High voltage traces on PCB

[archiees]

high voltage pcb

Hi,
I want to calculate analytically how much spacing is to be kept between the traces with high voltatge to prevent the arcing. Does any one has references of a text book or articles.
thanks

 

[hock]

pcb high voltage spacing

surface treaking is the thing you need to keep in mind.
how much is your voltage?
you can try to keep all high voltage trakes on one side and other trakes on the other side.
hock

 

[archiees]

high voltage trace spacing

Thanks Hock,
I am not sure i know what is surfance treaking? Can you please tell me more about it.
I have an RF voltage of +/- 215 volts. Also, i am making a board with soldermask on each side, which i guess give me a little advantage, i don't know why though.
However, i still want to do see some calculations/formulae.
Can't we derive from first principle, knowing the dielectric strength of FR-4, how much distance will cause arcing/corona discharge?
thanks again.
-arch

 

[cyberrat]

pcb high voltage clearance

2-2.5mm should do for 215 volts.
There are some formula/graphs available in some standards. cant remember them though.

 

[flatulent]

pcb spacing voltage

There are safety laws that deal with this. I vaguely remember that 220V mains traces have to be 3 mm from everything else and each other. The higher RF frequency may cause corona on the sharp edges of the traces. You will have to put some dielectric coating over them to keep this down.

The safety laws may also specify how far from human bodies they must be kept and what sort of interlocks must be used to turn off the RF power when the case is opened.

 

[IanP]

pcb voltage spacing

Q: Where would I find info on trace spacing based on voltage levels? How far should components be spaced for 1000V vs 4000V that type of thing?

Answer is attached (source unknown)
Regards,
IanP

 

[archiees]

high voltage spacing

Thanks IanP,
The documents are helpful for my design. I can use the clearance numbers for my 1 KV potential difference between 2 traces uncoated. However, i want to calculate them analytically also, which is important if i have to jutify it to someone.
If i use 750 V/mils, i get abt 1.4 mils which is not equal to 200 mils from the note and the IPC-2221 charts.
I understand that the FR4 has dielectric strength of 750 volts per mil perpendicular to the layers. But to calculate the clearance between uncoated conductor on the same layer, should i use dielectric strenght of air which is 3Kv/mm. Still won't be the corrent answer. Am i totally wrong here?

 

[IanP]

pcb clearance for high voltage

This 200mils from your post seems to be an error, is it not?
To calculate the clearance between uncoated conductor on the same layer I would use the same 750V 750V per mil insulation as the worst case scenario ..
Regards,
IanP

 

[cyberrat]

high voltage on pcb

Of course, if you look at the spacing between the legs of a TO220 transistor that can handle upto 1000 volts this sort of makes a mockery of what we use for spacings on the PCB, especially when the PCB can be coated.

That scenario has always puzzled me.

 

[archiees]

high voltage pcb spacing

Let me tell you how i use the IPC-2221 to determine my clearance for 1 KV signal.

Lets use the B2 colum of the IPC table for the clearance for uncoated external conductor at sea level.
For 500 volts it has to be 98.4 mils. Now, for every additional volt we have to add aprox. 0.2 mils. Thus for the remaining 500 volts, we need to add 500*.2 = 100 mils.
So in total we need 198.4 mils.
Does it make sense?
THanks for your help.

 

[vdaniel]

pcb high voltage

Dear Colleagues,

I have arisen a similar discussions several times, but, unfortunately, without a clear, but resonable figures in response. Some tables give about 10 mm/kV,
but I have made a High Votage converter PCB with about 5mm/kV, which was
tested to work successfuly on the 3600 m height.
Varuzhan