diy solar

diy solar

RFI issue

C

Cagire

New Member
Joined
Apr 3, 2023
Messages
28
Location
France
Hello,

I have a problem with RFI from my solar system. It's a cheap hybrid inverter (must pv1800) that can work on solar + batteries (offgrid) or work in bypass mode (from grid). I have massive RFI emissions from AC cables all over the house (I use AM radio for detection, I guess they act as antennas). As an amateur radio operator, it's a nightmare.

I put FT 240-31 ferrites with many turns everywhere, it helped, the inverter itself seems quiet now but the RFI comming from cables around the house (in the walls, ceilings ...) are still there. After hours / days of tests, I might have circled the problem, and I think it's a ground issue. Here are the results of some experiments:



rfi_problem.PNG

The only possible configuration for it to be quiet is when I turn OFF solar panels + disconnect the ground from the inverter. I feel like there is the solution behind these resuls but I can't get it.

I live in France, where the ground is tied to neutral in the grid and in the house (I have a ground spike).

Any thoughts ?



(Please don't tell me it's a crap inverter , I know that ... but I dont have 2 or 3k for a fancy one so we will deal with it)

Thanks
 
Last edited:
Show pictures of your ferrite chokes as you have them wound. You may need more ferrites. Another test is to disconnect grid, inverter out and PV one at a time and note change in RFI.
 
AntronX said:
Show pictures of your ferrite chokes as you have them wound. You may need more ferrites. Another test is to disconnect grid, inverter out and PV one at a time and note change in RFI.
Click to expand...

It's silent only when panels and ground are disconnected from the inverter. Grid connected or disconnected makes no difference.

Sorry for the mess in the cable but I'm testing so I need easy access ...

Battery DC + and - in the same toroid
1712698029847.jpg

Panels DC + and - in the same toroid (yellow is a test ground ...)
1712698029851.jpg

AC in , out and ground in the same toroid
1712698029871.jpg

I even added a filter on AC out :
1712698029856.jpg

And more 31 ferrites on AC in and out
1712698029842.jpg

Sounds like it is coming from a high frequency inverter. What kind on inverter are you using?
Click to expand...

Must PV 1800
 
Try this. Move PV side choke to inverter side and wind PV choke with 2 toroids. Wind 30 turns of ground wire on separate choke as shown.
rfi_problem.PNG
 
Traditional solution is to start with bonding and shielding and grounding. Wider conductors offer lower impedance to rf-that’s why you see those wide copper straps bonding cabinets together. The shield from big,old coax works well.

Put as much as you can in steel conduit. It adds a ton of impedance from the magnetic field, and high capacitative losses from the wires laying against the metal next to each other.
 
Save yourself some time and get rid of the cheap inverter. Go with a good low frequency inverter if you are really troubled by it.
 
Well well well. Guys, I think I won. Maybe not the war but definitly a major battle.

I made new filters as suggested by AntronX and it worked. I would say 80-90% RFI reduction.

Here is the new way:

rfi_problemv2.PNG
Note that I removed the ground from the inverter because I needed to bond the neutral to the ground anyway in offgrid mode, which I do with a relay) :

Filters with 2 x ft240-31 at DC-panels-IN and AC-out
1712759705292.jpg


Filter at ground
1712759673186.jpg

I went from S8 QRM to S1 at 7 and 14 Mhz which is already more than good.
3.5 mHz seems still a bit polluted but I need to wait for more testing to make sure it's not just atmospheric QRM.


Next step : I will rewind or move the toroid from the panels array and see what happens.
 
Last edited:
Toroid at panels does almost nothing that's why I suggested to move it somewhere useful. Try adding double toroid to grid input side. What happens if you connect ground to inverter?
 
AntronX said:
Toroid at panels does almost nothing that's why I suggested to move it somewhere useful. Try adding double toroid to grid input side. What happens if you connect ground to inverter?
Click to expand...

I also realised that panels do not create RFI. I have the same amount when panels are plugged and when they are not (underground cable only). It's really the (long) cable itself when connected to inverter that is making all that mess.

When I connect inverter ground to house ground it adds RFI. However, the neutral of the inverter seems already connected to the inverter ground chassis (I checked with multimeter). Then the neutral itself is bonded to house ground (to avoid floating neutral), so I guess it's fine for safety.
The grounding is very cheeky, if I change "the order" of my cables, some RFI comes back. I suspect some bad "loops" occuring here and there.

For now it's like this :

1712772841460.jpg
 
Why would you filter AC input ? I moved it out from the equation since it's a clean signal ...
 
Is that a solid state relay you have connecting post EMI filter neutral to ground ? Those have capacitive coupling and leakage current between in/out and don't block RFI. You need mechanical relay for that.
 
This noise I believe is a 15Khz sweeping freq and it comes from the MPPTs.
It is not a "Cheap" inverter thing, it happens with Sol-Arks and other expensive brands.
Once the Sun goes down the noise is gone.
 
AntronX said:
Is that a solid state relay you have connecting post EMI filter neutral to ground ? Those have capacitive coupling and leakage current between in/out and don't block RFI. You need mechanical relay for that.
Click to expand...

Yes it is. I don't understand what you mean by blocking RFI. It is just a switch aiming at bonding neutral and ground in offgrid mode, it has no RFI purpose. I'm confused.


robby said:
This noise I believe is a 15Khz sweeping freq and it comes from the MPPTs.
It is not a "Cheap" inverter thing, it happens with Sol-Arks and other expensive brands.
Once the Sun goes down the noise is gone.
Click to expand...

Noise was also there even with the connectors from panels disconnected = without any DC voltage through the cables.
 
A quick story of a recent experience I had. Not active on HF but had a terrible spur on a local GMRS repeater at 462.625 mhz. It went away when the sun set so I started going all over the two solar power systems I have here. Through a process of elimination, nothing I de energised helped. Was leaning toward the SolarEdge optimizers because they are the only thing I can't actually shut off which would have been a worst case scenario.

Was listening to some traffic on the repeater one night a few weeks ago, and everything was fine. Then the warden went downstairs to get something and Boom...the noise was back so strong it made the repeater inaudible. I asked what she did and was told she had set something on my desk and bumped the computer mouse. The light bulb went off in my head, all my computers and monitors go in to power saving mode an hour or so after I quit working for the day. When she moved the mouse, my off network computer woke up. Turned out to be an older Dell monitor. I replaced it and problem was solved. Drove me crazy for a long time because my schedule mimicked a solar device.
 
You must log in or register to reply here.

Similar threads

7
Best equipment for ham radio (no rfi interfere)
2 3 4
Replies
76
Views
4K
73powerstroke
7
G
Wiring Sungold To Sub-Panel
Replies
2
Views
136
G-Falco
G
ericfx1984
A chance to test my system before I actually start using it
Replies
7
Views
223
ericfx1984
ericfx1984
V
Advise needed for ground and neutral bond, thank you
Replies
5
Views
209
vietxtlife
V
Unwin
EcoFlow Delta Pro Ultra issues
Replies
1
Views
1K
AZ Solar Junkie
A
Back
Top