If you can explain what you do k now about transformers and what you think you need to but dont, it should be possible to fill in some gaps ;-)  In the meantime some general thoughts.

You do not say if the mechanical stimulus to be harvested is intermittent (the extreme of this would be the ones in the

qinatic switches) or more like a continuous system from say engine vibrations or wind flapping something.

In all but the highest voltage systems or the most mission critical cases,  a partial discharge test is unlikey to be helpful oe required in an energy harvester.

The most common fault is a shorted turn, and at manufacture that can be detected by an inductance / Q-meter test. During use that will manifest as a catestrophic losss of power.

I imagine you will be looking at sub-watt power levels, so the main thing will to be not to waste energy with more load than is required, so self test needs to be simple.

Also the design of a transformer for pulses or indeterminate frequencies or power levels is not the same as nrmal 50Hz power transformer design  - devices can be much smaller if heating is short term adiabatic rather than steady state.

Mike.

Hi Mike,

Many thanks for your insights on this topic. I'll take your inputs into consideration during my project.

Cheers,

Jona

What sort of transformers are you thinking of? They can vary between huge 50Hz things (as used in the public supply) to small 50Hz ones traditionally used in items of equipment or very small high frequency AF/RF or HF power conversion ones?

If its the large public-supply kind of thing, I would suspect the first parameter to measure would be temperature. (Although I'm not sure how kind such an environment would be for wireless comms...)

   - Andy.

I had the larger ones in mind, so thank you for your input regarding temperature sensors. 

I think you will be very lucky to build any sort of mechanical harvesting system that needs a power transformer, unless the energy harvester looks more like a generator !

But substation transformers, when monitored at all, usually look at oil temp as the main one, and then the  LV load balance (currents, and voltages) across the 3 phases.

An example of a commercial transformer monitor here, Note that since designing that range, Ash Wireless has since been taken over, and may not support that sort of product in the way they once did when they first developed it.

Mike.

I think you will be very lucky to build any sort of mechanical harvesting system that needs a power transformer, unless the energy harvester looks more like a generator !

I'm confused (easily done I know) - I'd imagined the OP was taking about powering a little sensor, plus it's wireless comms, from picking up stray magnetic fields from the transformer it's monitoring. Something along the lines of having only the secondary windings of a tiny transformer to power the sensor, it's being powered parasitically from the main transformer's primary. Apologies if I've got the wrong end of the stick.

   - Andy.

Aah. It is also possible that I have misunderstood - looking again, it could be read that way, But if you are connecting to TX monitor it why harvest power with great inefficiency from magnetostrictive vibrations,  when you already have a mains supply.,,,,

Mike.

But if you are connecting to TX monitor it why harvest power with great inefficiency from magnetostrictive vibrations,  when you already have a mains supply.,,,,

I can see it might be convenient - especially when retrofitting or if the monitoring is only wanted temporarily - just stick the device onto the side of the transformer - job done - no need for live working on maybe 1000A 400/230V connections and deal with 25kA+ fault currents.

   - Andy.

Maybe. But if you want to monitor output current and voltage you probably  need to get up close anyway.  I can see that clamp metering current on singles may be OK and if wire is unarmoured you could monitor voltage waveform capacitively through the insulation. An interesting idea, but I think the folk installing it would have the paperwork to go into the subastoin or up the pole anyway.

I must dig up a photo of an old style  open frame substation - bare bus-bars even when not open for maintainence...

Mike.