How to create a voltage division - but without resistors?

Thread Starter


Joined Jun 28, 2013

I have a surface-mount connector that I'm putting some high voltages on alongside low voltages and signals. The connector itself is has plenty of insulation to handle the voltage differences, however, the copper on the PCB that I'd mount that connector on is too close to each other. With the 0.5mm pitch, they are in danger of arcing. I am trying to adhere to a specification that demands 0.6mm separation ath the 0-50V range I am working with; the footprint copper currently gives 0.2mm separation.

My solution here is to separate high and low voltages and leave two "unused" pins between the used ones. The idea being that this gives me three spaces of 0.2mm, which provide my total separation of 0.6mm. However, I can't connect these pins to either neighbour or ground as they'd then undermine the whole idea. Which creates undesirable floating voltages! My solution to that was to create a small, 3-resistor potential divider with megaohm+ resistances, so that the copper would charge in steps across the voltage delta elegantly.

But... then I realised that for this (large!) connector I'd need 120 or so resistors, which seems a little... excessive! There are lots of pins that need to be isolated from each other. I am wondering therefore if there isn't another way to safely discharge unconnected pins in between two very different voltage pins? I'm quite ready for there not to be a solution to this, but I'm thinking if perhaps there is some partially-conductive paste I could smear over the pins that might put a couple millions ohms between the stuff it touches! Or some other similarly unusual idea...?


Mark Hughes

Joined Jun 14, 2016
What problem are you trying to solve? Mixing high voltages, low voltages, and signals is usually not a good solution to any problem. And forget about the copper traces for a moment, what's to keep an arc from jumping from one of the sharp edges on the connector pins to something on your board or another pin on the connector?

What's the problem that you are trying to solve with this 0.5mm pitch connector?


Joined Sep 24, 2015
Working on a DC-9 Series 50 for the NAVY, a 308 volt lead in a connector accidentally bent over and shorted to a ground pin. The 308 came from a ballast, so there was no breaker to break the circuit. The cockpit began to smoke and, not knowing what circuit was burning we shut the whole aircraft down. The hydraulic guys were pizzed because when they're running hydraulics you're not supposed to touch ANYTHING. Someone could lose their life. But since the hydraulic pump stand was shut down we didn't think there was any danger. When we shut it down the hydraulics supervisor came hollering about us turning power off, despite the cockpit with smoke rolling out the windows. He insisted power be restored, but when we refused he threw us off the plane. He had that authority. HE put power back on and left the cockpit. Quickly it began to smoke up again. Just then an entourage of NAVY dignitaries were touring the facility. When the VP of operations asked me why the cockpit was on fire I told him. He said "Shut the main generator down." So I did. He went back to his group and started to explain how he was going to get to the root of the problem and make sure their plane was ship shape. Moments later the hydraulics supervisor came screaming and hollering he was going to have me walked off the base. The VP stepped around the corner and asked him why the plane was burning. He had no answer, and HE was the one walked off the base by security. Took us three months to completely replace that one wire harness that had shorted out.

Do you really want to mix high voltages along side low voltages? Personally I don't think it's a wise idea. If it's a single or double layer board you can always lift one trace off the board to get your spacing, and like JWHassler said, remove the unused pins to ensure there is sufficient isolation.


Joined Apr 24, 2011
What is so undesirable about floating voltages on unused pins?

Skipping pins sounds like an excellent way of gaining spacing for isolation.

If the problem is coper on the PCB you may want to just skip placing pads on the isolation pins.


Joined Feb 20, 2016
Cut a slot in the PCB between the high and low volts.
Here is a pic I just found on line.
This is a normal practice for example, when there is mains voltage on a board.

Thread Starter


Joined Jun 28, 2013
Thanks for the responses (and, Tony, for that illustrative story :) )

I should probably clarify that the voltages aren't that severe - just 50V max. However, this is a system that we want to have very high reliability, and so we're trying to follow a quality spec which dictates this separation. Ideally, my four neighboring pins would be at 0, 17V, 33V and 50V - easy to do with resistors, of course, but trickier without...

Actually, strictly speaking, it demands a mere 0.13mm separation, as we will conformally coat the board, but we can't do that to the standoff area of the connector (the vertical risers) which have the same pitch - it is there that we are worries about arcs.

Pulling pins would help at the board interface, but it is these other parts of the connector where bits of copper would be close to each other that we want to control the voltages on. The connector is a Samtec QSH/QTH pair, and there would be quite a lot of cutting to remove all the leads! I don't think it is possible on this sort of connector to remove all the conductors on the unwanted pins int he separation positions - although this would be the ideal solution!

Papabravo 10 layers.

Ernie floating copper is liable to build up charge over time and eventually will discharge this into a neighboring conductor to dissipate it. The energy is low, so this isn't likely to cause and damage, but it could quite easily lead to a logical change on a signal line.