Maker Pro
Maker Pro

23KV per inch update

K

Ken Smith

Jan 1, 1970
0
I finally got tired of the noise and switched off the supply. The bit of
FR4 which had 23KV per inch on it shows no signs of a problem. This is a
lot longer than I need it to survive in the production case.
 
J

John Larkin

Jan 1, 1970
0
I finally got tired of the noise and switched off the supply. The bit of
FR4 which had 23KV per inch on it shows no signs of a problem. This is a
lot longer than I need it to survive in the production case.

I had a couple of old analog Kikusui scopes that had an awful pcb
layout in the CRT supply, probably 40-50 KV/inch. They would
eventually accumulate dust, arc, and zap some exotic diodes [1]. The
fix was to clean carefully and glop with thick conformal coating.

John

[1] This was in the cheapie Z-axis coupling circuit. Classic scopes
had two transformer-isolated negative supplies, one for the CRT
cathode and one for the grid, with the cold ends driven by the
intensity control and the unblanking pulse. The cheapie circuit
combines AC coupling with a modulated-level ac carrier that's
rectified on the high side to achieve sorta-net DC coupling to the
grid.
 
K

Kevin

Jan 1, 1970
0
Ken said:
I finally got tired of the noise and switched off the supply. The bit of
FR4 which had 23KV per inch on it shows no signs of a problem. This is a
lot longer than I need it to survive in the production case.

A few years ago I did some testing for a MEMS based optical switch and
we found that a 5mil spacing between traces on the surface (covered
with solder mask) broke down at about 2KV while buried traces broke
down at about 6KV.

That works out to about 400KV/Inch on the surface and 1.2MV/Inch for
buried traces.

We are using voltages of <200V with 5mil spacing.

One problem that did show up after long durations was corrosion and
electro-migration under high humidities around exposed copper/solder at
connectors even though the spacing was much larger.

kevin
 
K

Ken Smith

Jan 1, 1970
0
Kevin said:
One problem that did show up after long durations was corrosion and
electro-migration under high humidities around exposed copper/solder at
connectors even though the spacing was much larger.

I was running the test at RH=100%. There was a pool of water under it the
whole time.

ASCII ART

------------Plastic-bag-------------
=wires========PCB !
------ -------- !
! ! !plastic ! ! !
! ! ! Block ! water ! !
! --=======------------- !
 
I

Ian Stirling

Jan 1, 1970
0
Ken Smith said:
I was running the test at RH=100%. There was a pool of water under it the
whole time.

ASCII ART

------------Plastic-bag-------------
=wires========PCB !
------ -------- !
! ! !plastic ! ! !
! ! ! Block ! water ! !
! --=======------------- !
----------------------------


The bag was wrapped tightly where the wire came in to keep all the water
vapor in.

I would be concerned about salty environments being lots worse, if that's
possible.
 
K

Ken Smith

Jan 1, 1970
0
Ian Stirling said:
I would be concerned about salty environments being lots worse, if that's
possible.

I don't expect salt to get to the PCB. It does bring up a good point
about ions. I'll have to check into what ionic stuff can be in vapor
form.
 
R

Robert Baer

Jan 1, 1970
0
Ken said:
I was running the test at RH=100%. There was a pool of water under it the
whole time.

ASCII ART

------------Plastic-bag-------------
=wires========PCB !
------ -------- !
! ! !plastic ! ! !
! ! ! Block ! water ! !
! --=======------------- !
----------------------------


The bag was wrapped tightly where the wire came in to keep all the water
vapor in.
You forgot the frog!
 
J

Jasen Betts

Jan 1, 1970
0
I don't expect salt to get to the PCB. It does bring up a good point
about ions. I'll have to check into what ionic stuff can be in vapor
form.

carbon dioxide (carbonic acid), hydrogen chloride (hydrochloric acid), and
ammonia (ammonium hydroxide) are three fairly common gas forms of ionic
substances.

Bye.
Jasen
 
Top