Re: LPF problem on sbitx V2 (V3) with very low power on 17M!
1. ?Have you looked at the waveform going into the low pass filter to be sure that it actually looks like the right frequency? Is there a possibility that you have more second harmonic then fundamental? Looking at the wave form would answer that question. ? ?Also measure the fundamental. To be sure that it’s the right frequency
2. ?I’m assuming you don’t have a spectrum analyzer or something that you can use to analyze the transfer function of the filter?
3. ?By lifting a component or two you can isolate any of the filters. Put a 50 ohm resistor at the output and connect up an antenna analyzer on the input. ? It should show a good SWR all the way up to some thing near its 3DB point. ? That’s one way to map the filter response
I know that just doing any simple test can take an entire day, but you could help figure out some of these filters that way. ?When I mapped them with my spectrum analyzer I was very impressed with them
4. ? You can actually build or buy Low pass ?filters as well and just substitute right at the output of the transformer as a comparison
Maybe tell us what you’ve got for test equipment?
Gordon KX4Z?
toggle quoted message
Show quoted text
On Jan 1, 2024, at 18:35, John Terrell, N6LN <N6LN@...> wrote:
?I checked voltages on 17 meters along different points of the LPF chains first through LPF_A and then through LPF_B (the one that seems malfunctioning). CW, Drive=100
?
LPF_A??????? D11???????????? C302????????????????????????????? C303??????????????? D16????????????????????? total attenuation from D11-D16, down 48%.
?????????????????? 118V???? 80V (down 32%???????????????? 87V???????????????? 61V?????????????????????????????????????? 9.3 watts.
????????????????????????????? from previous stage)???????? (up 9%)?????? (down 30%)???????????????????????????? Seems to work.
LPF_B??????? D12???????????? C402????????????????????????????? C403???????????????? D17????????????????????? total attenuation from D12-D17, down 84%!
?????????????????? 104V???? 58V (down 44%???????????????? 31V???????????????? 17V??????????????????????????????????????? 0.72 watts.? Not good!
????????????????????????????? from previous stage)?????? (down 47%)????? (down 45%)??????????????????????
The capacitor bank, C401-404 reads a total of 1321 pF for LPF_B, whic adds up to 96% of the expected total so I don't think I need to desolder any of them and test individually.
All four R212-215 read 470K ohms.
I've tried reheating the solder joints of L41-L43 in case there was a cold solder joint, from the top. Perhaps I should remove the PCB and do the same from below.I can't see any toroid wire damage or shorts, although they are quite close together.
All 1N4007 diodes show a forward bias voltage drop of about 0.7 volts.
Perhaps there is a malfunction of one of the 1N4007 diodes in LPF_B under load, or else in Q27, the MOSFET FQN1N50C.
I may try switching those out unless someone has another suggestion. I have a couple of extra 1N4007s. I even got? number of FQN1N50Cs from DigiKey last year because I thought I might have to replace it in the sBitx DE solid state T/R board. Maybe that was a serendipitous order.
Does anyone have any other suggestions based on the voltage data I reported here?
Thanks
Jack
N6LN
|
Re: Sound Volume & TX Power Issues
Steve --
OK, so quick easy workaround until there is a fix.?? Plugged in a set of headphones.?
Problem temporarily solved, everyone in the house is much happier.
73 Roy
On Monday, January 1st, 2024 at 17:33, Steve Beckman via groups.io <n3sb@...> wrote:
Hi Roy;
Right now the way to turn down the FT8 monitor volume during transmit is to change the mode to CW and turn down the sidetone level, then go back to FT8.
73 and HNY; Steve, N3SB
|
Re: Sbitx with Raspberry pi 5?
2021, 2022: ?enormous effort was required to even FIND a raspberry pi. You could expect to spend well over $100, you might have to pay 170 and buy a complete starter set up to get one…
For example:
When I blew two raspberry pies in one of my radios, (trying workarounds around the horrible spurs) I was simply out of luck for almost a year because I just couldn’t find any at a reasonable price to buy
Our group has used raspberry pies for years in deployed systems, but it became cheaper to buy an entire laptop computer, the evolve three, which became available at $60-$80 in quantity! ?
Contrast that to the situation today:
$59-$65 !! ?Half the current price of an evolve three!!
So times have changed dramatically. ?
The biggest problem is that I have with the current sbitx and with the DE:
It took me months to get the DE to be within FCC guidelines. ?I changed circuitry and added changes to the software to reduce the spurs?
I bought a used V2 and between me and fate, blew 2 diode’s that weren’t well protected…Re resulting in signals easily getting around the low pass filter’s…. and again it took months to figure it out. ? I added protection to those diode’s.?
Just while I was packing the unit into a go box, smoke came out! ? Turns out a not well protected MOSFET got blown and that caused other systems to get blown. ?
The current CW reading algorithm is FLDGI which is one of the worst that I know of. ? ?I’m hoping that gets improved.
Thankfully the user community is helping usher out a lot, it’s almost impossible for one person to compete with China!!?
. ?The underrun buffer problem is being figured out just as I am finding places where the circuitry could be better protected
Our county group has radios in the field that have been serving for over five years (just with battery changes every few years ) and have finally been brought down by insects shorting out systems! ?TThat is reliability! ? ?Several in our group have purchased xiegu products. ?I’m almost the only electrical engineer in the group and when they see all the problems that I’m dealing with, well, Xiegus look pretty good….
Gordon
toggle quoted message
Show quoted text
On Jan 1, 2024, at 10:23, Dave, N1AI <n1ai@...> wrote:
?On Sun, Dec 31, 2023 at 02:57 PM, Gordon Gibby wrote:
I look at the Raspberries a little differently.? ?As the software improves, the need for computing resources for the basic performance may DECREASE.? ?
I'll offer a little push-back just to illustrate some other points of view.?? If we call this the optimistic point of view, what works in its favor is that the sbitx software is relatively new and relatively un-optimized and certainly can and is getting better with time.? If it gets optimized and its workload and set of use cases don't expand too much, then I could see an overall decrease in resources.? What works in favor of this direction is the project's designer wanting to use the 512 MB Pi Zero 2W with 512MB so there will be interest in constraining/managing resource growth. What works in the favor of a pessimistic view is that over decades of computing rarely do workloads and use cases fail to expand, and once a resource gets priced low enough code gets optimized in favor of using more of that resource.? ?From what I've observed there was a maniacal obsession on conserving memory in the 60s through the 80s, perhaps easing in the 90s, definitely more relaxed in the 00s and beyond.? ? My discussion of Pi 5 and future Pis is just to get people thinking about what the future trends could be in the hopes the sbitx and its follow-ons can be as useful as possible in the years to come.?
So I see the lower prices on R Pi 4's as a chance to get a spare or two.? ?
I'm not seeing those lower prices on Pi 4s yet.? As mentioned earlier, there is only a five dollar difference between Pi 4 and Pi 5 at the same memory size.? We also see the prices for new Pi 3s haven't fallen to dirt-cheap levels.? I think that's the way it will go.? The Pi team will want to gradually favor production of the new stuff and to do that they can't undercut the new by selling the old cheaply.? ?Another thing to consider is there are other embedded systems out there so Pi does have competition and that means they will keep releasing new things year after year. I'm sure there will be / are Pi 4s on the used market at good prices (I'm counting on it!) and that's fine for end-users but perhaps not so relevant for the vendor of sbitx.
If that will run the sBitx, then that radio is now functional and away I go.? ?At the moment I have one in an almost finished homebrew "go-box".? ?I am less concerned about cpu performance and more concerned about mounting the auto-antenna tuner, adding in a ferrite FT240-43 1:1 choke balun and I built an analog meter($8) based system that lets me watch either DC current draw or RF voltage output (proxy for power in a 50 ohm load).? ??
We should also consider that some people's enjoyment might come from replacing the current CW algorithm with one that uses AI techniques and would appreciate having a more capable CPU, or adding some new form of digital voice, or doing deeper FT8 decoding algorithms, or getting extremely good responsiveness when doing traditional ham radio things.?
Everyone is different, and every one of you play an important role!? ?So I cheer on those who will get us to RPi 5's and 6'ss and beyond, and those who will make the software run on the equivalent of two 2n3904's and a candle!? ?More power to you all!
Make the dadgum thing RELIABLE!? I see lots of improvement in that direction.? ?My team members still see the reliability of G90's as a huge factor.? ?
I agree!
Glad to hear you are active in Field Day, SHARES, EMCOMM, etc and have a great team around you.
Thanks for having a positive attitude to all of the above.
I wish you and everyone reading a great 2024!
?
--
Regards,
Dave, N1AI
|
Re: Sound Volume & TX Power Issues
I've read that sidetone suggestion.
Unfortunately I have CW sidetone at 5, sbitx volume at 40, and set desktop volume to 16.
What happens (additional info) is as soon as the FT8 TX starts, the desktop volume jumps to 76.? As soon as the TX ends, the desktop volume returns to 16.
Thx 73 Roy
On Monday, January 1st, 2024 at 17:33, Steve Beckman via groups.io <n3sb@...> wrote:
Hi Roy;
Right now the way to turn down the FT8 monitor volume during transmit is to change the mode to CW and turn down the sidetone level, then go back to FT8.
73 and HNY; Steve, N3SB
|
Re: LPF problem on sbitx V2 (V3) with very low power on 17M!
Jack - is anything in that LPF getting hot when you transmit?? If there's a short, that RF power may heating the shorted device up.
73 and HNY; Steve, N3SB
|
Re: Automatic IF Gain Control
#sBitx
FYI - there are two versions of sbitx.c - the first one includes Gordon KX4Z's mod to eliminate LPF relay chattering during CW transmit on the DE. The second version is more appropriate for those with V2 or V3 hardware.
73; Steve, N3SB
|
Re: LPF problem on sbitx V2 (V3) with very low power on 17M!
I checked voltages on 17 meters along different points of the LPF chains first through LPF_A and then through LPF_B (the one that seems malfunctioning). CW, Drive=100
?
LPF_A??????? D11???????????? C302????????????????????????????? C303??????????????? D16????????????????????? total attenuation from D11-D16, down 48%.
?????????????????? 118V???? 80V (down 32%???????????????? 87V???????????????? 61V?????????????????????????????????????? 9.3 watts.
????????????????????????????? from previous stage)???????? (up 9%)?????? (down 30%)???????????????????????????? Seems to work.
LPF_B??????? D12???????????? C402????????????????????????????? C403???????????????? D17????????????????????? total attenuation from D12-D17, down 84%!
?????????????????? 104V???? 58V (down 44%???????????????? 31V???????????????? 17V??????????????????????????????????????? 0.72 watts.? Not good!
????????????????????????????? from previous stage)?????? (down 47%)????? (down 45%)??????????????????????
The capacitor bank, C401-404 reads a total of 1321 pF for LPF_B, whic adds up to 96% of the expected total so I don't think I need to desolder any of them and test individually.
All four R212-215 read 470K ohms.
I've tried reheating the solder joints of L41-L43 in case there was a cold solder joint, from the top. Perhaps I should remove the PCB and do the same from below.I can't see any toroid wire damage or shorts, although they are quite close together.
All 1N4007 diodes show a forward bias voltage drop of about 0.7 volts.
Perhaps there is a malfunction of one of the 1N4007 diodes in LPF_B under load, or else in Q27, the MOSFET FQN1N50C.
I may try switching those out unless someone has another suggestion. I have a couple of extra 1N4007s. I even got? number of FQN1N50Cs from DigiKey last year because I thought I might have to replace it in the sBitx DE solid state T/R board. Maybe that was a serendipitous order.
Does anyone have any other suggestions based on the voltage data I reported here?
Thanks
Jack
N6LN
|
Re: Sound Volume & TX Power Issues
Hi Roy;
Right now the way to turn down the FT8 monitor volume during transmit is to change the mode to CW and turn down the sidetone level, then go back to FT8.
73 and HNY; Steve, N3SB
|
Sound Volume & TX Power Issues
I have been following the audio thread over the last days, so hopefully someone will have a suggestion for my problems.
When using FT8, the transmit volume is reset to 76 (as seen on the desktop volume slider) every time sbitx transmits.? You can lower the volume to (say) 16, but as soon as sbitx goes into the next transmit phase, that volume is right back up to 76.
As an added "feature", if you lower the transmit volume DURING the actual transmit, moving the desktop slider to lower the volume ALSO lowers the transmit power, all the way down to zero.? I have a watt / swr meter and confirmed what the radio is showing as the power drops.
I am desperate for a solution, as I have finally constructed an antenna at the house, allowing me to move out of my temporary / undesirable shack.? Unfortunately, the transmit screech is driving both the dogs and the XYL crazy... and I may not be "permitted" to operate this radio in my home office / shack.
Roy
|
Automatic IF Gain Control
#sBitx
I just submitted a pull request to the afarhan/sbitx github to implement an automatic IF Gain control. The changes are all in sbitx.c.
This change will hopefully make it no longer necessary to change the IF Gain setting when hopping from band to band. It seems to work well with W9JES JJ's sBitx Telnet Manager program.
Note that this PR also pulls in -some- of the changes for fixing high CPU usage (the whole Wisdom file discussion) because that change also touched sbitx.c
If anyone tries it, please let me know how it works for you.
73; Steve, N3SB
|
Re: sbitx application startup issues - preliminary solution found
#sBitx
Gyula - I sent you a zip file a couple hours ago. Just checking to see if you got it.??? N3SB
|
Dave - very interesting! Glad you are poking more in this area. At some point we may want to send sound to some other device, which will probably require reclocking. This could be why the V2 web interface sound would get out of sync with the receiver (just guessing, I have not looked closely at that code.)
73 and HNY;? Steve, N3SB
|
Re: Funny things with the Diode LPF Switching
Gordon,
The added Q is on the back of the board for the DE units.
--
Allison
------------------
Please use the forum, offline and private will go to bit bucket.
|
On Wed, Dec 27, 2023 at 07:53 PM, Steve Beckman wrote:
I found that the PCM Playback audio channel was frequently reporting buffer under-runs. To understand what was going on, I wrote a small test program to experiment with ALSA audio. It ran perfectly on a desktop Linux box, but would generate buffer under-runs on the Raspberry Pi. The solution was to run all the audio queues in a blocking mode. The queues actually don't block until the queues are totally filled up, which won't happen in the sBitx since the PCM Capture channel paces the entire audio system.
On Thu, Dec 28, 2023 at 02:02 PM, Steve Beckman wrote:
What really clued me in to what was going on was the test program I created to try some simple ALSA experiments. The program is set up to play sound for 10 seconds. It ran perfectly on a Linux PC. On the Raspberry Pi 4 in my sBitx, however, the program played 10 seconds of sound in 7 seconds!!? Somehow the audio samples in the output queue were either getting lost or getting overwritten. I was able to fix that problem by changing the buffer write mode from non-blocking to blocking. I then applied that change to the two output buffers being used by the sbitx.
I uploaded that test program to the bitx20 groups.io files section a short while ago if you want to play with it.
I am taking you up on that offer.?
I used Audacity to convert some known audio (ok, an Allman Brothers song!) to the right format for the program to play.? It worked just like your experiments, nonblocking mode was faster.?
Then I wanted to understand the difference between "hw:0,0" and "plughw:0,0".? ?This is because the sbitx install.txt talks about using hw:0,0 and not about using plughw:0,0.? I know they are similar but not the same thing.
The program worked with device="hw:0,0" and mode=0 but not mode=SND_PCM_NONBLOCK.?
Googling told me that if you use "plughw" you are using the "plug" plugin which does a bunch of things, one of which is rebuffering i.e. matching the rate of the sender (us) and the receiver (the codec).? When you use "hw" you are writing directly to the sound device.
If you use blocking mode you don't need this, the sender is blocked so it runs no faster than the rate of the codec, but as you suggest if the sender stalls for too long we get underruns i.e. we didn't deliver data to the codec on time.
If you use non-blocking mode and no plug plugin, I see 'short write' messages and I hear distorted sound.? I would expect this to mean we overrun i.e. we're writing faster than the codec can read and there's nothing to rebuffer so the audio gets distorted.? ?Yet I don't hear crackling, what I hear is the audio playing real slooow.? I don't have a good explanation for what is going on.
If we use non-blocking with the plug plugin the plugin will save the data (re-buffering if needed) till the right time to play it back.? ?As you say, this works because it's the codec that paces the system.? It's kind of a "dump and run" strategy.? The user mode code delivers samples as fast as it can and the plugin is playing them out at the right rate.? It seems the writer is never getting too far ahead of the codec, its rate is determined by the rate of the audio it is creating/playing.? It seems the writer does occasionally get behind if we use blocking mode.? It's better to let the writer deliver the data, for the kernel to rebuffer it as needed, and release the writer to get ready for the next block of samples instead of blocking it till the block plays.
Below is what the early part of the C code now looks like on my system.? I think we might consider making room for it in Git either in the sbitx repo or in an adjacent one.??
I've also written up how to use Audacity to make the audio the program wants:
More later...
--
Regards,
Dave, N1AI
?
--
Regards,
Dave, N1AI
|
Re: Funny things with the Diode LPF Switching
the easy fix for the 2n7000 is add a small (1/8th watt) resistor in series with
the Drain.? 100 ohms will certainly be enough.
--
Allison
------------------
Please use the forum, offline and private will go to bit bucket.
|
Re: Funny things with the Diode LPF Switching
GOOD NEWS!!? ?The sBitx Lives Again! Allison -- It was 250mA combined; I had previously set it at 360mA combined trying to improve the horrible spurs, which turned out to be due to shorted diodes in part of the complex circuitry that determines which LPF is selected. Although I could not find my trove of SMD 2N7000/2, I found the leaded ones and on my 2nd try I got one soldered in.? ?That was not easy for me!? ?I put a 200 ohm resistor in series with the DRAIN lead taking the suggestion by Dr. Schmidt, K9HZ -- excellent suggestion!? ? When the old (destroyed apparently) Q101 was pulled out, the probably basically disappeared.? ?When the new Q101 went in I can't tell for sure there is any difference, but it works, and that is what counts.? ?At 12.5 or 12.8 volts, NO OSCILLATIONS when the unit is powered up, or when sBitx is started.? ?And I also tested with NO LOAD.? ?Whew!? ?Passed everything. I have now managed to reconnect the "new" switching power supply and its associated inline current meter and wow! the startup current is indeed WAY LESS, so I was correct in my original concern -- that oscillation was chewing up a lot of power!!? ?I would guess now that it was 3Amps or so (but I don't have this well calibtrated). Here is a photo of the destroyed resistors that the oscillation caused:? (Look just to the left of the orange transformer secondary wiring)  Here is a photo of my amateur attempt at soldering in a new MOSFET to replace the destroyed Q101:  Yes, the iron is probably way too hot.? ?You can see the 200 ohm SMD resistor tombstoned on the drain solder point, and I have bent around tthe drain from the leaded MOSFET to match.? ?For me, this is ticklish work. I would STRONGLY suggest that the design get modified to put a current limiting resistor (or other solution) to protect Q101 ** Otherwise, I would suspect there are going to be a fair number of sBitx's that reach the secondary market because of "smoke released" and the owner is clueless why.? ? This has apparently happened to me TWICE now and most of my work on this particular unit has just been hours and hours of testing with a load attached, which significantly reduces the amplitude of the oscillation in my case.? ?Units that are turned on with NO LOAD may see far higher oscillatory power once Q101 fails, IF it fails in such a manner that it turns on the IRF4905.? ?(Some unknown percentage, right?)? ? ?This appears to be a preventable failure. Pretty sure my "Developers Unit" has a leaded Q101 (Qmumble) installed, so it will be considerably easier for me to insert a resistor in the drain lead there.? ?? Thanks for all the insights from Evan and Allison and Dr. Schmidt -- this took only about 24 hours to solve and get the unit working again thanks to your input! 73 Gordon KX4Z
|
Re: Funny things with the Diode LPF Switching
That was a first pass solution, later I went ta a 4N26( I had a handful)? optoisolator
for all logic level IN and OUT pins of the Rpi as I was not comfortable with people
reporting frying IO pins and the Pi itself.
The big issue with TX switching was two fold, at the time the RF switching and
that relay.? The other was switching off allows for 12V to collapse based on RC
time constant.? ?That resulted in the diode based RF switching with or with out
relays for selecting low pass filters.?
The issue with both on is solved mostly with the 2n7000 RC delay but depending
on gate thresholds and device switching times it may have a narrow (less than 1 uS)
spike when both are conducting (partially) at same time. It does depend on the
turn off time for the Pmos being generally less than 100nS.? Te as built on board
solution was without the RC delay and risky due to thresholds and pull up time.
The desired case was either 12V (or whatever was applied) and ground
without ambiguity, float or maybe.?
At one point that whole mess became a an Omron GS6 spdt to switch
the load side between 12V and ground and isolate the Rpi TX signal.
Didn't care the relay was noisy as it was a hard solution that allowed
for moving on to allow looking at other issues, which were plenty.
I? took the make it bulletproof, then clean up later approach.? I tossed
in the towel back in 2022 as it was clear issues were too plentiful and
life had other things I needed to deal with.
--
Allison
------------------
Please use the forum, offline and private will go to bit bucket.
|
Re: #sbitx VNC access to sBitx + wsjt-x works
#sBitx
OK, so this is my course of action. Thanks to all who contributed
to the thread also - have read and saved it all. Now to get some
fresh cards and burn my first card image.
John Seboldt K0JD,
Milwaukee, WI
On 12/27/2023 16:29, Dave, N1AI wrote:
toggle quoted message
Show quoted text
On Wed, Dec 27, 2023 at 04:39 PM, John Seboldt K0JD wrote:
So is the best course of action to recreate these
settings or just get a new SD card image?
My $0.02: yes to doing both.
I have learned the hard way to keep backup copies of the image,
see my earlier post.
I suggest you start with a copy of your backup of your
factory-fresh image, if you have one.
If you don't have one, start with a copy of the newest available
image which is the V2 image:
- as
shipped?with
version 2.1 software?for the
sBitx. Download, unzip and?.? More details
at?this post.?
Make sure your drive control works with that image.
Then, if it does, make a backup of that image.
Then, move on to V3 only once you know the V2 image works as
expected.
- Update the software in-place:?link
Once you know the drive control is working on V3, then if you
did not start with a backup of the original factory image either
live with the "conservative" settings or follow the calibration
procedures at??and
do calibration (presuming you have the equipment to do so).
?
--
Regards,
Dave, N1AI
|
Re: Funny things with the Diode LPF Switching
On Sun, Dec 31, 2023 at 07:46 PM, Gordon Gibby wrote:
360mA instead of the recommended 250mA.?
or either IRFZ44 or IRF510 180ma each is not unnreasoble (assuming good heatsinking).. IF that 360 mA is per device.. that's very high to the point of pain. ? -- Allison ------------------ Please use the forum, offline and private will go to bit bucket.
|
Re: Funny things with the Diode LPF Switching
OK.? New Discovery.
Dawned on me that the destruction of Q101 (aka Qmumble) might extend to its GATE--and that might explain the unexpected current drain through R6.
Managed to remove Q101 without doing TOO much damage.? ?
Immediate improvement.
No more signs of oscillation on startup even at 12.88 VDC
R6 -- both sides of it are now at supply.? ?Suggests that the 10K on the gate of the MOSFET driven by TX was indeed adequate to keep it from being turned ON by tthe 50K internal pullups of the Raspberry.? ?THAT part of the original design worked.? ??
Output of the 4905 -- now 0 volts.? .? ??
So the ability of the 4905 and the Q101 to be ON simultaneously would always seem to result in the destruction of Q101 and potentially to the problems I experienced with oscillation and loss of components connected to the highpower output terminals....? ?THAT part of the design would appear to need improvement.
Now to see if I can get a MOSFET back in there without destroying it, and with a series resistor.? ? Not sure if I'm game for the RC circuit but thatt appears to be another potential fix.
Gordon KX4Z
|
John Terrell, N6LN
Dave, N1AI