Weird wideband noise appearing after removing other RFI source.

[Lukasz Olszewski]

I've been plagued with PV inverter noise on upper bands like 10m/15m and to a lesser degree 17m and then 20m. I've ended up doing quite a bit of testing to confirm the source is my 3 cheap hybrid (battery + on grid) inverters.

I bought lots of ft240-31 and ft240-43 cores and I ended up using 7 per inverter which took the noise as received by an internal antenna down by 15dB and almost gone on my external antennas. During all of this I was focusing on upper bands.

The noise floor I receive with my multiband fan dipole (at 6m, fed by 35m od h155 coax, two chokes right before the shack) on most bands was about -120dBm as shown in pihpsdr.

However , after all this work on 40m/20m and a little on 15m it jumped to about -100dBm/-110dBm depending on the portion of the band(it seems on 40m it's one continuous region, on 20m there are two "humps)"). I confirmed it is not an internal source, because it goes away when I switch to the inside antenna. Also I tried removing all power from these inverters and my poe switch.It made no difference. 

One thing that changed is that I recently bought and installed a new 24 port MikroTik 10gig +Poe switch. But it is installed in the house 50m away and no cables connect it to the shack! I might be able to power it down to test, but i doubt it has anything to do with it. Most PoE rfi I saw before was not super flat across the entire band.

Very interesting is the fact this noise disappears and I go back to my original -120dBm background if I enable NB noise block filter. I never noticed much of a difference made by that filter ever before. So somehow it is very good removing this kind of wideband noise. It is so good it reveals many signals that otherwise would be gone in the noise. I tried every single combination before and that filter never did that before!

So now I wonder,is there some built in RF AGC in addition to our RX gain knob and post demodulation AGC? There must be.... Because where would that noise come from? I have no other explanation.

Can someone more knowledgeable about the chip explain this?

Also, do you live in a quiet rural location and use a dipole to receive? How does your band noise look like on 40m and 20m.

73, Lukasz 

[Rock Dubois]

Hi Lukasz,

I  live in a small town in Arizona, USA at 5100 feet ASL. My noise level on 20m is -83dBm and on 40m it is -90dBm. However there are times and days where

I have S6 to S7 noise on 40m.

73, Yvon

AE7YD

[Steve Haynal]

Hi Lukasz,

There is no builtin AGC in the HL2. I tried a hardware-based AGC at one point but was never happy with the results. Software has the option of implementing AGC. Your noise must be very regular man-made since the noise blanker is so effective.

I run my HL2s exclusively on a small off-grid solar power setup. I use a Victron SmartSolar MPPT 100 | 20 charge controller. I have not seen RF noise issues with it.

It is a 24V system so I do regulate down to 12V for the HL2 with a switching converter. Many cheap down converters I tried do introduce noise. I settled on this down converter:
  https://www.ti.com/tool/LMR14030SEVM

modified to generate 12V output. I bought the EVM when first working on the HL2. I'd love to find an off-the-shelf inexpensive down converter based on this design. I do use some ferrite on both sides of this converter but it is pretty quiet as is. I think it also provides some isolation between the radio and solar power system.

The ethernet cable to the HL2 can also introduce noise, especially if it is a long run and goes up to a remote HL2 which is significantly above the ground. You want to use shielded ethernet cable and also put some ferrite on each end.

I used to use an inexpensive gigabit switch but found that was introducing some noise after I switched to a dedicated ethernet card for my HL2s in the host computer. The dedicated link also helps reduce lost and out of order packets.

https://www.ebay.com/sch/i.html?_nkw=593743-001+HP+NC365T&_sacat=0&_from=R40&_trksid=m570.l1313

The fluorescent lights in my garage always produce a lot of noise when on.

You may need to individually turn off every breaker in your house to determine the noise source while your HL2 is running on battery and connected to a laptop.

73,

Steve

kf7o

[Ron Lewkowicz]

I assume you are completely off grid so this likely doesn't apply in your case but I have noise on 40m that also is completely removed by the NB2. It was intermittent and looked like a hump between say 7110 and 7140 and would be 10dB or more over the normal noise floor.  It's still there but generally only perhaps 2-3dB above noise floor.  NB2 takes it right out but if someone throws a strong carrier or other strong signal on the band it resets the NB algorithm I think and the noise is back until the strong signal goes away.   What some of the local Hams are finding here is that the lightning  arrestors being installed with new recloser installations on the power lines are noisy and that noise can travel  some distance on power lines.  They have been using Ultra-Sonic receivers to find the bad poles and get them repaired. 

[Lukasz Olszewski]

Thank you for replying Rock.

On Sun, 23 Feb 2025, 19:23 Rock Dubois, <rockdub...@gmail.com> wrote:

Hi Lukasz,

I  live in a small town in Arizona, USA at 5100 feet ASL. My noise level on 20m is -83dBm and on 40m it is -90dBm. However there are times and days where

I have S6 to S7 noise on 40m.

73, Yvon

AE7YD

Very interesting. I'd love to also hear similar info from others.

73, Lukasz 

[Clifford Heath]

Folk here might find this article enlightening. I certainly did. It convinced me that I didn’t want to do this, so I installed a passive solar array instead of this microinverter crap:

<https://www.dropbox.com/scl/fi/g1y4sfwvrv9uq0kn4zqr5/QST-Amateur-Radio-Solar-Power.pdf?rlkey=9a2cwcoy64ps4tm5786aq9mnm&st=wqxtkhiq&dl=0>

Clifford Heath.

--
You received this message because you are subscribed to the Google Groups "Hermes-Lite" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hermes-lite...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/hermes-lite/CACwB4R7oq0%3D%2BmRK-icpviB3-FhaB-6y3pe%2BG3e6dx%2BtmLnVzYw%40mail.gmail.com.

[Lukasz Olszewski]

On Sun, 23 Feb 2025, 21:57 Steve Haynal, <softerh...@gmail.com> wrote:

Hi Lukasz,

There is no builtin AGC in the HL2. I tried a hardware-based AGC at one point but was never happy with the results. Software has the option of implementing AGC.

Thanks for clarifying. This makes it even weirder. No extra devices were added when that noise appeared. (other than the pretty expensive L3 managed switch I mentioned before, but I excluded it already by cutting power to my entire "server room" with the noise being unchanged). The nearest neighbour is 300m away. It is possible he got "something" new.

Your noise must be very regular man-made since the noise blanker is so effective.

Indeed, it is very effective , but on occasion it has problems(especially on 40m). Today it didn't work at all on 40m. I suspect due to the very strong pirate/mil signal at the middle of 40m.

I run my HL2s exclusively on a small off-grid solar power setup. I use a Victron SmartSolar MPPT 100 | 20 charge controller. I have not seen RF noise issues with it.

The noise issues I saw with my solar were always limited to 10m (and a bit at 600kHz, but I used filters to cut that out). 

Unfortunately if I chose Victron gear to build the system I'd have to sell my house and land to be able to afford it.... It's a 30kW of inverters 17kW of solar, 63kWh of battery (lifepo) storage and a 7kW auto start diesel generator as backup system. The inverters are mid-range Chinese brand (Y&H). I never had RFI issues with them other than previously mentioned at 10m. Even that could be worked around as it was a slightly smeared carrier every hundred kHz or so. So I could use the portions of the band inbetween. The signal strength was 30dB over the noise floor inside the shack ~20dB immediately outside. It was undetectable to me at the edge of my property some 60m away.

It is a 24V system so I do regulate down to 12V for the HL2 with a switching converter. Many cheap down converters I tried do introduce noise. I settled on this down converter:
  https://www.ti.com/tool/LMR14030SEVM

modified to generate 12V output. I bought the EVM when first working on the HL2. I'd love to find an off-the-shelf inexpensive down converter based on this design. I do use some ferrite on both sides of this converter but it is pretty quiet as is. I think it also provides some isolation between the radio and solar power system.

The ethernet cable to the HL2 can also introduce noise, especially if it is a long run and goes up to a remote HL2 which is significantly above the ground. You want to use shielded ethernet cable and also put some ferrite on each end.

Yes, I confirm I found the same, but for me a ferrite on one end took care of it (the end that goes into hl2).

I used to use an inexpensive gigabit switch but found that was introducing some noise after I switched to a dedicated ethernet card for my HL2s in the host computer. The dedicated link also helps reduce lost and out of order packets.

https://www.ebay.com/sch/i.html?_nkw=593743-001+HP+NC365T&_sacat=0&_from=R40&_trksid=m570.l1313

Fun story, after I bought the current switch the next day my wife said to me "the TV is not working". We have DVB TV here (terrestrial digital) on 6 muxes that provide some 50 channels or so (all around 400- 600MHz). The TV antenna is in the attic right in the middle of my "server room". 

I thought, that can't be my new switch because I installed it two days ago and the TV was fine yesterday. So I went to rescan channels and I found only muxes above 550MHz working. Nothing below. I then had an idea to switch off a pc that uses 10 gig port on the switch and channels returned. You know what was the reason? 

A plastic female to female coupler that joins two LAN cables. I have another server pc that also uses 10 gig and the cable is right next to the other one. That one (using a continuous shielded cable) makes no difference. The channels came back permanently when I took the link speed down to 5gig for the pc until a properly shielded coupler arrives.

(I did cut power to the switch to verify it has nothing to do with this ham bands issue).

The fluorescent lights in my garage always produce a lot of noise when on.

All my lights are LED in the house. Lots of small gu10 7W lights (a bit under a hundred in the entire house). I haven't noticed a problem (yet?) Perhaps because they are all better brands.

You may need to individually turn off every breaker in your house to determine the noise source while your HL2 is running on battery and connected to a laptop.

I've done that before, but this is new. I may have to do it again.

It is really is a massive pain to have to cut power like this  I've built an elaborate system so power cuts don't affect me and here I'm cutting power myself every couple of weeks. It really sucks the cheap crap made these days can't not emit RFI. Sadly buying expensive stuff is not enough to assure this problem doesn't happen. Often the expensive stuff is the same as the cheap variety just in a nicer box.

73, Lukasz 

[Lukasz Olszewski]

On Sun, 23 Feb 2025, 23:40 Ron Lewkowicz, <ron.le...@gmail.com> wrote:

I assume you are completely off grid so this likely doesn't apply in your case but I have noise on 40m that also is completely removed by the NB2. It was intermittent and looked like a hump between say 7110 and 7140 and would be 10dB or more over the normal noise floor.  It's still there but generally only perhaps 2-3dB above noise floor.  NB2 takes it right out but if someone throws a strong carrier or other strong signal on the band it resets the NB algorithm I think and the noise is back until the strong signal goes away.   What some of the local Hams are finding here is that the lightning  arrestors being installed with new recloser installations on the power lines are noisy and that noise can travel  some distance on power lines.  They have been using Ultra-Sonic receivers to find the bad poles and get them repaired. 

Interesting. I'm not entirely off grid. I have a grid connection and a small on grid pv system that sells power there. However,I don't use the grid for the huge majority of the year.

Very interesting about the strong signals messing NB.

There is a power line 300m away. Maybe they installed something new...

Regards ,

Lukasz 

[Lukasz Olszewski]

On Mon, 24 Feb 2025, 01:35 Clifford Heath, <cliffor...@gmail.com> wrote:

Folk here might find this article enlightening. I certainly did. It convinced me that I didn’t want to do this, so I installed a passive solar array instead of this microinverter crap:

<https://www.dropbox.com/scl/fi/g1y4sfwvrv9uq0kn4zqr5/QST-Amateur-Radio-Solar-Power.pdf?rlkey=9a2cwcoy64ps4tm5786aq9mnm&st=wqxtkhiq&dl=0>

Clifford Heath.

This is very useful . I too stay away from micro inverters, optimisers etc. It is far easier to fix one box than 40.

This article was my inspiration for buying all these fair rite cores :-)

73, 

Lukasz

[Lukasz Olszewski]

I now think the noise is not new, it was always there I just never noticed it... How?

I run pihpsdr on another computer I use occasionally and it started at 40m with NB on and this noise is only visible there at the very edges of the panadapter (this is how NB removes it usually).

So it seems the noise is not new. It may be the nearby power lines or I have no idea what.

Also I checked my old Kenwood TS-690 I've used for many years on which I remember mostly about S3 level of noise on 40m (unless some space disturbance is ongoing). So I fired my old Kenwood and I noticed AIP was on (automatic impulse filter). When I switched it off the noise went to S9.

So I guess I'm lucky I have the kind of noise that responds well to noise blocking mostly.

I too observed that the NB fails briefly when wideband strong signals are present.

I have to make myself a ferrite Rod antenna and walk around to see if I can locate the source. The power lines few hundred meters away are medium voltage (few kV) so I'm not sure these can be the reason. There is also a transformer. So faulty lightning protection can be a possibility .

However I've now established it is not an issue of Hermes Light 2. 

Thank you for all the replies.

73, Lukasz 

[G4ZAL]

Two years ago I had a solar system professionally designed and installed using Victron MPPT 250/100 'Smart controller' and Multiplus 48/5000 inverter.
I have 14 x 385w panels feeding the smart controller into a 10kW battery storage system.

It has the capabilty to be 'off grid' and to feed excess back to the grid and it does all this seamlessly.
That much I like about the system.

To my horror, the interference was terrible from the smart controller.

My installer was next to useless in understanding the problem.

This is what it looked and sounded like initially...

https://www.youtube.com/watch?v=y3HDqF1V2gI

https://www.youtube.com/watch?v=CQWs0HW-7Jk

The inverter also makes a little noise on 17m when it's working hard, but that is tolerable.

I gave up on my installer and pretty much fixed the system myself.

1) Twisted the solar cables from the roof panels to the smart controller - helped a little

2) Put some big ferrites on the solar cables as close to the controller as I could - helped a little ( https://www.hamgoodies.co.uk/problem-solver-ferrite )

3) The improvement that made a huge difference was to add some 2kv ceramic disk capacitors across the solar DC cables and from the positve to ground and the negative to ground as close as possible to the smart controller.
Those caps costing pennies has totally transformed the system from a noise point of view and I can now live with the odd few spikes on some bands.  Thetis' ANF and MNF can easily take care of them.

It's the solar panel cables that are acting like antennas for the pulsed switching 'noise' the smart controller is generating as it tries to maximise the voltage/current being fed to the system.

Those caps provide a path to ground for the switching RF noise.

Just don't ask me for a recommendation of Victron 'smart' equipment ;-(

Nigel, G4ZAL

[Lukasz Olszewski]

I'm now convinced this noise may be natural or if manmade is not local.

I've tried switching off every single breaker in my house.Also  the shack and other buildings I have here. The noise stayed the same.

I also built a little ferrite antenna (5 turns or so, got glued to 140mm ferrite rod - no idea thd grade but the resulting inductance was about 4uH) and an old receiver's adjustable capacitor (up to 370pF). Normally these rod antennas have pick up coils. This time I just connected it to the Mini SA ultra direct.

When walking around I'd find the noise  near metal structures. When I walked into a middle of an empty field (muddy and wet, average ground, 50m from a metal fence and 150m from a 5kV electric line) it was undetectable until I raised the antenna and the SA as far above my head as I could. Then I did pick it up but i couldn't establish any directionality. Every time I found a null it wasn't there the next turn around.

I think 

On Mon, 24 Feb 2025, 10:31 G4ZAL, <devon...@gmail.com> wrote:

Two years ago I had a solar system professionally designed and installed using Victron MPPT 250/100 'Smart controller' and Multiplus 48/5000 inverter.
I have 14 x 385w panels feeding the smart controller into a 10kW battery storage system.

It has the capabilty to be 'off grid' and to feed excess back to the grid and it does all this seamlessly.
That much I like about the system.

I too like my patchwork system as well. I can disconnect from the grid completely for 9 months out of the 12. And while I'm disconnected a tiny 2kW on grid deye is sending the power to the grid so I can get 80% of it back in winter (that's a very favourable deal early solar adopters like myself were getting some years ago).

To my horror, the interference was terrible from the smart controller.

It is funny how the part that is the least involved in making the powerful AC or DC often is the source of RFI.

73, Lukasz 

[Steve Haynal]

Hi Nigel,

Thanks for the 2kv ceramic disk capacitor tip. I will have to try it out. My system is pretty small. I have 2 x 250W panels, MPPT 100 | 20 controller and no inverter. I bought the panels used out of someone's car in a grocery store parking lot. I think they were warranty pulls from a professional installation. I use it with batteries from my 20 year old EV which died. It is a 24V system. With full sunshine, I will see at most ~450W. I use it to run my radios, charge some batteries and power some computers in the summer. It so happens that Chinese trucks use a 24VDC system so there are rice cookers and hot plates which run on 24VDC. I enjoy cooking food only with the sun on sunny days!

https://www.amazon.com/LElectric-Portable-Multifunctional-Steamer-Electric/dp/B095P2G2YN

https://www.aliexpress.us/item/3256807479426953.html

I do have cheap inverter which I use occasionally, but that does generate a lot of RFI. Perhaps I see less RFI since I rarely use the inverter.

73,

Steve

kf7o

[G4ZAL]

Hi Steve,

IIRC, I used 100nF (0.1) caps, 1 each across + to - & + to ground and - to ground on each 'string' close th the MPPT controller  - I have 2 strings of 7 panels each.

There are several complaints regarding the 'newer' MPPT controllers on the Victron forum (but no cure from the manufacturer) - maybe you got lucky with an 'older' version?

Mine are 50v 'Pylontech' batteries.

It's great when the mains power fails and our place is a little beacon of light in the area !!

I'd be interested if it makes any difference for you.

Regds,

Nigel G4ZAL

[Lukasz Olszewski]

On Sat, 1 Mar 2025, 22:40 G4ZAL, <devon...@gmail.com> wrote:

Hi Steve,

IIRC, I used 100nF (0.1) caps, 1 each across + to - & + to ground and - to ground on each 'string' close th the MPPT controller  - I have 2 strings of 7 panels each.

There are several complaints regarding the 'newer' MPPT controllers on the Victron forum (but no cure from the manufacturer) - maybe you got lucky with an 'older' version?

Mine are 50v 'Pylontech' batteries.

It's great when the mains power fails and our place is a little beacon of light in the area !!

I'd be interested if it makes any difference for you.

In my install I see no difference on MPPTs, but there is noticeable difference on the battery leads. Mind I already have a ft240-31 with 8 turns on the mppt wires so others may very well see a difference like Nigel.

I also have 3 ft240 cores (2*31 and 1*43) on the DC leads, but I still have two "smeared carriers" on 80m. Thankfully it's on frequencies I don't use often, but they exist. Interestingly only when it charges the battery but only so so.

If it really pushes the current (100A+ etc) there is no 80m rfi. If it only pushes in 20A that's when it is the worst.

If I touch one or both of the DC cables with a 10nF 2kV ceramic cap I can see it drop by 5-6 dB on a spectrum analyzer . Unfortunately that translates to only 3db at the entenna.

A capacitor connected between the plus-minus does nothing. Also if I have one already connected, adding another to the opposite polarity wire makes little difference (maybe 1dB). This sort of confirms it is mostly common mode current.

Also just for fun I asked one of the paid AIs I use to calculate what voltage rfi is on the cables (based on 10cm antenna being parallel, touching the insulation and so on, on a spectrum analyzer ). And it came up with a value of 2.3mV. No chance of seeing that properly on an oscilloscope! But nevertheless it causes issues. Such small voltage is enough to mess with a noise floor at -120dB.

Im planning to get some material 78 cores and more 31s to get rid of the rest of it.

I'll also be ordering bigger cores than ft240 to put  thicker cables through it.

73, Lukasz 

[Steve Haynal]

Hi Group,

The power went out in my neighborhood for 6 hours due to a storm. I was able to measure noise floor levels when the power was out. The noise floor was 5 to 10 dB lower when the power was out versus typical power on conditions. The biggest reduction in noise was on 30M. I am surprised there is this much man made noise in my neighborhood.

73,

Steve

kf7o