I recently had a PCB designed which has several analog, digital, and power elements. The board is powered by 24Vdc which is stepped down using an LDO to 3.3V and 5V. These voltages are used to power the many IC and MCU on the board. The issue I am having is when I touch the ground layer on the board to measure voltages with my multimeter a very intense voltage oscillation occurs. This first picture shows the voltage spike on a 1 volt signal on the board. enter image description here

This second picture shows the oscillation on the 24Vdc power line. enter image description here

Some other pieces of information:

I first noticed this issue while debugging the MCU code. I was verifying the DAC output and noticed when a single probe on my meter touched the board the debugger would lose connection. Thus I used the oscilloscope to see this disturbance.

This phenomenon is very random when it occurs. It is not instantly repeatable and when it occurs the amplitude of the disturbance varies.

We've used 2 different 24Vdc supplies and both create the same issue.

I've also recreated the issue using a stray wire as opposed to a multimeter probe.

My current thoughts:

I haven't provided the PCB layout because I don't need someone to solve this problem for me (although I will provide pieces of it if need be). What I would like to know is what could cause an issue like this? To me it seems that because it only takes a single open wire to create this issue, the stray wire is acting as an antenna or some sort of conductor for a massive discharge on the board. The spike always starts with a voltage dip and it seems like it takes time to "recharge". Let me know what you guys think about this. I have seen grounding issues on a PCB before that cause signal integrity issues but never anything that cause the high voltage spikes.

  • 1
    \$\begingroup\$ Very high impedance input, high gain, no guard traces, trace inductance, etc. What you may be seeing, and this also may explain the "not instantly repeatable" issue, is the transfer of charge from/to your measuring probe and the circuit responding to those sudden initial conditions damping out through oscillation. Once the charge balance takes place, the 2nd time you touch it doesn't necessarily happen. But given time, charge accumulates and the transfer can re-occur. But I'm a hobbyist. My betters will likely chip in soon with better advice. \$\endgroup\$ – jonk Nov 8 '19 at 19:13
  • \$\begingroup\$ I was also thinking it was some type of ESD issue caused by the board as well. It makes sense given the "charge time" I have noticed. Can this be a side effect of a poorly laid out ground plane? \$\endgroup\$ – dbeng Nov 8 '19 at 19:51
  • \$\begingroup\$ I don't have extensive experience. My interest/experience is physics, not electronics for which I haven't even had one DC class. But from the physics perspective, it's very very easy to see metals as massive oceans of free electrons. If one ocean level is just slightly different from another ocean level, when you touch them there is a massive transfer, very very fast, to level things out. So it's easy to visualize from that viewpoint. \$\endgroup\$ – jonk Nov 8 '19 at 20:14
  • \$\begingroup\$ My experience is limited to a few situations with a very high input impedance, high gain, and high bandwidth front end used as part of a phosphor thermometer. (You can literally paint phosphor onto a jet turbine blade, fire up the jet engine, and monitor the blade temp in-situ while it runs at 100k rpm. Had to use every trick, including encapsulating the entire front end into a can. But definitely got there. Luckily, the signal was optical!) \$\endgroup\$ – jonk Nov 8 '19 at 20:16

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.