I set the rules for .050'' traces and .050" spacing on my vacuum tube PCB's. I generally leave as much extra spacing as available around the plate pins on output tubes since the voltage on the plate can reach 1 KV in normal operation. It can see 2 to 3 KV (actual measured values) on a P-P guitar amp cranked full tilt into a guitar speaker near its resonant frequency.
I will also make heater traces thicker if the current is higher than 600 mA.
I will also make heater traces thicker if the current is higher than 600 mA.
I had issue recently on a PCB with 430VDC. I could not explain why I had distorsion and resorted with using an old board with success. So my new board was defective.
I had a 0.8mm trace clearance. Not enough. I re-ran the PCB with a 1.5mm clearance and did a new batch. All good this time.
Since then, I always clear 1.5mm with high voltage tracks.
I had a 0.8mm trace clearance. Not enough. I re-ran the PCB with a 1.5mm clearance and did a new batch. All good this time.
Since then, I always clear 1.5mm with high voltage tracks.
I spent several days chasing a weird issue on a home made board, I finally came to believe that it was the fault of the board itself, and I made another identical board. It was OK.
Discussions, and testing with one of the materials experts at work revealed minute leakage currents in the PCB itself. He explained the two main reasons this could have happened. We believed that both came into play here.
Some budget G-10 material (similar to FR-4, but not Flame Retardant) can be hygroscopic (absorbs moisture from the air) given the average humidity in south Florida is in the 90% range for most of the year, a likely possibility.
The same moisture absorbing properties could have caused some of the etchant (with copper in solution) to be absorbed into the board.
I have added an acetone wash to my process which should help take the surface moisture out.
The same materials guy took one look at my blown head gasket and told me that I had never changed my car's coolant. He was right. The old predominately iron car engines didn't care. A modern engine is made of iron, aluminum, magnesium, copper and others. With old fluid the cooling system becomes a battery and the copper ring in the head gasket gets eaten! "Galvanic corrosion"
@Tubelab_com
Very interesting, are some papers about this phenomenon existing?
In this case, all printed board assemblies needs to be coated by manufacturer then it´s become a real safety issue.
One method to reduce safetly creepage is to go multilayer and put high voltage in inner layer: FR4 is good isolator (roughtly 1mm/10kV) and no inner layer air cavities after manufacturing process. I´ve use this process in PCB´s for nuclear power plant.
Don´t forget creepage on PCB is equal to air gap!
JP
Very interesting, are some papers about this phenomenon existing?
In this case, all printed board assemblies needs to be coated by manufacturer then it´s become a real safety issue.
One method to reduce safetly creepage is to go multilayer and put high voltage in inner layer: FR4 is good isolator (roughtly 1mm/10kV) and no inner layer air cavities after manufacturing process. I´ve use this process in PCB´s for nuclear power plant.
Don´t forget creepage on PCB is equal to air gap!
JP
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.