Ferrite Choke questions

[NukeEngineer]

Actually it's PLC, but could be BPL elsewhere. PLC is "power line carrier".

[NukeEngineer]

Generally, a smart meter uses a signal like Zig-Bee in the 2-5GHz band. I don't think it uses ac cable "carrier current" communication - cause that would stop at the transformer.

You (or a neighbor) may have some other gadget installed that uses carrier current (wireless intercom, X-10 home automation) and that's mucking up your inverter comms.

Nope, some use BPL which is carrier current. It's a strong enough signal to pass through the service transformer and reach the distribution substation, but not further.

See this link from ARRL and search for "BPL" http://www.arrl.org/smart-meters

[Mike90250]

You're not following.... The interference is coming from the smart meter, going through the AC lines to the inverters. Nothing is radiating.

Generally, a smart meter uses a signal like Zig-Bee in the 2-5GHz band. I don't think it uses ac cable "carrier current" communication - cause that would stop at the transformer.

You (or a neighbor) may have some other gadget installed that uses carrier current (wireless intercom, X-10 home automation) and that's mucking up your inverter comms.

[sdold]

This sounds very odd to me. I thought smart meters only radiated, using internal antennas, in the ISM bands. I don't think they would be able to operate on lower frequencies on the power lines because I don't think it would make it past the service drop transformer. How did you measure the conducted RFI and what was the frequency that you saw? I'm not saying you're wrong, but I'm having a hard time understanding how it would work.

[DanKegel]

What's the system you're trying to protect from the RFI? How far is it from the system? And why do you suspect the smart meter?

Does a handheld broadcast FM radio pick up roughly the same RFI?

[bcroe]

That's what I was afraid of. I suspect everything is radiating, including the micro inverters themselves, so they have to be placed in a shielded enclosure, their wires run through toroids, but worse, the noise is probably back-feeding into/onto the panels themselves. Not sure what can be done about that.

The panel inputs can be filtered if that is what is needed. RF is slippery stuff, it doesn't follow nice
simple paths. Getting some real filters will produce better results. Bruce Roe

[NukeEngineer]

You're not following.... The interference is coming from the smart meter, going through the AC lines to the inverters. Nothing is radiating.

[kb58]

That's what I was afraid of. I suspect everything is radiating, including the micro inverters themselves, so they have to be placed in a shielded enclosure, their wires run through toroids, but worse, the noise is probably back-feeding into/onto the panels themselves. Not sure what can be done about that.

[NukeEngineer]

I've got three chokes on each inverter this time, with both hots and neutral going thru them, so no overload worries. Let's see if this makes a difference...

Didn't help.

[bcroe]

I'm a ham as well and will be facing this demon myself. From what I understand, microinverters/optimizers were never designed to be RFI shielded. Every single box, and its associated panel, and all the wiring, ends up forming a giant transmission antenna. Adding toroids way out at the end of one set of wires does nothing cutting down on the radiated energy. It's an big enough problem that I'm not sure that it is solvable short of moving away from microinverters/optimizers and replacing it with a traditional string converter.

If I had to use micro inverters, I would shield each one and filter all their wiring. No radiation to panels or
connecting wires. Bruce Roe K9MQG

[NukeEngineer]

I've got three chokes on each inverter this time, with both hots and neutral going thru them, so no overload worries. Let's see if this makes a difference...

[kb58]

This may help, though of course they're trying to sell you something: http://palomar-engineers.com/rfi-kits/solar-system-rfi

[jflorey2]

I've got limited experience with these. So, is there is current limit on these? I'm clamping around the individual L1 and L2 lines, not both in the same choke. With a fairly small current, the bead I attached early this evening was buzzing. I'm worried that at 50 amps they might overheat or something.

Try both wires in the same choke. That makes it a common mode choke - and you are most likely seeing common mode noise that's re-radiating.

[NukeEngineer]

You need to know the general frequency range of the interference, how strong it is, and how it is
propagating. Then an appropriate filter can be chosen. Throwing beads at it is just a shot in the dark.
Bruce Roe

100% agree, but was worth a low-priced shot. I'm having a hard time finding any info on what frequency and signal level meters tend to transmit on.

What's interesting is that the meter guy was surprised that you could hear the meter communicate audibly, just standing next to the meter and inverters. He indicated that he had never noticed this before.

He did mention that when the communications from the substation is generated, it IS audible and loud, and it's the kind of sound that makes you want to run. It supposedly sounds like shorting 277V to ground with a dirty conductor so that it's not a clean 60hz hum, but a sizzle sound. I would LOVE to find a video of this, but can't find anything yet.

[NukeEngineer]

Wait, I'm confused, you say above that you wouldn't expect the chokes to do anything, yet earlier:


Which sounds like the RFI may be radiated instead of conducted - or both. How am I mixed up?

If the interference was radiated and being picked up by the panels, wiring, etc. then chokes on the AC lines would be pointless. However, the interference is conductive, and being generated by the smart meter.

Analogy: The bad guys are coming in the front door of the house, so I tried a filter for front doors. If they were coming in the windows, then I would have expected a front door filter to be pointless.