1
\$\begingroup\$

I have found that PoE uses isolated power delivery. If I splice into my ethernet cable, will I be able to hook up a scope to one of the 8 lines inside safely without frying anything?

The scope and the PoE switch will both be connected to earth ground. If I understand this correctly, the PoE spec specifies power delivery must be performed with an isolated power supply. If this were not the case, then I could expect touching the ground return lead on the probe to one of the "+" PoE wires would cause a short to ground. But since PoE is isolated I hope that this could be safe.

\$\endgroup\$

1 Answer 1

2
\$\begingroup\$

When in doubt, put a 100k resistor between the two grounds and check the voltage across the resistor. If you see anything except 0V, a differential probe is a good idea.

\$\endgroup\$
6
  • \$\begingroup\$ Nice. thank you for the general troubleshooting tip. Do you think it is equivalent in terms of safety compared to: (1) Touch probe tip from scope to the other device's ground and check the voltage reported by scope? (2) Use a battery multimeter in DC voltmeter mode to measure the voltage difference? I am guessing that the choice of using a 100KOhm resistor here is to allow any transient charge to dissipate relatively quickly (as opposed to w/ a 10x scope probe or multimeter which would have at minimum 1MOhm impedance. \$\endgroup\$
    – Steven Lu
    Commented May 22, 2022 at 23:11
  • \$\begingroup\$ Thought about it some more... It may be pretty nonsensical to attach a scope ground to the PoE cable on which each wire is carrying a high frequency signal. Wouldn't matter if I connect it on the PSE or PD side of the shunts and wouldn't matter which wires I've got the shunts on: Nothing good would come of it, the best case should be that I do no damage to the scope and merely disrupt the ethernet signaling by introducing a large ground plane to it. It definitely seems to me like a differential probe is the answer, or I suppose two channels and math could be used for something similar. \$\endgroup\$
    – Steven Lu
    Commented May 23, 2022 at 8:36
  • \$\begingroup\$ Because of capacitive coupling, I don't think the probe check will be reliable. A properly designed POE switch / injector will provide isolated power. However, there are lots of crappy designs out there that are not isolated. \$\endgroup\$ Commented May 23, 2022 at 9:24
  • \$\begingroup\$ Even with a differential probe you may have trouble with reliable data communication with the probe connected. The POE negotiation will work without data, so depending on what you are trying to measure this might not be an issue. \$\endgroup\$ Commented May 23, 2022 at 15:07
  • \$\begingroup\$ I'm using this Micsig DP10013 Diff probe and maybe there are some more settings I should tweak on the scope but I'm getting a large amount of noise (hundreds of mV) with the leads shorted. But this is a high voltage diff probe. I'm not using the right tool here but the tools I "need" are prohibitive expenses. \$\endgroup\$
    – Steven Lu
    Commented May 23, 2022 at 23:14

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.