RFI From Inverter

[johnb2713]

So a quick update. Schaffner responded with ...."not enough info to make a conclusive determination on products to use" "pick some filters and try them" (my words) kind of reply. Really quite disappointing but it is what it is.
I was contacted by another radio guy who has the same RFI problem, with the exact same inverter model installed in a neighbors property, so I have been trying to help him and have shared a lot of info. He also went to Ofcom and got the same response as I from them.
I also spent a bit of time the other night looking through the installation manual and noted that on page 16 (see attached manual and photo). there is instructions to install "Magnetic Rings" (Listed on page 12 parts list) on both the Battery DC Cables, and the Battery Monitoring Cable prior to terminating them. I asked the guy if he could get some photos of his install and send them too me, only to find that NO MAGNETIC RINGS have been fitted during his neighbors installation. I have emailed the installer in my case asking if these have been installed. I shall leave it up to readers to determine if they think they will have been or not! I do not know what impact on the RFI would be with the potential omission of these Magnetic Rings in my case, however the manufacturer clearly felt that they were needed, and that is why they have been included in the kit.

The connections you have circled are the solar PV input, the manual is referring to the battery and BMS connections (neither wearing their ferrite rings). A slightly more serious problem is the CHNT double pole MCB, it looks like it's on the battery cables but isnt rated for DC operation, there is the potential for a fire if a fault occurs, you maybe should inform your neighbour of this. The installation is looking more and more like a cowboy job.

 

[Pwablo]

The connections you have circled are the solar PV input, the manual is referring to the battery and BMS connections (neither wearing their ferrite rings). A slightly more serious problem is the CHNT double pole MCB, it looks like it's on the battery cables but isnt rated for DC operation, there is the potential for a fire if a fault occurs, you maybe should inform your neighbour of this. The installation is looking more and more like a cowboy job.

Thanks for your reply. The photo is not of the installation concerning my case. It is of another case whom I am currently working with in tandem to mine to find a resolution. If you look at the Orange DC Battery Cables you will see they are clearly within the circle. As the PV Input cables are in front of the DC cables, it is a bit difficult to not encompass them in the circle. Thanks for the info regarding the MCB I shall pass this along to the other party.

 

[Bob Minchin]

Those ferrite toroids will need to be pretty beefy to avoid magnetic saturation from the DC flowing from the panels (10's of amps)and even more beefy for the battery leads handling several time the panel current. Once saturated their performance at RF will be greatly reduced. The sort of ferrites used in high power three phase motor drive inverters will possibly do the job.
It is a thorny problem you have. Electromagnetic compatibility is a mix of emissions and susceptibility as you will know. there is not such thing as "zero RF emissions" from an electronic device and your receivers will be exceptionally sensitive and impossible to filter out and in band signals and therefore susceptible. All you can really do is to make your front end extremely linear to reduce intermodulation product and have as narrow band filter as possible such as cavity resonators. If perchance the interference is harmonically stable from say fixed frequency PWM in the inverters then you might just be able to build notch filters to reduce the problem.
When I was working in electronics applied research and radio back ends needed micro controllers in them I devised a scheme for the digital clock signals to be multiples of the radio channel spacing which was a great help in reducing interference.
Keeping in touch with other Amateurs and RSGB publications should be helpful. As you have found, the authorities are of little help.
Good luck

 

[Pwablo]

Just a quick update. The Installer has replied stating that there were no Ferrite Cores supplied in the kit of parts with the inverter. They even opened a new inverter from stock and found no cores. I have asked them to get in touch with Sunsynk and enquire why they are in the manual but not included in the kit. Clearly the manufacturer at some point, identified a risk of RF leaking from these cables, and that is why there is instruction to fit suppression? The mind boggles, more questions than answers!!

 

[binky]

Just a quick update. The Installer has replied stating that there were no Ferrite Cores supplied in the kit of parts with the inverter. They even opened a new inverter from stock and found no cores. I have asked them to get in touch with Sunsynk and enquire why they are in the manual but not included in the kit. Clearly the manufacturer at some point, identified a risk of RF leaking from these cables, and that is why there is instruction to fit suppression? The mind boggles, more questions than answers!!

Chinese manuals.... I've fitted many solar and battery systems, ferrites are never supplied.

 

[Pwablo]

Well a quick update. I have all the EMI filters and Ferrite Beads at hand now, but a few more things have transpired over the past few weeks. The regulator called and informed me that they now have a 3rd case with the same inverter model, but this time it has been installed by another amateur radio operator. The manufacturer is apparently sending a modified inverter to the guy to test.
I also chased them up again, and they actually replied..!!!! saying that the previously identified ferrite rings for the Battery and BMS cables have been recently added to all their new inverter kits, however the one installed next door was prior to this change. They also endorsed fitting EMI filters and beads on DC and AC cables. Clearly they know full well that RF leak paths exist on their products. Will likely be a few weeks yet before I get any more progress. Will post back with results once/if I get across the finish line.

I also came across an interesting video, showing the effects of ferrite beads on common mode currents, link is there-------> Ferrite Bead Testing.
Basically it demonstrates the diminishing returns you get by adding clip on ferrite's....First bead is approx -10dB attenuation, 2nd -14dB, 3rd -16dB etc up to 7 beads which choked common modes to around -25dB.......

 

[Radiotowers]

Well a quick update. I have all the EMI filters and Ferrite Beads at hand now, but a few more things have transpired over the past few weeks. The regulator called and informed me that they now have a 3rd case with the same inverter model, but this time it has been installed by another amateur radio operator. The manufacturer is apparently sending a modified inverter to the guy to test.
I also chased them up again, and they actually replied..!!!! saying that the previously identified ferrite rings for the Battery and BMS cables have been recently added to all their new inverter kits, however the one installed next door was prior to this change. They also endorsed fitting EMI filters and beads on DC and AC cables. Clearly they know full well that RF leak paths exist on their products. Will likely be a few weeks yet before I get any more progress. Will post back with results once/if I get across the finish line.

I also came across an interesting video, showing the effects of ferrite beads on common mode currents, link is there-------> Ferrite Bead Testing.
Basically it demonstrates the diminishing returns you get by adding clip on ferrite's....First bead is approx -10dB attenuation, 2nd -14dB, 3rd -16dB etc up to 7 beads which choked common modes to around -25dB.......

Well done, nearly choked when I read Ofcom are doing something positive for a change. It will be interesting to see what difference can be seen.
Might be worth taking a bit of time to check all bands day and night again especially as the active sun is causing some radio blackouts attenuating some bands from propagated noise.

 

[Pwablo]

Well done, nearly choked when I read Ofcom are doing something positive for a change. It will be interesting to see what difference can be seen.
Might be worth taking a bit of time to check all bands day and night again especially as the active sun is causing some radio blackouts attenuating some bands from propagated noise.

Well I could debate all day if they actually did anything....I don't think their position has changed at all and they were just passing on the message. Was still nice to be remembered ....I did'nt expect to hear anything from them again. They also want me to let them how the mitigation works out and provide some data ..... I wonder if they will pay me for doing their blinking job for them ?

 

[Pwablo]

Well, out of the blue this morning I got a call from the Installers who called to say they had just installed the supplied ferrite rings from Sunsynk and could I check the radio for interference. There was an immediate improvement in the RFI but still a reasonable amount was present. I handed them the box with all the EMI filters and enclosures and they carried out the install.
I have attached the final install diagram showing the filters used, and where I though the clip on ferrite beads would work best, although I see that the installer just clipped them on randomly, but I dont think this is a problem.

The results are not perfect but there is an overall improvement in the RFI and it doesnt look like it is now as much of a hobby killer as it was before the measures were put in place. Its never going to be zero RFI again, which nobody should expect unless the inverter model is changed, but I think that if I move my primary antenna from its current location, which is only 20m from the source, and stick the mast on a tilt base another 20m away into the garden, then I may be able to reduce the RFI a little bit further.
Some before and after photos below off my radios spectrum scope, to show the changes in received RFI on each frequency. It is a pretty sunny day today, and the RFI moves about depending on the inverter load, so I will see what it is like during the night. It will take a few days of monitoring to really assess what the improvment is really like, but first impressions are that the RFI has reduced significantly on alternate peaks, as it is predominant every 60KHz now, where it was every 30Khz before. I can still hear the carrier tone when on side band every 30KHz, but there is not much in the way of a signal. After observing the changes as they were put in place, I now get a distinct feeling that the MPPT charge controller in the inverter may be the primary cause of the RFI, and believe that this would be virtually impossible to eliminate completely, unless design changes were carried out internally.


8Mhz Before

8Mhz After

10 MHz Before


10MHZ After

Continued on next Post

 

[Pwablo]

14MHZ Before

14MHz After

18MHz Before

18MHz After

21MHz Before

21MHz After

24MHz Before

24MHz After

28Mhz Before

28MHz After

Continued on next post

 

[Pwablo]

EMI Filters that were Installed are as follows.....

2 x Schaffner FN2200-25-33 1.2Kv DC 25 Amp High Performance Inline EMI Filter on each Solar Panel String
1 x Schaffner FN2410-32-33 250VAC 32 Amp High Performance Inline EMI Filter on AC LIne to consumer Unit

2 off FN2200-25-33 on Solar Panel Strings

1 off FN2410-32-33 on AC Cable to Consumer Unit

Sunsynk Supplied Ferrite ring on DC Battery Cables

Sunsynk Supplied Ferrite Ring and Ferrite Beads on BMS Cable

A Mass of Ferrite Beads on DC Panel Input cables



I hope that readers have found this useful, and that by documenting this sage for the wider public will go on to help others in the future who face interference issues with solar installations.


All the Best.

 

[binky]

nice tidy job. From what I remeber rightly, emf or any form of radio wave has a logorithmic 'power' scale. ie doubling the distance away sees greater than having it's strength, but I could be talking crap!

 

[Pwablo]

nice tidy job. From what I remeber rightly, emf or any form of radio wave has a logorithmic 'power' scale. ie doubling the distance away sees greater than having it's strength, but I could be talking crap!

Think your correct, I notice that when i rotate my antenna away from the source direction it drops the signal quite a lot, and is not visible on the scope , however the carrier tone is still detectable. I'm actually quite disappointed that the RFI has not been reduced more considering that every port going into the inverter has now had some serious suppression put on it. I'm pretty sure that the bulk of the noise is in fact on the charge side, and the DC cables would maybe benefit from another wrap and the cables tightened up around the ferrite core. But that said, 14Mhz thru 24Mhz still have a crap ton of RFI, so I doubt that any amount of mitigation will tame this inverter, and that the design and/or component quality needs to be addresses to have any chance of it being relatively noise free.

Ive contacted the manufacturer again and provided them with all these images so will see what they come back with. Perhaps nothing and I will have to live with what I have now got, but I would rather not. Either way they need to know.

 

[Geoff1946]

Is the only earth on the inverter through the mains cable?
If so, I wonder if short direct links between the inverter case and the filter cases would be beneficial.
This just intuitive ! I'm no EMC expert.

 

[Radiotowers]

Think your correct, I notice that when i rotate my antenna away from the source direction it drops the signal quite a lot, and is not visible on the scope , however the carrier tone is still detectable. I'm actually quite disappointed that the RFI has not been reduced more considering that every port going into the inverter has now had some serious suppression put on it. I'm pretty sure that the bulk of the noise is in fact on the charge side, and the DC cables would maybe benefit from another wrap and the cables tightened up around the ferrite core. But that said, 14Mhz thru 24Mhz still have a crap ton of RFI, so I doubt that any amount of mitigation will tame this inverter, and that the design and/or component quality needs to be addresses to have any chance of it being relatively noise free.

Ive contacted the manufacturer again and provided them with all these images so will see what they come back with. Perhaps nothing and I will have to live with what I have now got, but I would rather not. Either way they need to know.

That now looks better,
As I said previously with my level of noise I can easily work around any noise issues and so should you.
The noise reduction you refer to is the inverse square law where doubling the distance should reduce the intensity to a quarter.( I think)

I can see a nice article in Radcom being written soon .

Regards

 

[apprentice87]

Think your correct, I notice that when i rotate my antenna away from the source direction it drops the signal quite a lot, and is not visible on the scope , however the carrier tone is still detectable. I'm actually quite disappointed that the RFI has not been reduced more considering that every port going into the inverter has now had some serious suppression put on it. I'm pretty sure that the bulk of the noise is in fact on the charge side, and the DC cables would maybe benefit from another wrap and the cables tightened up around the ferrite core. But that said, 14Mhz thru 24Mhz still have a crap ton of RFI, so I doubt that any amount of mitigation will tame this inverter, and that the design and/or component quality needs to be addresses to have any chance of it being relatively noise free.

Ive contacted the manufacturer again and provided them with all these images so will see what they come back with. Perhaps nothing and I will have to live with what I have now got, but I would rather not. Either way they need to know.

Shameful you should have to do all this...

john...

 

[Pwablo]

Shameful you should have to do all this...

john...

Tell me about it John.

 

[SPECIAL LOCATION]

Thanks again for further updates to this thread...

Although I have no direct experience or knowledge I can input, it is a very interesting read..

And highlights a real problem many electricians may never be aware it existed!

 

[apprentice87]

Tell me about it John.

What about this??

 

[johnb2713]

What about this??

View attachment 16707

Playing devils advocate, could b) not apply to the telecoms equipment?