, 2020, 10:04:27 pm »
p
« Last Edit: June 09, 2020, 10:07:04 pm by doktor pyta »
 

Offline doktor pytaTopic starter

  • Frequent Contributor
  • **
  • Posts: 492
  • Country: pl
Re: 9kV CW multiplier potting process
« Reply #3 on: June 09, 2020, 10:09:20 pm »
p

Online coppercone2

  • Super Contributor
  • ***
  • Posts: 11342
  • Country: us
  • $
Re: 9kV CW multiplier potting process
« Reply #4 on: June 09, 2020, 10:11:59 pm »
depending on the viscosity of your epoxy, you might want to mix it under vacuum. need specialized equipment. If its anything more then a light syrup, you only get the best results if you mix it under vacuum or mix it in a nozzle. If viscosity is low, there is no benefit.

If you can't do this, I recommend getting a blank epoxy cartridge with a mixing nozzle and refill it with silicone epoxy.

What is your epoxy type? I am interested. I assume its some kind of epoxy silicone.

Also, for complex things, you need to do a vacuum pour, if you pour it in and pull a vacuum, you can make situation worse. But this depends all on viscosity. This is particularly relevant for inductor/transformer winding's.
« Last Edit: June 09, 2020, 10:14:51 pm by coppercone2 »
 

Offline doktor pytaTopic starter

  • Frequent Contributor
  • **
  • Posts: 492
  • Country: pl
Re: 9kV CW multiplier potting process
« Reply #5 on: June 09, 2020, 10:16:53 pm »

Online coppercone2

  • Super Contributor
  • ***
  • Posts: 11342
  • Country: us
  • $
Re: 9kV CW multiplier potting process
« Reply #6 on: June 09, 2020, 10:18:47 pm »
https://abbess.com/wp-content/uploads/2016/03/Degassing-Mixing-V2-Secure.pdf

you need to control pressure. check the tables to see what your procedure should be.

Based on your viscosity (2000 Cps), you can get away with atmospheric mixing and degas (so long pressure is controlled), and vacuum mold filling is either recommended or necessary depending on mold geometry.

Read page 4.
« Last Edit: June 09, 2020, 10:21:03 pm by coppercone2 »
 
The following users thanked this post: doktor pyta

Online coppercone2

  • Super Contributor
  • ***
  • Posts: 11342
  • Country: us
  • $
Re: 9kV CW multiplier potting process
« Reply #7 on: June 09, 2020, 10:23:37 pm »
also be careful of ultrasonic cleaning, you should do a flush with clean solvent and bake it afterwards I think.

Keep in mind any bubbles will cause corona damage possibly. I say this because HV potting is different then say, mechanical potting, which I think is less demanding, and there is alot of great info on vibration hardening that might not be as useful for reliable HV
« Last Edit: June 09, 2020, 11:22:17 pm by coppercone2 »
 

Offline ahbushnell

  • Frequent Contributor
  • **
  • Posts: 751
  • Country: us
Re: 9kV CW multiplier potting process
« Reply #8 on: June 09, 2020, 10:54:41 pm »
depending on the viscosity of your epoxy, you might want to mix it under vacuum. need specialized equipment. If its anything more then a light syrup, you only get the best results if you mix it under vacuum or mix it in a nozzle. If viscosity is low, there is no benefit.

If you can't do this, I recommend getting a blank epoxy cartridge with a mixing nozzle and refill it with silicone epoxy.

What is your epoxy type? I am interested. I assume its some kind of epoxy silicone.

Also, for complex things, you need to do a vacuum pour, if you pour it in and pull a vacuum, you can make situation worse. But this depends all on viscosity. This is particularly relevant for inductor/transformer winding's.
Filling under vacuum works very nice.  No bubbles introduced when pouring. 
 

Offline texaspyro

  • Super Contributor
  • ***
  • Posts: 1407
Re: 9kV CW multiplier potting process
« Reply #9 on: June 11, 2020, 04:39:54 am »
When making composite rocket propellant. it is common practice to add a few drops per kg of thin silicone oil to help minimize bubbles and voids.  Typically RC car "shock oil" is used.  Shock oil is used to fill things like shock absorbers and steering dampers.   You want to use the thinnest version you can find.   The binder used in most propellent formulations is HTPB...  basically a polyurethane.
 

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 7681
  • Country: ca
Re: 9kV CW multiplier potting process
« Reply #10 on: June 11, 2020, 06:26:41 am »
I add PC board holes and slots to ensure the potting compound flows in nooks and under parts. It also helps the compound bond to the board and not peel off a side, as a big slab.
I get most air bubbles added during mixing and pouring.

For HV multipliers, an output resistor is added to limit diode current if the load can ever arc.
I run the boards in air in total darkness to see where the corona hot spots are, if the PCB layout is not so great you can see the glow. Sometimes a sharp edge or spacing is missed.

Some HV rectifier diodes are sized to be potted i.e 2CL71MT is SMA yet rated to 8kV and I estimate ~2.5mm gap between pads. You wouldn't use that in air.
 
The following users thanked this post: doktor pyta

Offline ahbushnell

  • Frequent Contributor
  • **
  • Posts: 751
  • Country: us
Re: 9kV CW multiplier potting process
« Reply #11 on: June 11, 2020, 02:52:48 pm »
Just two things, I’m surprised that the pcb does not have any cut slots, below diodes and caps, moreover when diodes are not placed in a triangle shape. Also, usually hv pcbs to be potted do not have any soldermask, just bare pcb.
The filling process seems ok for the voltages your dealing with
I agree.  We always make slots under HV components. 
 
The following users thanked this post: doktor pyta

Offline doktor pytaTopic starter

  • Frequent Contributor
  • **
  • Posts: 492
  • Country: pl
Re: 9kV CW multiplier potting process
« Reply #12 on: June 13, 2020, 12:04:12 pm »
Thank you for valuable feedback !


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf