eHam
eHam Forums => RFI / EMI => Topic started by: AK4YA on December 06, 2022, 07:58:19 PM
-
I have a fairly large solar system (12KW of panels on a 12KW inverter) and its definitely making a ton of RFI noise in a many of the RF bands.
Got a bunch of the mix75 (Palomar's suggestion is to use the very low LF ones to choke out the actual source of the noise) from mouser:https://www.mouser.com/ProductDetail/Fair-Rite/0475176451?qs=%252B6g0mu59x7I%252BCmoSR0e6cg%3D%3D&countryCode=US¤cyCode=USD (https://www.mouser.com/ProductDetail/Fair-Rite/0475176451?qs=%252B6g0mu59x7I%252BCmoSR0e6cg%3D%3D&countryCode=US¤cyCode=USD).
Very strange though, is that as soon as the ferrite is snapped around the AC main breaker inside the inverter then that breaker starts hissing and sounds like arcing and sparking. I cant see or smell anything bad there but its definitely there and LOUD. Heres a pic of the inside of my inverter, the ferrite I just installed that is causing the problem is circled. next to it are the AC breaker for the generator and then the critical loads panel.
https://i.ibb.co/02DkPvq/20221206-204644.jpg (https://i.ibb.co/02DkPvq/20221206-204644.jpg)
I did verify the length of the wire leads there just under the breaker. That is definitely not the problem, they are making solid contact within the breaker even with the ferrite there. I get the same thing with other ferrites. The ferrites have a solid plastic casing all around, nothing is getting shorted. Literally AS SOON AS I feel the ferrite halves making contact with each other is when that breaker goes all hissing.
Not sure where to go now.
-
You generally do not want to place the two legs of a split phase circuit through the same ferrite closed form due to the imbalance of AC current and RFI current between the two legs.
A better solution would be to wind each wire onto its own ferrite rod. This may necessitate extending the wire length in order to achieve sufficient turns.
- Glenn W9IQ
-
My recommendation is to use ferrites for common mode and a line filter for differential mode. The way you have the ferrite over only two legs of the 3-phase, is not common mode and exposes the ferrites to large low-frequency currents which is causing the problem. The inverter probably does not like the inductive load at all, and the cores are probably heating/saturating. For a better way to do it, see figure 4 of my April, 2016 QST article (https://www.fair-rite.com/wpapers_anotes/can-home-solar-power-and-ham-radio-exist/ (https://www.fair-rite.com/wpapers_anotes/can-home-solar-power-and-ham-radio-exist/)), which shows the solution for two phase AC connection. It can be extended to three phase by using a three phase line filter. Note carefully the fact that the enclosure does not provide a ground path outside the common mode chokes - ALL conductors must pass through the INSIDE of the chokes.
Other notes:
Solid cores are much better than snap-ons. They have higher impedance (due to no air gap) and they are much more permanent as the plastic snap housings can 'creep' over time and allow the gap to open up.
Mix 31 is preferred as it does a better job of covering the HF-VHF spectrum than 75. 75 is ideal for MF such as 160 meters but tails off in the HF region.
-Tony, K1KP
-
its not 3 phase, its standard 240v split phase service. The 3 big breakers you see arent 3phase, they're separate breakers for AC mains, generator, and critical loads, and each has its own ferrite. I see from another post though that I should use 6 separate ferrites there.
and are you saying that also the ground wire must also have a choke? (not just the hot and neutral?)
Ill take a look at your stuff though thanks
-
the guy from palomar said not to worry about the ground wires, to just use the ferrites for the black and white wires.
also he really suggested the 75 hard, well over the 31 since he said it was better to cut the source of the noise itself which should then automatically take care of the higher odd harmonics. In my case all the odd harmonics are spaced about 29.3KHz apart.
-
the guy from palomar said not to worry about the ground wires, to just use the ferrites for the black and white wires.
also he really suggested the 75 hard, well over the 31 since he said it was better to cut the source of the noise itself which should then automatically take care of the higher odd harmonics. In my case all the odd harmonics are spaced about 29.3KHz apart.
When implementing a common mode choke, you want to pass all the current carrying conductors through the ferrite. If you do not do this, the core will also carry flux from the desired differential mode current instead of the intended common mode current alone. Closed form inductors can become saturated which will lead to reduced performance and sometimes their destruction.
The normal operating (differential) currents in a US split phase circuit are carried in the two hot wires and the neutral wire. These must all pass through the core of a proper common mode choke. The ground wire does not normally operating carry current but it can in the event of a neutral fault or other fault. It can also be a source of radiated RFI. For these reasons, it is wise to include the ground in the common mode choke.
Since you are only passing the two hot wires through the ferrite core, the differential current is not balanced in your clamp -on common mode choke. The noise you hear may well be from the ferrite assembly itself - hobbyists sometimes call this "coil whine". It can occur as the core nears or passes into saturation.
- Glenn W9IQ
-
In my application though there really is no split phase, at least not inside that inverter there isnt.
Both of those 3 breakers there are separate 240vac circuits. Now, that 3rd one is the feeder for my critical loads panel, and that panel IS setup to split phase off of that 3rd breaker though to feed all my 120v critical load breakers.
Sorry for being dense, but Im still not understanding your post at 7:00:03AM (CST), are you saying that in my case when I clamp on the ferrite over the wires going into that 1st breaker and it starts hissing, that it should stop if I also get the ground wire in the ferrite as well?
-
In my application though there really is no split phase, at least not inside that inverter there isnt.
Both of those 3 breakers there are separate 240vac circuits. Now, that 3rd one is the feeder for my critical loads panel, and that panel IS setup to split phase off of that 3rd breaker though to feed all my 120v critical load breakers.
Sorry for being dense, but Im still not understanding your post at 7:00:03AM (CST), are you saying that in my case when I clamp on the ferrite over the wires going into that 1st breaker and it starts hissing, that it should stop if I also get the ground wire in the loop as well?
If you trace the connections from your inverter to your meter/mains, you should find two hot wires (generally red or red/black), a neutral wire (white) and a ground wire (bare or green). The neutral and ground wires will not pass through the breakers, they will be hard wired to the meter/mains connection. If in your picture the mains connection is the lower left conduit, then you can see the bare ground wire and the white wire wire near the conduit fitting.
The most important wire to add to your core assembly is the neutral wire. Ideally, you also wish to include the ground but that will not fix the present problem without the neutral being included.
- Glenn W9IQ
-
ah, I got it now, thank you very much.
now I just need to figure out if its worthwhile to rip all the wire our and redo it with longer lengths to support wrapping of the neutral and ground.
or maybe better to just get some massive toroids and put them around those waterproof entry conduits
-
Better yet, I would get a scope, voltmeter, and freq counter and measure what is causing the problems. Then you could develop a plan to cure the problems. Much better than guessing!
-Mike.
-
Well I was able to put the red, black, and white wires of that 1st AC breaker there through a ferrite, and its still buzzing.
I did some checking with a well insulated glove, and 100% the buzzing is indeed coming from the ferrite. theres no vibrations from the breaker, and vibrations on the wires getting stronger the closer down to the ferrite I get, with solid vibration feeling on the ferrite itself. I suppose I could try to get the ground through there as well...
Better yet, I would get a scope, voltmeter, and freq counter and measure what is causing the problems. Then you could develop a plan to cure the problems. Much better than guessing!
-Mike.
that sounds fun.... I can prob get some low cost equipment for this, where would you start?
-
You need, in order:
1. A much larger cross section ferrite.
2. A solid, not a split, ferrite.
3. To wind the wires of each side in an opposing direction so the current from each phase causes a magnetic field in the opposite direction of the other one.
Grover
-
I don't know about "low cost".
However, www.circuitspecialists.com has some quality stuff you can check out. They used to advertise in QST. (not affiliated with them).
-Mike.
ps: Xmas is coming!
-
OK I got all 4 wires through and the buzzing has stopped!
Also, my RFI is gone (at first glance it is, ill check more across the bands but 40m was the worst and now all those tones out there that were all roughly 29.385 KHz apart are gone on 40m!)
Now, it may have not been that 1 single ferrite on that 1st AC breaker in the picture that solved it all, as also today I put a buttload of those ferrites on the DC solar wires and generator cable that are not shown in this picture.
You guys are terrific!
-
OK I got all 4 wires through and the buzzing has stopped!
Also, my RFI is gone (at first glance it is, ill check more across the bands but 40m was the worst and now all those tones out there that were all roughly 29.385 KHz apart are gone on 40m!)
Now, it may have not been that 1 single ferrite on that 1st AC breaker in the picture that solved it all, as also today I put a buttload of those ferrites on the DC solar wires and generator cable that are not shown in this picture.
You guys are terrific!
Glad you solved it. The ground wire is often the dumping ground for line filtering circuits which can lead to substantial RFI by conduction or radiation.
- Glenn W9IQ
-
I should add that you may want to investigate a potential faulty or shorted to ground neutral connection on that circuit. It would be unusual for the ground wire to be required to bring the common mode filter into balance.
- Glenn W9IQ
-
Well now the ferrite with all 4 through it has started buzzing real loud. I cant find anything wrong inside that inverter - but the electrician who wired the home and solar all up for me is coming by next week.
-
You may want to run a cable tie or two around the core so that the gap is minimized and regulated.
- Glenn W9IQ
-
making it really tight with a ziptie makes it worse, which is pretty disconcerting.
now, its a pretty big mix75 ferrite though
(https://www.mouser.com/ProductDetail/Fair-Rite/0475176451?qs=%252B6g0mu59x7I%252BCmoSR0e6cg%3D%3D&countryCode=US¤cyCode=USD)
-
What is the Fair-Rite part number of the core you are using?
- Glenn W9IQ
-
odd- I could have swore I put the part# and link on the previous message.
Its a Fair-Rite 0475176451 https://www.mouser.com/ProductDetail/Fair-Rite/0475176451?qs=%252B6g0mu59x7I%252BCmoSR0e6cg%3D%3D&countryCode=US¤cyCode=USD (https://www.mouser.com/ProductDetail/Fair-Rite/0475176451?qs=%252B6g0mu59x7I%252BCmoSR0e6cg%3D%3D&countryCode=US¤cyCode=USD)
-
I suspect that the AC current through the core still is not balanced. You may want to ask your electrician about that when he or she is there.
Another possibility is that the switching harmonics are so strong that their current is saturating the core. Unfortunately there isn't a larger type 75 core available. You would need to switch to type 31.
- Glenn W9IQ
-
If it was the latter then would using a bunch of those larger mix75 toroids just outside the enclosure work?
-
Yes, that can help. The question is how many? Those cores are not inexpensive.
- Glenn W9IQ
-
Perhaps you can conduct a temporary experiment by putting the three cores that are on the three protected circuits onto the one troublesome line.
- Glenn W9IQ
-
I had 3 extras of those big ones so I was able to leave the 1 there in the inverter and put the others in the AC mains panel where that same inverter line enters it. Its maybe 8 ft total of cable as they are almost directly next to each other. I was able to get all 3 new of these big clamp on ferrites around all 4 conductors as well.
The buzzing is not as bad now but theres basically zero power going through (heavy clouds all day) right now. so the buzzing is really at a min. Cannot feel it but can still hear it if I put my ear close enough. Also the RFI trash is still there albeit subdued, but thats likely due to the low solar output instead of the new ferrites.
-
Any update after some sunny days?
- Glenn W9IQ