Harmonics in pole power transformers?

This can't not work, unless the power feed is in a metal conduit. Fluke and others sell voltmeters with field sensors in them. You can even buy screwdrivers with sensors in them. Even if some of the feed is in conduit, a lot of installations have some portion that is not shielded.

Jon

Unfortunately it usually is :-(

Not these because they are outdoors but I am sure hoping for some leakage around the breaker box plus maybe some E-field remnants due to non-ideal grounding of the conduit but that will be like trying to listen to grass grow. However, that wouldn't be the first time so it doesn't faze me too much.

After seeing how they handled phase angle issues around here I am not so sure about that. 15 years ago we had voltage surge issues that cause light bulbs to blow and it took so persistence on my part to have someone come out. Turns out they just connected a capacitor bank slam-bam style via timer, regardless of any measured values.

We'll see. One problem is that all the cabling is in metal conduit.

Yeah but I am thinking you don't need 'instant' notification of faults - maybe a second or two is OK and the mains is a fairly fixed frequency you can filter the hell out of the 60Hz to get your all OK signal

Oh yes, we can take minutes or even tens of minutes. Though our times of running the electronics is limited because the whole thing is more or less battery-operated. So we can't have an ARM-Cortex crunch numbers for

Of course it CAN fail, a 220VAC drop, with center tap to ground, will indicate one phase if your probe is placed on one side, the opposite phase if it is placed on the other side, and SOMEWHERE in the middle, your probe won't detect anything. Also, if your transformer fails with an open phase, the center tap and one leg will still have voltage; it'll have the 60 Hz field, but won't drive your 220V motor.

The obvious, correct, way to detect line power, is to plug into the line. The 'convenience outlet' is a common adjunct to many electrical installations, one oughtn't do anything exotic until you know it isn't available here.

Build clamp with three sensors. Select the best signal or sum (absolute value) of the three. This doesn't solve the conduit problem, though.

I think the problem is that he's trying to solve a problem for all installations. Not all will have a convenience outlet. The solution may be to include multiple sensors (perhaps options) and select the one that works best. Perhaps use them redundantly and vote, or some such. I don't think one solution is going to work for all possible installations, give Joerg's long list of restrictions.

Assuming the transformer is on the top of the pole and power is brought down via a two wire cable; and that the cable leads to the pumps switch which is a few feet at least from the transformer. Put a foot or two or three of hookup wire on each side of the power cable & connect them to a length of coax, which, with adjusted length, will serve as a good voltage divider and provide an opam with fet inputs a 60 cycle signal to monitor.

Hul

Joerg wrote:

Unfortunately that requires an electrician which we'd like to avoid, mostly on account of the remoteness of the installations.

The hookup wires are not electrically connected in the common sense. They serve only as a capacitor. One electrode is one of the hookup wires and the other electrode is one of the power leads, with the cable insulation and the hookup wire insulation between them. No "electrical" connection. And, I'd bet ($2) a lm358 would serve for an opamp. I checked a bunch of them recently and found the input current notably less than spec would suggest.

Hul

Joerg wrote:

You have not stated why a direct connection to the input side of the pump switch is not possible. What is wrong with a small power transformer ?

If you need better insulation, get a certified voltage transformer, some of which are intended for outdoor installation.

Yes he did. No connection to any of the supply or load circuit allowed. One requirement is that no electrician be required for installation.

Apparently they need a Central Psychic (tm) or one per district that will inform management when service is required and whom to send.

float switch on the trough.

yeah, but you save some that you'd otherwise have to spend on on cow-proof solar panels and batteries...

Can't you create your own power flow, by installing a low-current shunt and measuring current through that?

It would still require an electrician because customers would not allow their staff to open switch panels, switches, conduit or the like to lay a snippet of hook-up wire in parallel. Opening is the only way to get there. Unless we are able to sniff the 60/180Hz from the outside and that's what we'll try to do. A LM358 wouldn't quite suffice, it'll be at least partly discrete circuitry. Partly on account of a very low supply voltage but also because we'll need well in excess of 10M input resistance.

You need those anyhow. How else would you convery a mains power failure? Deducing that from a mote not transmitting isn't secure enough because then a damaged mote would trigger a utility truck roll. Which will come with a hefty bill if they found everything to be fine out there on the utlity's side.

An electrician would need to come out for that. Too expensive in remote locations.