Radio frequency interference with recently installed solar panels

[Mike90250]

i suspect a bad optimizer, something like a cracked casing or a bad internal component. Having a Good array next to a bad one, makes troubleshooting very easy. Walk a radio around till you find the noisy component. Or a sheet of cardboard on a pole to shade each panel , find out which one has the bad optimizer.

[bcroe]

That is just fine, it is their emissions and you are supposed to find a way to live with them.
The fact that one system is NOT causing a problem, shows that is in NOT inherent in the
design. I would go after the EMI radiating array, and localize the point of radiation. If it
is the other type inverter radiating and conducting along connections, you have some
options. One is get another inverter, maybe the same as the quiet one. Another is
contain the inverter EMI with a Faraday cage and filter leads.

I would certainly myself run each array to see how quiet each is by itself. Follow that
up by getting up close with your radio to locate the noise. Some trial filters might
tried for effect. A basketful of clamp on ferrite beads is probably not a cure, but might
give some hints of where to go.

AM interference is common, but me thinks it really is excessive to interfere with FM.
Bruce Roe

[sdold]

He's basically confirming that one of the arrays, its cabling or the inverter is emitting EMI. It's nice that it's not the communications between the optimizers and the inverter, but I don't think that matters. I think his suggestion to install an antenna "outside of the footprint of the array" isn't much of a fix unless you can take a portable radio to that location and confirm that there's no interference. And with an external antenna you'd need to connect it to whatever radios are having the problem. Re-radiating it with a small amplifier and antenna in the house could work, but you need enough signal isolation between the top antenna and the one in the house to not have feedback, but enough gain to make it work with all of the radios, it's tricky to set those up. I don't know what involvement the FCC would have, but I don't think they would expect you to try an outside antenna first. The line about it being an RF Shielding/absorbtion issue is meaningless unless it turns out to be radiation from the inverter itself (instead of cabling).

[Timmyscott]

I really appreciate all the effort and advice I’ve received. We went back to SolarEdge to ask more questions and to have their tech explain how he remotely tested the system. His reply is below. I still don’t understand it fully. Does this make sense? Do these tests show that it’s the panels themselves that are causing the interference?


“Again testing was as follows:
We first current throttled inverter 731A8262 (Larger inverter and array) which had no effect on your radio. When then did the same thing to 73118103 in which we heard your radio signal improve. At this point being that I left my optimizers communicating and sending full frequency I knew that the optimizers were not the issue (RFI) and it was current related (EMI). As you should have still had the issue without turning off the communication as in HAM radio as an example (RFI). We then took it a step further and let the larger inverter and array 7311A8262 generate at full capacity which would provide full broadcast signal and it had not effect on your radio. Only current throttling the optimizers to modules (solar panels) produced an EMI effect.

Again this can be resolve with an external antenna outside of the foot print of the array. The FCC would consider this to be the first reasonable step to take and has more than enough documentation on our product to know that we are not the cause. We take every claim serious and that is why we ran the initial testing as there can all ways be outliers. We knew going into this that this was an RF shielding/ absorption issue,” but I still needed to confirm.”

[bcroe]

Hopefully the designer learned something about real world designs from this.
Bruce Roe

[sdold]

Since the interference is noticed mostly in the FM band, an FM radio might work, but I bet an AM radio tuned to an open channel might work better because FM radios aren't good at responding to noisy signals which more resemble AM than FM. Funny this came up, I just spent all day tracking down switching noise from power supplies using a spectrum analyzer. Here's what a highway patrol radio sees when the officer turns on his dash camera, and I bet inverters can look pretty similar. This is from a switching power supply in in the camera system running at about 200 KHz, black is before turning on the camera, red is after. It shows how evenly-spaced harmonics can be. Luckily it's a prototype system not yet in service.

150-170 LPR and cameras only - and all off baseline.JPG

[J.P.M.]

I think the testing has to happen mid day, with the system live. if it's disconnected, it's off and not doing anything. Walk along the array and use a portable radio on a stick over each panel, to find the one causing the RFI.

I wonder if the shade a radio creates might change the output of the panel. Maybe there's room under the array ? Just wondering.

[Mike90250]

I think the testing has to happen mid day, with the system live. if it's disconnected, it's off and not doing anything. Walk along the array and use a portable radio on a stick over each panel, to find the one causing the RFI.

[SunEagle]

An optimizer at a panel is another possibility. Bruce Roe

Yes. Technically that is a load and an inverter. Still the OP needs to disconnect his system and then trace where the rfi is coming from. Remember it may be more than one place

[bcroe]

More than likely the rfi is coming from the inverter and not the panel string. The issue is usually the "load" and not the solar panel that creates rf.

An optimizer at a panel is another possibility. Bruce Roe

[bob-n]

sdold has a great suggestion. It might take an FM radio, but a portable radio would be a great debug tool.

Walk around with it and see where the interference is better and worse. Turn things on and off.

Your system will have a DC disconnect and also an AC disconnect. It's always safe to turn off the AC disconnect but leave the DC disconnect alone. There are safe sequences required for turning off the DC disconnect without arcing.

It wouldn't hurt to also experiment with turning off other things in the house, just to make sure that you've isolated it. An easy way to do this is with one person holding the radio where interference is worst and another throwing circuit breakers in your main service panel.

[sdold]

My wife and I thank all of you for your suggestions and advice. Really gave us something to think about. Now we think the SolarEdge tech sort of sold us a bill of goods and that the rfi is caused by SE equipment.
Followup question - how would one ID which optimizer could be making rfi?

Have you shut the system down to confirm that it's the Solaredge system? I wouldn't think you'd need to tell them which optimizer it is or if its the inverter, or a particular string, just turning the system on and off should confirm it. But an AM radio moved along the array might narrow it down.

[Timmyscott]

My wife and I thank all of you for your suggestions and advice. Really gave us something to think about. Now we think the SolarEdge tech sort of sold us a bill of goods and that the rfi is caused by SE equipment.
Followup question - how would one ID which optimizer could be making rfi?

[SunEagle]

Any chance that application of ferrite core inductors on the lines might quiet things down? I do not claim to be RFI expert but I picked up military lighting system last year used on military vehicles that use HID ballasts and IR sources, at every lightbar has a HID ballast and some sort of IR LED control board with large ferrites cores attached around the cables before they exit the lighting housings. My assumption is they are there to cut down the RFI profile of the vehicle and prevent interference to the electronics in use in the vehicle. The cores are readily available and cheap plus the split core versions are easy to install, just clamp them over the lines and use a UV resistant Tie wrap.

As a side note this lighting system appears to be designed to day light up a couple of blocks around a vehicle. One fixed lightbar with multiple heads on all four sides of the vehicle plus two motorized spotlights. Flip a switch and the HIDs go black and LED IR lamps take over for black out mode. No such thing as common wiring, every lightbar has its own dedicated wiring to a terminal box mounted inside the vehicle. I dont plan to simulate a war but the components are worth far more than I paid for the system that was new in boxes. I will use what I need and sell the rest piece by piece on Ebay. Unfortunately I have not found a wiring diagram so I had to point to point the entire system and create one. To make it interesting at some of the connectors they change wire colors from one side to the other

More than likely the rfi is coming from the inverter and not the panel string. The issue is usually the "load" and not the solar panel that creates rf.

[peakbagger]

Any chance that application of ferrite core inductors on the lines might quiet things down? I do not claim to be RFI expert but I picked up military lighting system last year used on military vehicles that use HID ballasts and IR sources, at every lightbar has a HID ballast and some sort of IR LED control board with large ferrites cores attached around the cables before they exit the lighting housings. My assumption is they are there to cut down the RFI profile of the vehicle and prevent interference to the electronics in use in the vehicle. The cores are readily available and cheap plus the split core versions are easy to install, just clamp them over the lines and use a UV resistant Tie wrap.

As a side note this lighting system appears to be designed to day light up a couple of blocks around a vehicle. One fixed lightbar with multiple heads on all four sides of the vehicle plus two motorized spotlights. Flip a switch and the HIDs go black and LED IR lamps take over for black out mode. No such thing as common wiring, every lightbar has its own dedicated wiring to a terminal box mounted inside the vehicle. I dont plan to simulate a war but the components are worth far more than I paid for the system that was new in boxes. I will use what I need and sell the rest piece by piece on Ebay. Unfortunately I have not found a wiring diagram so I had to point to point the entire system and create one. To make it interesting at some of the connectors they change wire colors from one side to the other