Hi Quentin,

I have just completed an investigation into this issue as I experience it here also.? I am writing a report which I will share with the groups in a few days.

Basically the problem is caused by your Pi power supply or other similar wall warts in the immediate area of the SHARI antenna. What is happening is your SHARI is coupling RF energy into the power lead to the Pi from the power supply.? This energy at your chose transmit frequency is pulse modulated at a 60 Hz rate by the AC power rectifier diodes which only conduct over part of the 60 Hz cycle resulting in sidebands every 60 Hz (i.e. 60,120,190 ....Hz) with a sin(x/x envelope shape about your transmit frequency.? The radio you are using to listen to your node (call it an HT) receives the RF signal from your node plus this pulse modulated signal.? What you hear depends on the strength of each signal at your HT antenna? and how the FM detector in your HT works (especially the capture effect),? Sometimes the signal from SHARI wins, sometimes the modulated signal wins and most the time it is a combination.? You move your HT a few inches and it sounds fine, move it a few more and the buzz is incredible.

If you have a phone recharging battery, use it to power your Pi and the buzz should go away if there are no other wallwarts nearby.? (other nearby wallwarts can cause the same problem but at a lower level)

I have "debuzzed" my Pi supply by using two FT114-61 ferrites right at the power supply with 4 turns of the power supply lead through each ferrite.? This keeps most of the SHARI RF energy out of the wall wart and thus the sidebands are vastly lower.? ?I have some Fair-rite type 61 cores which should be identical to the FT114-61 cores on order from Digikey (much cheaper than the Amidon cores) and will test them soon.

The problem can also occur on your audio that others hear as the RF from your HT is subject to the same problem.? The closer you get to your node and thus your power supply, the worse the problem.

Another solution to this problem is to use an external antenna on SHARI so that your radiated RF is not near your Pi power supply or any other wallwart.

I created a photo album in the photofolder showing this modulation phenomena.

I find this assessment a real ear-opener. I've always been able to detect who has a Shari node and who has something else, because almost to a fault, those stations using a Shari have the buzz. I don't specifically know any other kind of Allstar node that uses a Raspberry Pi other than a homebrew one, like the one I'm building with a URI cable, a fob, and my own radio, so after reading this, I'm thinking my node will be subject to this problem, too, telling me it's thankfully not due to the Shari transceiver. I'm also thinking mine might have the problem even more than a Shari node might because it's using ten times the power--five watts--but the antenna, a tri-band mag mount, is at twice the distance--outside on a terrace. Definitely good information to have when I get transceiving and discover I have the buzz, too.


Regarding the choke on the Pi power supply, the Pi 3 that runs my hotspot has one, but unles it's built into the in-line switch, the power supply on my 4 B? which will run my Allstar node does not. Curiouser and curiouser.


On 5/1/2020 5:11 PM, N8AR wrote:

Hi Quentin,

I have just completed an investigation into this issue as I experience it here also.? I am writing a report which I will share with the groups in a few days.

Basically the problem is caused by your Pi power supply or other similar wall warts in the immediate area of the SHARI antenna. What is happening is your SHARI is coupling RF energy into the power lead to the Pi from the power supply.? This energy at your chose transmit frequency is pulse modulated at a 60 Hz rate by the AC power rectifier diodes which only conduct over part of the 60 Hz cycle resulting in sidebands every 60 Hz (i.e. 60,120,190 ....Hz) with a sin(x/x envelope shape about your transmit frequency.? The radio you are using to listen to your node (call it an HT) receives the RF signal from your node plus this pulse modulated signal.? What you hear depends on the strength of each signal at your HT antenna? and how the FM detector in your HT works (especially the capture effect), Sometimes the signal from SHARI wins, sometimes the modulated signal wins and most the time it is a combination.? You move your HT a few inches and it sounds fine, move it a few more and the buzz is incredible.

If you have a phone recharging battery, use it to power your Pi and the buzz should go away if there are no other wallwarts nearby.? (other nearby wallwarts can cause the same problem but at a lower level)

I have "debuzzed" my Pi supply by using two FT114-61 ferrites right at the power supply with 4 turns of the power supply lead through each ferrite.? This keeps most of the SHARI RF energy out of the wall wart and thus the sidebands are vastly lower. ?I have some Fair-rite type 61 cores which should be identical to the FT114-61 cores on order from Digikey (much cheaper than the Amidon cores) and will test them soon.

The problem can also occur on your audio that others hear as the RF from your HT is subject to the same problem.? The closer you get to your node and thus your power supply, the worse the problem.

Another solution to this problem is to use an external antenna on SHARI so that your radiated RF is not near your Pi power supply or any other wallwart.

I created a photo album in the photofolder showing this modulation phenomena.

I have performed additional testing to better understand the buzz phenomena.

I fabricated a current probe that measures the current on the power supply cable and cabled it to an HP8901 modulation analyzer.? The power supply cable was intentionally routed next to the SHARI? antenna to produce a worst case situation. Pictures of the setup have been added to the Files folder.? I found that with no ferrite filter on the power supply the AM modulation was almost 20%.? Adding 1 Laird 28B1020-100 core (similar to mix 31) with 2 turns through it reduced the modulation to less than 0.5% (essentially gone).? With the core installed, no buzz could be located for any position of the receiving HT antenna.? I also found that? curling my hand around the antenna and vary its position was a good way to test the suppression as different hand positions revealed buzzing when no ferrite was used.

I also tried a Laird split core, P/N 0431164181, with similar results

I tested both VHF and UHF and found the buzzing phenomena was worse at VHF than UHF.

I also verified that the same modulation situation occurred when I transmitted with the HT (rather than listening) which is why someone might tell you they hear buzzing on your audio.? With the ferrite, the modulation also disappeared.

The bottom line is the buzzing on both transmit and receive is being caused by the Pi power supply modulating the RF antenna current coupled onto the Pi supply power cable.? It can be eliminated with one ferrite core at the Pi power supply.? This buzzing is not caused by SHARI.? Any node with a built-in radio or the HT you are using to talk to your node can cause this to occur.

Steve - N8AR

One of the cores I mentioned isavailable from Mouser.
The split core is available from Digikey??

I'm blind so I have to ask... split core as in two halves split down the middle, and you just pass the cable through the hole?

On 5/4/2020 6:33 AM, N8AR wrote:

One of the cores I mentioned isavailable from Mouser.
The split core is available from Digikey??

Darn good question, Chris, blind or not. I know those who have supposedly perfect (20/20) vision who stare at these things, asking "how's a wire supposed to get in there?"

I have a couple of these kinds of choke sitting around. They are cylindrical, with a hole in the center going from end to enc. There's a hinge on one side and a clamp to keep the thing shut and cylindrical on the other. A little finger pressure--OK, sometimes a lot of finger pressure--and the thing opens like a banana opens after you pull the peel down. When closed, it usuall closes around the wire tightly enough to keep it in place.

HTH

On 5/4/2020 12:01 PM, Chris Smart wrote:

I'm blind so I have to ask... split core as in two halves split down the middle, and you just pass the cable through the hole?

On 5/4/2020 6:33 AM, N8AR wrote:

One of the cores I mentioned isavailable from Mouser.
The split core is available from Digikey??

One of the cores I mentioned isavailable from Mouser.
The split core is available from Digikey??

One of the cores I mentioned isavailable from Mouser.
The split core is available from Digikey??

You can find a USB linear power supply on Amazon.

Also larger brick style USB switching power supplies tend to have better filtering than the small cubes.? I had noise on a small cube power supply and changed to a larger brick style power supply and it was perfectly quiet.? Ive used this one on VHF Shari and its 100% quiet.

If you want to use a cube power supply and try to quiet it down, you can also order ferrite rings from Amazon.? I put two of these on the USB cable closest to the cube power supply and it cut out 95% of the power supply noise.

Hope this helps.

KN6GUW - David

On Wednesday, May 20, 2020, 07:23:02 AM PDT, Steve Matzura <number6@...> wrote:

Are either of these cores better than the other? Also, the split ones, I presume they hinge open, you place the wire inside the split cylinder, then close it? If the fit is not super precisely tight, does it matter if the core moves along the cable? Also, does it matter where along the cable a core is placed?

Also, I just looked at the one from Digikey, and the manufcturer'manufacturer's lead time is sixteen weeks. FOUR MONTHS? Really? LOL!

On 5/4/2020 6:33 AM, N8AR wrote:

One of the cores I mentioned isavailable from Mouser.
The split core is available from Digikey??