Purchased a V3 from a ham on QTH. Told it was 2 weeks old and it’s of course built and tweaked at the factory.

S/N 0994 so someone could ascertain it’s true build date.

It has several issues but the most glaring one is internal receiver noise.

Read all the forums, blogs, etc.

5v Buck boost module has a systemic problem. As some users units exhibit the exact issue I’m observing. I’m guessing some V3’s work or there would be more of a footprint here on this issue.

Tried adding of the suggested caps on the enable line also on the input and outputs of the 5v bb module. Slight if any improvements. Still unusable on 40 meters. Its absolutely worst there.

Purchased the suggested Drok module. Still problematic. Used RC low pass on the output. Some improvement but come on.

Im spending too much time on the hardware as I purchased this to help my Linux bone.

No issues updating the V3 or Pi configuration. Some struggle with the time zones correctly being displayed. No problem as having Linux experience helped.

Is there anyone who has a handle on this internal noise problem??

Dont need “what worked for me” comments.
Whats the solution?

Should be relatively easy for the folks who designed this to weigh in and implement a hard fix.?

Not complaining just need pertinent help.

Many other issues with this radio that more than likely can be fixed via the code.

AGC after unkey hangs to long except when AGC is off. Easy fix.

No S meter. Should be able to implement that via code.?

Of course the screen resizing for WJST etc.
Just to mention the obvious.

Anyway if your going to buy one of these and aren’t ready to deal with these problems?
stay away until it’s more mature. Both hardware and software.

Im just trying to get it function hardware wise before starting any major code experiments.

Gulya, you’re one of the heavy hitters here so your response is interesting. I appreciate the attempt at constructive criticism.
Your inference seems to make this about me.

I much prefer to stay on the topic.

Ive been in contact with HFSignals. Not sure of your affiliation but you may get the same info.

Asked to submit data on the internal interference observed. Video’s and details sent.

Frankly these SBITX are up to a V3 and in my opinion there’s still basic hardware integration issues.

Analog RF chain next to an unshielded Pi4 and display. What did you think was going to happen?
I purchased my V3 for stand alone ops and possible Linux refresher. Unfortunately I’m having to spend time on troubleshooting and conversations such as this. Not what I intended, but here we are.

Dont need “what worked for me” comments.
Whats the solution?

I’ve got the 5 volt issue reasonably working. The Drok module combined in series with a LP filter has significantly stabilized the critical 5v line. Had to play with grounding but ended up grounding at the header and then on the chassis ground.
2.5-3 amps is just what’s needed.

Ive done some experiments with shielding the Pi4 and the ribbon cables.
Using copper foil tape or similar.
I used a bare HAT board wrapped in copper tape.
This is installed on the exposed bottom on the installed Pi4. Then carefully cut long strips of copper tape and cover the long internal ribbon cable. Use clear packing tape over the copper to insulate.
Still working on the main display RFI shield.
With the already suggested 5v DC-DC modifications and the promising results from my amateur copper tape shielding the internal RFI is being attenuated. So headed in the right direction.

?

Reached a point of diminishing returns on this internal noise issue.
Here’s what is effective:

5v DC-DC converter (Drok) upgraded.
Single board LPF on converter output.
Proper grounding on converter.
RF shielding on 22 pin DPI ribbon
RFI cover over display module
HAT RFI plate on Pi4
These enhancements have attenuated the internally generated RFI.

Especially shielding the DPI cable.

Looked at the receiver chain.

Improper IF level can cause what I’ll call pulse oscillation on the higher bands 20 meters and above. On this unit going above 45-60 just causes increased internal noise. S/N suffers greatly especially in LSB.
On this unit it would pulse when IF level setting was between 30-40. Strange thing it stops when you increase or decrease the level.
Checking the receiver sensitivity. I compromised to maintain decent sensitivity and minimal to almost no internal generated RFI. Forget weak signal work but it will catch 90%. Doesn’t appear to have a significant effect on FT8.
RFI mods are working to a point.
When I received this unit 40 meters had an oscillating noise floor and was unusable. Now FT8 on 40 is no problem.
On this unit the IF level ended up being 15-20.
With 15 being were I left it. Yours will be different.
So the IF gain setting is (of course)sensitive and critical for optimum signal with tolerable receiver noise.
This coupled with the internal shielding
has improved the overall operation of the SBITX V3.

Thanks everyone for your assistance and tolerance.

Reached a point of diminishing returns on this internal noise issue.
Here’s what is effective:

5v DC-DC converter (Drok) upgraded.
Single board LPF on converter output.
Proper grounding on converter.
RF shielding on 22 pin DPI ribbon
RFI cover over display module
HAT RFI plate on Pi4
These enhancements have attenuated the internally generated RFI.

Especially shielding the DPI cable.

The DPI/DSI 22 pin ribbon cable is shielded but not grounded to either at the Pi4 or the display board.
Two separate modules. Where to ground properly?

Please send me a picture or video of your V2, V3. Antenna not attached. IF to 70.
Put your SBITX in LSB on any band. Now take a video or a picture of your spectrum/waterfall display.

I’ve attached a picture of my V3 and the obvious noise. This is after I did the 5v BB and internal shielding. It was much worse.

Here’s another picture,
Internal noise on 80 meters. Same on all bands. Appears on LSB. Very audible.

I found using my Oscope noise on the power cable going to the display board.

Appears DSI/DPI ribbon cable coupling RFI onto unshielded power cable.

I could actually move the front panel away from the board and the RFI was almost gone.
I put a 2x 3.5mm ferrite bead on the power cable. RFI noise has been attenuated greatly. It’s no longer audible.
Now just some birdies. I’ll deal with them.
So successful troubleshooting.

Took awhile to track down.
Fairly simple fix.

I found using my Oscope noise on the power cable going to the display board.

Appears DSI/DPI ribbon cable coupling RFI onto unshielded power cable.

I could actually move the front panel away from the board and the RFI was almost gone.
I put a 2x 3.5mm ferrite bead on the power cable. RFI noise has been attenuated greatly. It’s no longer audible.
Now just some birdies. I’ll deal with them.
So successful troubleshooting.

Took awhile to track down.
Fairly simple fix.

After 2x 3.5mm ferrite beads on Pi4 display power leads. Shielded DPI/DSI cable too. Just a couple of birdies now.
Excellent results.

It was type 43 material for the ferrites.
Being a ham I have a lot of different sized ferrite cores, beads etc.
The shielding of the 22 pin 150mm DPI/DSI (display) ribbon did help.

The real issue was the input DC power cable to the display panel.
It was already twisted but not enough for effective common mode reject.
The DPI/DSI ribbon even being shielded was still radiating into/onto the DC power cable.
Twisting the DC power cable further helped.
I installed another LP filter on the DC input to the display board plus two 43 ferrites. This definitely eliminates the DPI/DSI ribbon cable coupling noise.

Still when reassembling the front panel try and keep the DPI/DSI cable away from the display DC power cable. I know they are very close but do your best.

Now able to run the IF level up to high levels(over 70) without internal RFI wiping out the receiver. Running about 55-60 now on IF level verses 15-20 before.
Big improvements in overall S/N.

Also installed ferrite on the main DC input line to the board.

Ferrite on 5v Dc-Dc converter output.

Still have a couple of birdies in LSB on all bands. It’s in the audio passband which makes it sound like a carrier.
I’ll take a closer look at these as it appears to be coming from the Pi4 which I attempted to shield with a cover on the exposed area. More analysis needed.

I hope anyone with a V3 who is experiencing these issues will use this data to make your SBITX perform like it was intended.

Hopefully HFsignals will take a good look at these effective solutions. Possibly incorporate some or all. Cost is minimal.
$20 in parts, 2 hours labor if your old like me. 63 yo…

These modifications are documented. Before and after videos and pictures.

If and when I make progress on attenuating the leftover birdies I’ll inform the group.

开云体育

maybe some "switching" noise from the display??

my vna v2 had some noise on the power rail caused by the display ... some added caps and it was solved

dg9bfc sigi