Re: ubitx audio amp problem discovered
toggle quoted message
Show quoted text
Probably, Raj.? Ultimately I'm trying to understand how to implement Farhan's mod.?
Since I don't yet know how to create the emitter follower he described, I'll go back to applying fiberglass to the bottom of my boat.?
Thanks, Andy
On Mon, Jan 22, 2024 at 12:22 AM, Raj vu2zap wrote:
Andy,
You are talking about a hard to get audio transformer!
Raj
On 22/01/2024 8:21 AM, AndyH wrote:
Could this be fixed with a 9T:2T transformer between C63 and the mixer pad, and rebiasing Q6 for a gain of 2-3?
And/or, could someone flesh out the emitter follower and specify which parts from the emitter of Q6 should be removed? (Just C62 and R64?)
? ?Thanks!
? ? ?73, Andy, KG5RKP/MM
On Sat, Jul 15, 2023 at 09:18 PM, Ashhar Farhan wrote:
Thanks for the analysis. There are two other significanr issues with the current mic amplifier.
1. The output impedance is 1000 ohms, the modulator needs a 50 ohms drive.
2. The gain is not less but too much. The electrer mic produces about 50mV of audio, the modulatoe needs less than 150mv drive. A gain of two or three is more than sufficient. The apparent loss is due to the mismatch between the 1000 ohm output and the needed 50 ohms drive.
These can be solved by adding an emitter follower with a 47 ohms resistor in series with the output of the emitter follower and removing the emitter bypass of the original mic amplifier.
- f
I've been working on the uBitx audio circuit for a while now and I think I've discovered a problem that I've never seen anyone mention before. If anyone has experience with this circuit, please review and let me know if I've missed anything. Here are my notes:
?
?Audio Amp notes.
?
The uBitx has always had challenges with the audio amp. It has insufficient gain for the condenser mic that is provided. Solutions include:
?
a. Use a mic with an amplifier built in.
b. Add an additional amplification stage such as the SSM2167.
c. Increase the mic circuit gain.
?
Mic amplifier circuit with suggested modifications. Also showing the load of the following stage.
?
SEE ATTACHED FILE
?
Analysis:
Given that the receive transistor Q70 is connected directly to the output of the microphone amplifier, the output of the mic amp must be less than 0.6v. We¡¯ll say 550 mV to give us a bit of safety margin.
The input drive required to achieve 550mV out of the mic amp is 200mV. Most have claimed that the mic amp has a gain of around 20 due to the collector resistance of 1k divided by the emitter resistance of 47 ohms. This simple analysis is incorrect because it neglects the load impedance of the mixer circuit. Spice analysis and actual observation shows the actual gain to be about 2x or 3x.
A big problem with this mic amplifier is that it is incorrectly biased. With the originally specified bias resistor R61 set at 2.2k, the transistor runs out of emitter current at about 160mV of output voltage. In addition to limiting output, this causes distortion of the audio waveform. By increasing R61 to 10k, the circuit can easily output the 550 mV (1.1V P-P) originally suggested.
In my own uBitx, I have chosen to use the SSM2167 as a preamp in order to enjoy some voice compression. However, the simplest solution is to decrease R63 to 10 to 22 ohms. I¡¯d advise not to go lower than 10 ohms, since that circuit should fully drive the mixer with a mic output of only 40 mV. But before doing anything else, increase R61 to 10K.
|
Gyula,? just confirmed what you said.? I just turned the rig on and the first contact showed "logged" twice but does not appear in the log file. It looks like the second contact worked normally and auto-logged.?? So simply manually logging the first contact seems to work around the issue.
--
Mike WB5LLI
|
Re: ubitx audio amp problem discovered
Probably, Raj.? Ultimately I'm trying to understand how to implement Farhan's mod.?
Since I don't yet know how to create the emitter follower he described, I'll go back to applying fiberglass to the bottom of my boat.?
Thanks, Andy
toggle quoted message
Show quoted text
On Mon, Jan 22, 2024 at 12:22 AM, Raj vu2zap wrote:
Andy,
You are talking about a hard to get audio transformer!
Raj
On 22/01/2024 8:21 AM, AndyH wrote:
Could this be fixed with a 9T:2T transformer between C63 and the mixer pad, and rebiasing Q6 for a gain of 2-3?
And/or, could someone flesh out the emitter follower and specify which parts from the emitter of Q6 should be removed? (Just C62 and R64?)
? ?Thanks!
? ? ?73, Andy, KG5RKP/MM
On Sat, Jul 15, 2023 at 09:18 PM, Ashhar Farhan wrote:
Thanks for the analysis. There are two other significanr issues with the current mic amplifier.
1. The output impedance is 1000 ohms, the modulator needs a 50 ohms drive.
2. The gain is not less but too much. The electrer mic produces about 50mV of audio, the modulatoe needs less than 150mv drive. A gain of two or three is more than sufficient. The apparent loss is due to the mismatch between the 1000 ohm output and the needed 50 ohms drive.
These can be solved by adding an emitter follower with a 47 ohms resistor in series with the output of the emitter follower and removing the emitter bypass of the original mic amplifier.
- f
I've been working on the uBitx audio circuit for a while now and I think I've discovered a problem that I've never seen anyone mention before. If anyone has experience with this circuit, please review and let me know if I've missed anything. Here are my notes:
?
?Audio Amp notes.
?
The uBitx has always had challenges with the audio amp. It has insufficient gain for the condenser mic that is provided. Solutions include:
?
a. Use a mic with an amplifier built in.
b. Add an additional amplification stage such as the SSM2167.
c. Increase the mic circuit gain.
?
Mic amplifier circuit with suggested modifications. Also showing the load of the following stage.
?
SEE ATTACHED FILE
?
Analysis:
Given that the receive transistor Q70 is connected directly to the output of the microphone amplifier, the output of the mic amp must be less than 0.6v. We¡¯ll say 550 mV to give us a bit of safety margin.
The input drive required to achieve 550mV out of the mic amp is 200mV. Most have claimed that the mic amp has a gain of around 20 due to the collector resistance of 1k divided by the emitter resistance of 47 ohms. This simple analysis is incorrect because it neglects the load impedance of the mixer circuit. Spice analysis and actual observation shows the actual gain to be about 2x or 3x.
A big problem with this mic amplifier is that it is incorrectly biased. With the originally specified bias resistor R61 set at 2.2k, the transistor runs out of emitter current at about 160mV of output voltage. In addition to limiting output, this causes distortion of the audio waveform. By increasing R61 to 10k, the circuit can easily output the 550 mV (1.1V P-P) originally suggested.
In my own uBitx, I have chosen to use the SSM2167 as a preamp in order to enjoy some voice compression. However, the simplest solution is to decrease R63 to 10 to 22 ohms. I¡¯d advise not to go lower than 10 ohms, since that circuit should fully drive the mixer with a mic output of only 40 mV. But before doing anything else, increase R61 to 10K.
|
On Mon, Jan 22, 2024 at 02:30 PM, WB6GJE Mike wrote:
FYI, Jeff Geerling covers the use of USB and PCIe booting for Raspberry Pis in great detail on his blog (; for example - ) and on YouTube ().
I definitely love Jeff's work and watch every video as they come out.? The only caution I would raise is that he tends to be working on the bleeding edge with Pi, so some of his older articles don't age too well, because once the stuff he is working with becomes mainstreamed usually easier ways to do things have been invented or discovered.? This is not the case for PCIe right now, it is still "bleeding edge" and his stuff is still the best reference IMO.? For USB I posted the Tom's Hardware link because I think it's easier for most people to follow.? In particular Imager can be used to do some things Jeff did manually. ? -- Regards, Dave, N1AI
|
FYI, Jeff Geerling covers the use of USB and PCIe booting for Raspberry Pis in great detail on his blog (; for example - ) and on YouTube ().
|
toggle quoted message
Show quoted text
On Jan 22, 2024, at 8:56?AM, Jeff Kelly via groups.io <jkelly@...> wrote:
?Mint Sbitx v3 with all that comes from HF Signals.
Used for receive.
$350 plus ship. Conus only.
Jeff
K2SDR
|
Re: An update to the sbitx software - v 3.02
Can you share the contents of your build file?
Otherwise - recommend renaming the sbitx application to sbitx.orig then executing the build script again.
73, Steve, N3SB?
|
My interesting cases today in pictures. In the picture you can see that saving to the LOG file was done twice, while in reality it was once. The program does not allow you to save the same thing again in the log file within two minutes. So it can be seen from this that at first he only wrote to the screen, but did not save. For the second one, I pressed the SAVE button manually, then the saving was done. The interesting thing is that when I turn on the device, it does this the first time. The subsequent qso is saved to the log file.  With the sbitx (rf output 10W) and the vertical loop antenna, this would have been my first South American qso. It failed to confirm during the past 6 minutes, but it was automatically saved to the log file three times. This happened because of the logic of saving to a log file.   -- Gyula HA3HZ
|
test message from WW8LA.tom
toggle quoted message
Show quoted text
On Jan 22, 2024, at 12:03?PM, Dave, N1AI <n1ai@...> wrote:
?On Mon, Jan 22, 2024 at 11:41 AM, test message from WW8LA.tom wrote:
Where is the ¡°BOOT_ORDER¡± parameter within the code.
It is stored in EEPROM.? shows how to access it: <dummyfile.0.part> You will need the right firmware on the Pi 4 for this to be operative.
I would like to try booting from external media but am not that familiar with where to find it.?
?is a user-friendly guide for how to get things set up. ? -- Regards, Dave, N1AI
|
Re: Introducing sBITX Scanner and All-In-One Installer for my apps
#sBitx
#sBITX_v3
Thanks for writing these programs JJ, they very useful!
|
On Mon, Jan 22, 2024 at 11:41 AM, test message from WW8LA.tom wrote:
Where is the ¡°BOOT_ORDER¡± parameter within the code.
It is stored in EEPROM.? shows how to access it:  You will need the right firmware on the Pi 4 for this to be operative.
I would like to try booting from external media but am not that familiar with where to find it.?
?is a user-friendly guide for how to get things set up. ? -- Regards, Dave, N1AI
|
Re: An update to the sbitx software - v 3.02
My case, what is the solution?
|
Re: V3 versus V2 SBITX Board - What are the differences?
Hi Mike,
The problem of the diode LPF switch current comes up when measuring the spurious content for 80 meters.? The variations of Diodes and other filter components will come into play.? I would guess that most will not have the issue.? I do not have a V2 to test the assumption.? As far as I know, the change does not impact any of the other parameters.? If I had a V2 board, I would make the change.
This also points out that since the sBitx is sold as a kit, it is up to the Ham to verify conformance to local signal purity standards.? The purchaser should test the kit to ensure conformance.
73
Evan
AC9TU
|
Re: An update to the sbitx software - v 3.02
Hi - can you share the contents of your build script?? The build script that's been part of the distribution does not call the make program.
When the make program does not find any source files that need to be built, it will report "Nothing to be done for 'all'. " This is only for information - it does not imply that an error has occurred.
73; Steve, N3SB
|
test message from WW8LA.tom
Dave, Where is the ¡°BOOT_ORDER¡± parameter within the code. I would like to try booting from external media but am not that familiar with where to find it.? --Tom WW8LA
toggle quoted message
Show quoted text
On Jan 22, 2024, at 10:59?AM, Dave, N1AI <n1ai@...> wrote:
?On Mon, Jan 22, 2024 at 10:16 AM, Ken N2VIP wrote:
Also, Dave mentioned it, but I want to highlight the need to remove the microSD card from the RPi 4 - I suspect (but do not know for certain) that the RPi 4 boot order will have the RPi 4 boot from the microSD no mater what is attached to the USB ports.
All good points, and more on the topic of booting: There is a boot order variable the Pi firmware supports.? Its documentation is?.? You can google BOOT_ORDER to learn more about it.? I don't know what the resulting boot order will be by following the Tom's Hardware procedure I linked earlier, maybe someone will report it. You can see it via the command 'rpi-eeprom-config'. Some of my non-sbitx Pi devices have a NVMe board that is buried under the Pi and I don't want to open the box and disconnect it so I let the other things have priority in the boot order: <dummyfile.0.part> Note the hex digits are read low-order first so '1' is the first thing it tries not 'f'. For the factory sbitx box the sdcard is buried inside the box by default and it's a pain to get to, so you probably want any external devices to have priority. That makes booting slower since it takes it time to look for USB devices but IMO it's worth the tradeoff of not having to pry open the box. Someone here posted??for a sdcard extension and ran the ribbon cable to the outside of the box and taped down the card holder. I haven't tried that, but I plan to. I have been USB-booting my Pi 5 with a USB hub then a USB-to-NVMe adapter.? I often see timeouts during reboot.? One 'fix' is to power-cycle the Pi and the hub together which is easy for me since they are on the same power strip.? ?The same document above describes some bootloader USB settings that can be used to make the timeouts longer, but I haven't messed with these since I intend to replace that USB-to-NVMe adapter with PCIe direct.? Yet the odd thing to me is that the same USB-to-NVMe adapter and the same NVMe device worked without timeouts on the Pi 4 so I suspect they have changed something on the Pi 5 with regard to USB timeouts.? It'd be interesting to know if others notice something similar. ? -- Regards, Dave, N1AI
|
Mike,
I understand. A few things work interestingly with ft8 autosave.
When I turn on the device, the first connection says twice that it has been logged. Actually once.
If the double logging is not visible on the display, it is worth pressing the manual log entry, because it is not in the log.
It only does this the first time you turn it on.
Before that, I look at the log file, close it, and then leave until the data is in the LOG window on the GTK screen.
It happens that you end the connection earlier and then delete it from the LOG window. You have to pay attention to the operation.
So for the first QSO, I always press a PrtSc i.e. Print Screen button, you can find the picture in the /home/pi folder.
Because if you don't save it in the log file, then I have data for later input.
I'm writing these things because that's what I'm currently doing in order to function well.
There are more things, but so far no one has responded to my entries. Everything remains the same.
--
Gyula HA3HZ
|
On Mon, Jan 22, 2024 at 10:16 AM, Ken N2VIP wrote:
Also, Dave mentioned it, but I want to highlight the need to remove the microSD card from the RPi 4 - I suspect (but do not know for certain) that the RPi 4 boot order will have the RPi 4 boot from the microSD no mater what is attached to the USB ports.
All good points, and more on the topic of booting: There is a boot order variable the Pi firmware supports.? Its documentation is?.? You can google BOOT_ORDER to learn more about it.? I don't know what the resulting boot order will be by following the Tom's Hardware procedure I linked earlier, maybe someone will report it. You can see it via the command 'rpi-eeprom-config'. Some of my non-sbitx Pi devices have a NVMe board that is buried under the Pi and I don't want to open the box and disconnect it so I let the other things have priority in the boot order:  Note the hex digits are read low-order first so '1' is the first thing it tries not 'f'. For the factory sbitx box the sdcard is buried inside the box by default and it's a pain to get to, so you probably want any external devices to have priority. That makes booting slower since it takes it time to look for USB devices but IMO it's worth the tradeoff of not having to pry open the box. Someone here posted??for a sdcard extension and ran the ribbon cable to the outside of the box and taped down the card holder. I haven't tried that, but I plan to. I have been USB-booting my Pi 5 with a USB hub then a USB-to-NVMe adapter.? I often see timeouts during reboot.? One 'fix' is to power-cycle the Pi and the hub together which is easy for me since they are on the same power strip.? ?The same document above describes some bootloader USB settings that can be used to make the timeouts longer, but I haven't messed with these since I intend to replace that USB-to-NVMe adapter with PCIe direct.? Yet the odd thing to me is that the same USB-to-NVMe adapter and the same NVMe device worked without timeouts on the Pi 4 so I suspect they have changed something on the Pi 5 with regard to USB timeouts.? It'd be interesting to know if others notice something similar. ? -- Regards, Dave, N1AI
|
Re: V3 versus V2 SBITX Board - What are the differences?
Re-reading over the "sbitxv3 PA changes and measurements" doc* I see this item, but I don't recall reading of anyone making the changes (nor the other 2 changes preceding this).
3. The standing current in the diodes that switch the LPFs was found insufficient on 3.5 MHz, hence, the resistors R201, R202, R203 and R204 are changed from 470 ohms to 330 ohms.
Just how "insufficient" was the current, and is it recommended? I have V2 hardware.
Thanks,
Mike
*
|
David Saylors WK4DS recently offered to make available for sale working SDs with versions 2 and 3.01 software on them.? They are copies of his operating system.? He did mention that some of the macros had been changed, as he apparently does quite a bit of POTA & some of his macros reflect this.? (I will be ordering some.)? For those who have seen FLDIGI, you know that it is possible to set up 4 banks of messages and select any bank of 12 messages at any time.? Is there any interest in something like this for the standard sBIT software?? Alternatively, make the macros a separate file, and have separate files with the ability to select the desired macro set from multiple files at run time. I haven't actually used my sBIT yet, and don't know if I would use this feature, but perhaps it's something to think about.
Gerry Sherman
Sent by the Windows 11 Thunderbird
|
Hi Gyula,? I did a reinstall after upgrading my v2 folder to v3.02.? I had v3 operating fine in a separate folder.? The problem was with FT-8 and some repeat transmits the macros as well as showing "logged" twice.?? I did a clean install again last night and ran ./build.? So far on my tests this morning it is operating normally again.? I did foece a few manual log entries too so that may have started the autologging up again.
--
Mike WB5LLI
|
Dave, N1AI
Jeff Kelly
HA3HZ