Power out of range? (3V)

[jakematic]

Moved UAP to port 4, same fluctuations

[Dave]

hmmmm...can you put more loads on the switch to see if it is a light load issue..i can imagine that the 24 volts might jump around a bit until we get enough of a POE load, but I have no idea why the 48 volt power would jump around so much...

ummm, what part of the world are you in...is it 50 or 60 hz power?

[jakematic]

Ehhh... don't have anything else right now I can load the PoE with.

North Carolina, USA, 60Hz

[Dave]

that's to bad....ya got me scratching my head....beats me..i did a lot of POE load testing & never saw this....& we haven't had anyone else complain.....hmmmm..what happens if you don't use the UPS & go straight of the main power grid??

[jakematic]

Going off the main grid is how I had the last WS8 setup and it arrived in PA on the 17th for RMA analysis.

Granted that one wasn't surge protected and I forgot to ground it so I got the requisite reminder from Chris.
I've since improved my install :)

The RMA'd one was powering 2 UVC-G3 and an mFi, and eventually went down this same path.
Eventually it just died.
There is an old TS5 where it was now (that the WS8 had replaced) and it 'seems' to be working fine but doesn't have the UI Netonix does.

Very perplexing.

Am hoping against all hope the WS12 doesn't start going down this path too !

[jakematic]

OK interesting update on this.

Was putting everything back, and noticed those pre-made cables didn't seat well - kept knocking the NTP servers offline.

Data points for process of elimination:
- the WS12 has PoE cables that I made powering another UAP and an mFi
- RMAd WS8 had 1 pre-made to the mFi and 2 cabled I made to the cameras, the uplink was also one I made
- new WS8 now has a cable I made up connecting to the UAP

Voltages seem a little bit more stable, but are still fluctuating.

[sirhc]

Added surge protectors, moved some equipment, verified electrical grounds (tester), verified correct wiring (tester), added ground straps to everything in sight that wasn't already, and put in an APC SMT1500.

Just so you realize if this is a ground potential issue damaging the units AC Surge Protectors, Ethernet Surge Protector, and APC UPS have no way to protect against ground potential shift, they are not design to protect against this as explained in this grounding post: http://forum.netonix.com/viewtopic.php?f=30&t=1816

Surge suppressors shunt excess voltage to ground but they can do nothing with ground current flow which is often low voltage high amperage.

Does all equipment involved get AC power from the same AC service (meter)?
If yes that is good, if no then that is bad.
I ask this because I wonder as you mention garage and possibly this is a separate building with a sub panel but at the very least a possibly long AC cable run which would be a different ground potential than a switch in the house if it is a long wire running to garage.

Can you draw a simple diagram of where everything is and where it gets power from. This includes the switches and what they power. As I said a crude drawing is OK. Label the cable lengths between devices and note where they are located.

Can you check your electrical service ground rods to make sure they are solid in the ground (can not move them with your hand) and all connections look good (no corrosion).

[jakematic]

Understood - have learned a lot reading your posts on grounding.
Short of driving a ground rod at every switch, I've made sure they are all grounded to a common thing - the electrical ground.
Not sure what else can be done in this situation other than switching to fibre between everything and/or shooting an M9 across the grounds, which doesn't sound like an exciting project...

Everything is connected to the same service on the same load center in the attached garage.
The WS12 is opposite corner from the load center, so yes that is a long run.
The WS12 replaced a Netgear as did WS8B

Ground rods are very solid, doesn't appear to be any corrosion, connections are solid.

The WS8 at the top of this picture is the one you have back, and there is a TS5 in there at the moment.

[sirhc]

Actually a ground rod at every device is a BAD idea unless all ground rods were bonded together with HEAVY wire as this is what causes ground potential differences. Every ground rod will have a different ground potential and then they will try to equalize across your Ethernet cables. This is why I stress Ground is not Ground as every grounding rod/system has its own different ground potential.

Just like plugging one switch into your house then running a cable to your neighbors house switch plugged into his electric service is VERY BAD. This will result in damaged equipment every time.

I will review your drawing later today, headed to pool with kids. (last day of vacation)

[jakematic]

Actually a ground rod at every device is a BAD idea unless all ground rods were bonded together with HEAVY wire as this is what causes ground potential differences.

That is what I meant to say.
Sorry am trying to handle 16 things right now, on a webex training session, doing other work and I didn't get the complete thought out.

THank you both for looking this over. It's getting really frustrating... and expensive... which I'd much prefer avoiding more of...