|
I have a different experience. V2 with some V3 mods but stock 5 V bb.? I replaced the original DSI cable with a longer one that allows the case halves to lie flat open as it's a PITA to disconnect and reconnect when you split the case. Screen shots with antenna disconnected at IF 70 and 90.
--
73
??? Bob? KD8CGH
|
As seen in these 2 images on the same frequency with matching IF levels.
With my RFI/EMI modifications you’ll notice the noise floor is ~3-4db better at the 90 IF settings. Look at the waterfall you’ll observe much less snow(noise).
The birdies seen are very similar to your units.
Im investigating the source and the possible fix.  
|
wanted to add some comments as i trouble shoot this same issue before i was a part of this IO group.
I had this exact issue and it was VERY bad, huge birdies on the waterfall and very high level of noise at the target frequency.
i noticed the major noise ( for my sbitx V3 at least) was from the flat display ribbon getting close to the radio board. while the radio was on i could move it closer or further away and recreate the issue. if i gently formed it (bent it) away from the radio board the noise dropped of considerably.? i re-organized all of the cables to "encourage" the display ribbon to stay away from the radio board.?
but the but big improvement came when i added aluminum tape to the display ribbon to act as a shield. same tape that is used for shielding on guitars (pretty much the same as the aluminum foil tape for duct work) and covered the aluminum tape with electrical tape so it can not short anything. I was concerned this could cause some capacitance on the ribbon and thus display issues but , no,? no issues to be found.?
Wow what a difference.? i can crank the IF to 100 without getting the noise anymore.??
after reading the comments form this thread i am going to open it back up and add the solutions you have found as well, thank you for sharing all your findings.?
hope this helps someone, W9BLW
|
W9BLW, thanks for sharing your experience. I implemented this idea that I shared in message ? #114315? regarding a program error. The idea came from this topic and it was only a matter of time before I implemented it.
I am satisfied with the result and I recommend it to others.
--
Gyula HA3HZ
|
Good on you for taking this on! ?I know many others will appreciate it also. ?We all just deal with it. ?I have a drok, twisted pair and a ferrite on the display power and a flat ferrite made for ribbon cables on the display cable. ?I have knocked down quite a bit, but the birdies are very annoying more visually than audible for me as you can move them out of the pass band on 4.2 64bit. ?I have a fan case on top of the pi5 that I may add the copper tape to act as a shield and am thinking of trying an aluminum shield between the rf board and display. ?This would cover a good portion of the board. ?
Ryan
kk6dzb
|
HA3HZ,?
nice write up, your shielding looks really nice, much nicer than mine. :)? what was the purpose of shielding under the RPI??
Ryan,?
i was thinking the same thing, need to shield the whole board.? I just reopened mine and i think we can 3D print a "mold" that we could then cover in shielding tape or paint ( the tape works better in my experience).? Looks really tight with the speaker in there but I'm going to take some measurements and see what i can come up with.? ?the output side will / would be much easier to 3d print a mold for.? thinking jut to keep the ribbons away from the filters if nothing else.
w9blw
|
We've discussed this since the V2 days as both of the models suffer from the same issue. There is no shielding whatsoever inside this mass-produced radio which causes all sorts of issues. Mitigating internally generated noise can be accomplished by shielding the DSI cable, replacing the buck converter with a shielded one, making a shield cover for the RF input area, grounding unused pins, etc.
?
?
-JJ
|
On Thu, Jan 2, 2025 at 01:25 AM, w9blw wrote:
HA3HZ,?
nice write up, your shielding looks really nice, much nicer than mine. :)? what was the purpose of shielding under the RPI??
That's exactly why. Because of the lack of shielding.
Since there is shielding between the pcb and the RPi and the monitor cable is also shielded, I can now increase the sensitivity to a much higher level without increasing the noise. Of course, increasing the sensitivity also increases the noise, but not as much as without shielding.
--
Gyula HA3HZ
|
I have begun working to shield mine. ?We will see if it was worth it.
?
Ryan
KK6DZB
|
Thanks for the photos.? ?Documentation of exactly what physical techniques are used will help others.? ?
Gordon?
On Thu, Jan 2, 2025 at 10:51?AM Ryan Wesolowski via <cosmo1stgen= [email protected]> wrote: I have begun working to shield mine.? We will see if it was worth it.
?
Ryan
KK6DZB
|
Ryan, I mocked up some forms to act as a mold for shielding tape yesterday.? The output side had plenty of space to add a full shield and seems like a no brainer. The input side did not have any room to spair so i had to cut the middel part of the form out. The speaker literally touches some of the coils. I was able to add some shielding to the sides but i don't know if its really attributing to anything.? i added the .stl files i used to this post if anyone wanted to try this them selves, as a disclaimer I am self taught with 3d design so I'm sure someone could do much better. you will need to print them on their back and use an auto tree for the over hangs.? They print really nice, i use a bambu P1S. i printed 100% fill and they still print fast (~1 hour) with a .06 hotend . as you can see you will need to cut away at them depending on how you want them to fit but that is very easy to do.? big question is how is the noise. ill put it this way, if the radio performed as it is now i would had never thought there was an issue.? do i still see things when i crank the IF past 90 yes.. but it doesn't look abnormal, no birdies or hard hash in one spot. Ill be the first to admit the radio is a ton of fun with or without this shielding but it has seemed to make a difference and its a fun project.? some photos with notes.?  ?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
|
toggle quoted message
Show quoted text
Ryan, I mocked up some forms to act as a mold for shielding tape yesterday.? The output side had plenty of space to add a full shield and seems like a no brainer. The input side did not have any room to spair so i had to cut the middel part of the form out. The speaker literally touches some of the coils. I was able to add some shielding to the sides but i don't know if its really attributing to anything.? i added the .stl files i used to this post if anyone wanted to try this them selves, as a disclaimer I am self taught with 3d design so I'm sure someone could do much better. you will need to print them on their back and use an auto tree for the over hangs.? They print really nice, i use a bambu P1S. i printed 100% fill and they still print fast (~1 hour) with a .06 hotend . as you can see you will need to cut away at them depending on how you want them to fit but that is very easy to do.? big question is how is the noise. ill put it this way, if the radio performed as it is now i would had never thought there was an issue.? do i still see things when i crank the IF past 90 yes.. but it doesn't look abnormal, no birdies or hard hash in one spot. Ill be the first to admit the radio is a ton of fun with or without this shielding but it has seemed to make a difference and its a fun project.? some photos with notes.?  ?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
|
Wb9blw. ?
Looking good! ?I also had the problem with the speaker. ?I made a low profile shield from .030 thick lexan sheet and covered it with copper tape. ?It is secured with 2 of the board of mounting screws and has a bonding ground to the case next to the pi. The shield continues under the pi. ?Picture attached. ?I will add the aluminum shield on top of this. ?
|
On Sat, Jan 4, 2025 at 12:11 AM, w9blw wrote:
big question is how is the noise. ill put it this way, if the radio performed as it is now i would had never thought there was an issue.? do i still see things when i crank the IF past 90 yes.. but it doesn't look abnormal, no birdies or hard hash in one spot. Ill be the first to admit the radio is a ton of fun with or without this shielding but it has seemed to make a difference and its a fun project.?
Thank you for introducing your work and for these words you wrote here:
--
Gyula HA3HZ
|
|
|
Just incredible work, wow...? few questions for you,?
the silver box with yellow red and black wires at the bottom, what is that, replacement buck converter?? can you send me a link to what you used?
id you 3d print a new end on the right side??
did you change to the DS3231 RTC or did yours come with that?
Your fan, are you managing the speed from CPU temp or just an always on?
did you notice any impact to overall Rx performance??
Again, im really impressed with your work
|
Guys, on the shielding of the display ribbon cable, how many ground connections were used. I wrapped mine and grounded the shield at the display but not at the PI end due to lack of an easy ground point. So far, I have not noticed any change to the amount of junk I hear, especially on 20Meters.
Really interesting thread.
TU
Barry
K3BO
?
|
Hi Barry,
I'm no expert so just sharing my experience.? When i was first playing around investigating i had the radio open and noticed a huge change in hash noise at the target frequency just by moving the ribbon away from the main board.? My suggestion is to make sure, shielded or not, that the ribbon s carefully bent to stay as far from the main board as possible. just doing this make a bigger difference than the shielding from what? saw.? I started gently with the bending but at the end was fairly aggressive to get it to go straight up after leaving the pie. The tape i used also held it more upward away from the board.? the ribbons are cheap and readily available online if you were to adversely damage it.? I also re organized my cables to encourage it to stay away from the board as i closed it back up.?
to your question when i was first testing i did not have the shield grounded, i later grounded it at the PI's hold down screws, it did not make a notable difference. In networking and industrial automation, my professions, we never ground at both sides, not sure this applies here though. maybe someone else can chime in with good engineering practices.? ?hope this helps.?
|
W9blw,
?
Need to get your name on here!? I might have to look it up on QRZ!
?
The silver box is the 3 amp Drok regulator
?
The piece on the side is 3d printed from Jeff's work at DuceRF.? I have a PI5 which does not line up properly with the original end plate and I have the DuceRF QRO board which is mounted to the plate.? Jeff's plate corrects the misalignment of the PI5 connectors and adds mounting bosses for the QRO board.?
?
The RTC came with my SBITX.? Thought the PI5 has its own and just needs the battery I have not implemented that.?
?
The fan is being controlled by the native fan control circuit on the PI5.? I spliced in the original connector to the new fan.? As the temp goes up so does the fan speed.? Full PWM control.
?
The noise floor on the SBITX is significantly lower.? So much so that I thought I had shorted something out when I first powered it on.? It will rival my direct sampling 16bit radio in sensitivity even when the band is busy.? Unfortunately the spurs still exist which is the most annoying thing to me so I need to get a LPF for the Drok regulator and try it.? I may rig one up soon.? I also seem to have pretty bad intermod from strong AM broadcast.? I need to rig up a HPF on the antenna to try.? Or wherever in the RX chain of the radio.?
?
Ryan
KK6DZB
?
?
?
|
Oh yes, my name is Brandon??
On Sat, Jan 4, 2025, 2:42 PM Ryan Wesolowski via <cosmo1stgen= [email protected]> wrote: W9blw,
?
Need to get your name on here!? I might have to look it up on QRZ!
?
The silver box is the 3 amp Drok regulator
?
The piece on the side is 3d printed from Jeff's work at DuceRF.? I have a PI5 which does not line up properly with the original end plate and I have the DuceRF QRO board which is mounted to the plate.? Jeff's plate corrects the misalignment of the PI5 connectors and adds mounting bosses for the QRO board.?
?
The RTC came with my SBITX.? Thought the PI5 has its own and just needs the battery I have not implemented that.?
?
The fan is being controlled by the native fan control circuit on the PI5.? I spliced in the original connector to the new fan.? As the temp goes up so does the fan speed.? Full PWM control.
?
The noise floor on the SBITX is significantly lower.? So much so that I thought I had shorted something out when I first powered it on.? It will rival my direct sampling 16bit radio in sensitivity even when the band is busy.? Unfortunately the spurs still exist which is the most annoying thing to me so I need to get a LPF for the Drok regulator and try it.? I may rig one up soon.? I also seem to have pretty bad intermod from strong AM broadcast.? I need to rig up a HPF on the antenna to try.? Or wherever in the RX chain of the radio.?
?
Ryan
KK6DZB
?
?
?
|
Bob Benedict, KD8CGH
Kb4yho
HA3HZ
