I work as a radio engineer and have Victron 10/50 MPPT, shunt and Cerbo GX deployed at a marine VHF remote helicopter access repeater site for remote monitoring via a LTE router.
Since a recent upgrade, we have been plauged by a RFI source destroying the site noise floor and by association the effective radio coverage area.
Isolating the DC to the Cerbo GX stopped the source of radiated interference, all other inputs were removed with no effect. Torriods and ferrite sleeves on all associated wiring were tried with no improvement. The spectrum analyiser screen shots below confirm the radiated signal and the level was surprising. Approximately 530 kHz appart in a re-ocurring comb.
If any one can contribute a solution or mechanisim, it would be greatly appreciated.
OGPS
(Ed @ Off-Grid Power Systems - offgridps.com)
2
Yikes. I’m a radio amateur and I was more concerned about HF interference in my RV. The MPPTs were my worst culprit.
WiFi and Bluetooth are operating at 2.4GHz so I would not expect them to be the cause, but just in case, disable the internal AP, disconnect from any WiFi AP’s you’re using, and disable the BT adapters. Does anything change?
Are you using a hardwired ethernet connection with the LTE router? If so, disconnect the ethernet connection and check the SA. I’m guessing this is the source. If so, use shielded ethernet patch cables and see if that helps. If not, use some isolating ethernet filters and/or multiple turns through a mix 31 torroid on each end of the ethernet patch cable if it’s “long”, or just one in the middle if it’s “short”.
Your plot from the SA looks just like ethernet noise to me. You’ll need to diagnose further to see if it’s conducted, radiated, or both.
Please let us know what you find. There aren’t a lot of RFI discussions on this forum so it would be good to have the resolution documented for future users.
Many years ago when I was involved in radio design, I used a small probe to look for conducted emissions. This was made with 4 turns of thin wire about the end of a small piece of pcb, and with a shielded connection to the analyser input.probe end was about 1cm x 1.6mm, and proved useful for localising emissions.
OGPS
(Ed @ Off-Grid Power Systems - offgridps.com)
4
H probes. Very handy. You should have seen the RFI coming from the terminals of a SmartSolar 150/100. Connected to wiring in my RV it wiped out all of HF for 30 meters/100 feet. The ultimate fix? I went to a 48V system so I could use the much quieter 150/35 and 150/45 MPPTs
I’ve just resurrected an old M710 that has PLL problems, now just got to sort a reasonable antenna… AND then find any EMC problems… 2 x MP2 3kVA operating in parallel from 24V..
Did one install with a Chinese inverter, the MPPT on that wiped out the entire MW band.
OGPS
(Ed @ Off-Grid Power Systems - offgridps.com)
6
I have found the Multi’s/Quattro’s to be RFI quiet. At least 3kVA and 5kVA units. I haven’t “played radio” around any 10k’s to know. The Orion’s haven’t given me any troubles I couldn’t remedy. The MPPTs are a different matter. Use 24V or 48V batteries and small MPPTs and it works pretty well, but there are still some birdies to work around on HF. If I’m trying to do some weak signal work on HF I might still have to turn off an MPPT or two during the contact. But I’m talking an S2 signal with an S0 noise floor. I don’t have any noticeable issues on VHF or above.
