Re: uBitxV6 SSB transmit has RF out with no audio input, voice distorted when present
Hi Dave, I have the same problem too. In my case I have a carrier of about 0.5W out of 10W modulated by a frequency of 85Hz. The microphone has great sensitivity, such that it modulates even in the presence of only domestic environmental noises. As soon as I can I will do the suggested tests and will let you know the results.
73, by IK0XBC - Andrea
|
Re: uBitxV6 SSB transmit has RF out with no audio input, voice distorted when present
Almost EVERY piece of current?digital technique freebie software includes a spectrum display.? ?JS8, WSJT-X,, FLDGI are very common ones that all have a spectrum display.
If your BFO is in the middle of the filter....you will have a RESTRICTED frequency range of white noise coming from the antenna.? ?It will be like 0-1500 Hz for example. If you get your BFO?just to one side (where it should be) then your white?noise will expand to 2500-3000 Hz wide.?? That is another clue.
If upper sideband signals (think 20meter voice) work o the UPPERsideband setting, then you areon the correct side of the filter. If they are BACKWARDS, then you are on the wrong side, just change to the other side by changing the frequency of the BFO up or down about 2.5 KHz.
Loads of experimental learning available from this effort.? ?Exactly the same as what my chemistry students will be doing in just a few hours
Gordon
On Mon, Nov 6, 2023 at 3:12?AM Evan Hand < elhandjr@...> wrote: On Sun, Nov 5, 2023 at 07:43 PM, DaveW, KL7HJF wrote:
Gordon,
Your explanation does describe what I am seeing.
Are you familiar with the KD8CEC "IF Shift" function? Perhaps there is a way that might be used as a tool for determining whether the BFO is adjusted to the edge of the filter?
Dave,
I am familiar with the KD8CEC IF Shift function (though, as with Gordon, it has been a couple of years).? It will not help in this case.? The relationship is between the SSB crystal filter and the BFO.? That does not change with the IF shift function in the KD8CEC software.? The shift function adjusts the relationship between the BFO and the IF frequency.
The best way that I found to measure the filter reduction is with a spectrum analyzer and a tone generator.? The one that I used at the time was an SDRPlay RSP1A and the spectrum analyzer software on the SDRPlay website.? Like you, I had an Antuino and tried that.? The issue is the minimum resolution is about 3kHz.? It's too wide to see the difference between the tone and the carrier.? That was true for the other low-cost analyzers that I could find.? The TinySA became available, but it is also limited.? The newer TinySA Ultra can go down to 200Hz resolution bandwidth (RBW), BUT it takes over 8 seconds for a scan.? That could work if you are patient and have a good, constant tone.? It would be a test, adjust, and test again with as much as 20 to 30 seconds between adjustments.
If I needed to do the analysis again, I would return to the RSP1A, as it could go down to 10HZ RBW with a good update time.? It has other weaknesses that keep it from becoming my go-to analyzer.? The ideal is either a Siglent or Rigol analyzer.? Maybe this year, I will break down and buy one of them.
Without the spectrum analyzer function, the method Gordon suggests is the only one I can think of.
Have fun experimenting and learning!
73
Evan
AC9TU
|
Re: uBitxV6 SSB transmit has RF out with no audio input, voice distorted when present
On Sun, Nov 5, 2023 at 07:43 PM, DaveW, KL7HJF wrote:
Gordon,
Your explanation does describe what I am seeing.
Are you familiar with the KD8CEC "IF Shift" function? Perhaps there is a way that might be used as a tool for determining whether the BFO is adjusted to the edge of the filter?
Dave, I am familiar with the KD8CEC IF Shift function (though, as with Gordon, it has been a couple of years).? It will not help in this case.? The relationship is between the SSB crystal filter and the BFO.? That does not change with the IF shift function in the KD8CEC software.? The shift function adjusts the relationship between the BFO and the IF frequency. The best way that I found to measure the filter reduction is with a spectrum analyzer and a tone generator.? The one that I used at the time was an SDRPlay RSP1A and the spectrum analyzer software on the SDRPlay website.? Like you, I had an Antuino and tried that.? The issue is the minimum resolution is about 3kHz.? It's too wide to see the difference between the tone and the carrier.? That was true for the other low-cost analyzers that I could find.? The TinySA became available, but it is also limited.? The newer TinySA Ultra can go down to 200Hz resolution bandwidth (RBW), BUT it takes over 8 seconds for a scan.? That could work if you are patient and have a good, constant tone.? It would be a test, adjust, and test again with as much as 20 to 30 seconds between adjustments. If I needed to do the analysis again, I would return to the RSP1A, as it could go down to 10HZ RBW with a good update time.? It has other weaknesses that keep it from becoming my go-to analyzer.? The ideal is either a Siglent or Rigol analyzer.? Maybe this year, I will break down and buy one of them. Without the spectrum analyzer function, the method Gordon suggests is the only one I can think of. Have fun experimenting and learning! 73 Evan AC9TU
|
Re: uBitxV6 SSB transmit has RF out with no audio input, voice distorted when present
No, sorry, it has been years since I last turned on one of the ubitx series.? ?But this is a common problem when the filter bandpass differs a bit from unit to unit, and the si5351 also differs a bit from unit to unit.
I don't know how much you are familiar with the theory of single sideband transmission and reception via balanced mixers and sharp crystal filters, but it is what is important to understand here.? ? The ubitx series?generally put the BFO on the sharp side (I think) of the crystal filter, and then flipped a different mixer between two local oscillator frequencies in order to get the best SSB performance in both modes.? ? What you're experiencing sounds exactly like having the BFO not quite off the edge.
By moving the BFO in one direction until your transmitted carrier quiets down and then testing how well things are working (is upper side band really upper sideband?? ?is lower sideband really lower sideband?? ?Things sound OK?)? ?and if things are BACKWARDS, then move the BFO backwards the OTHER direction until your transmitter carrier quets?down THE OTHER DIRECITON, and test things again.? ?That should be a fairly straightfoward way to figure it out.
This is part of being a true radio ammateur -- adjusting, leaarning, fiddling.? ?Have fun!? ?I'm busy struggling with TWO radio mysteries right now in different areas....
Gordon
On Sun, Nov 5, 2023 at 8:43?PM DaveW, KL7HJF < david@...> wrote: Gordon,
Your explanation does describe what I am seeing.
Are you familiar with the KD8CEC "IF Shift" function? Perhaps there is a way that might be used as a tool for determining whether the BFO is adjusted to the edge of the filter?
|
Re: uBitxV6 SSB transmit has RF out with no audio input, voice distorted when present
Gordon,
Your explanation does describe what I am seeing.
Are you familiar with the KD8CEC "IF Shift" function? Perhaps there is a way that might be used as a tool for determining whether the BFO is adjusted to the edge of the filter?
|
Re: uBitxV6 SSB transmit has RF out with no audio input, voice distorted when present
If your BFO frequency?is not off to the edge of the filter, then you're going to have exactly what you are describing:
a) you will have some transmitter output, because the carrier is not being suppressed properly because the diode mixer isn't perfect and the filter should get rid of the rest b) you will have detection of both upper and lower sideband
I would suggest that if you are able to measure your filter somehow, you do so, and position the BFO to one side.? ?
Otherwise, you could try making careful adjustment up or down on the bfo until the transmitted energy without mic input drops to a small level (say,, 1/4 watt?) and then notice whether you can still hear the same received signal in both upper and lower.? ?If you have guessed the wrong direction to go, try the other direction.
Good luck -- this is an important part of setting up a filtering ssb system properly!
Gordon KX4Z
On Sun, Nov 5, 2023 at 8:26?PM DaveW, KL7HJF < david@...> wrote: Raj,
I'm sure the 1N4148 diodes (which I installed temporarily) are soldered securely in place. I'm reasonably sure the BAT45 was fully connected, because ohmmeter checks of the circuit were made from its legs. There has been nothing to indicate an intermittent connection. I will probably re-install the BAT45, but according to Ashhar's circuit description of uBitxV6, well matched 1N4148 diodes work as well as BAT45. So, there is no rush.
Evan,
I have done the BFO calibration several times. I do not use Ashhar's online tuning aid, because my computer runs Linux, and Firefox refuses to recognize my microphone. Rather than spending days chasing that issue, I installed a free audio spectrum analyzer program (JAAA, Jack/Alsa Audio Analyzer), which is easily configured to show the same parameters as the online tuning tool.
However, I suppose it is still possible that I have mistakenly tuned to a false image. As mentioned, I have two Raduinos. One runs CEC v1.2 firmware on an Arduino Nano. The other runs CEC v2 firmware on RPi-Pico. When calibrating the v2 board, I used AJ6CU's "calibration wizard" on his Settings Editor.
It does seem strange that a given SSB signal is received clearly in both LSB and USB modes. My understanding is that the selection of LSB versus USB is accomplished by changing the clock#1 frequency to the second mixer. It doesn't seem likely that both clock frequencies are present at the same time. But if they were, there would be some unexpected products in the mix.
Thank you both for your insights. If I make some progress, I'll let you know.
|
Re: uBitxV6 SSB transmit has RF out with no audio input, voice distorted when present
Raj,
I'm sure the 1N4148 diodes (which I installed temporarily) are soldered securely in place. I'm reasonably sure the BAT45 was fully connected, because ohmmeter checks of the circuit were made from its legs. There has been nothing to indicate an intermittent connection. I will probably re-install the BAT45, but according to Ashhar's circuit description of uBitxV6, well matched 1N4148 diodes work as well as BAT45. So, there is no rush.
Evan,
I have done the BFO calibration several times. I do not use Ashhar's online tuning aid, because my computer runs Linux, and Firefox refuses to recognize my microphone. Rather than spending days chasing that issue, I installed a free audio spectrum analyzer program (JAAA, Jack/Alsa Audio Analyzer), which is easily configured to show the same parameters as the online tuning tool.
However, I suppose it is still possible that I have mistakenly tuned to a false image. As mentioned, I have two Raduinos. One runs CEC v1.2 firmware on an Arduino Nano. The other runs CEC v2 firmware on RPi-Pico. When calibrating the v2 board, I used AJ6CU's "calibration wizard" on his Settings Editor.
It does seem strange that a given SSB signal is received clearly in both LSB and USB modes. My understanding is that the selection of LSB versus USB is accomplished by changing the clock#1 frequency to the second mixer. It doesn't seem likely that both clock frequencies are present at the same time. But if they were, there would be some unexpected products in the mix.
Thank you both for your insights. If I make some progress, I'll let you know.
|
Re: uBitX V6 Serial Interface Issue
Hi Jim,
The 5V from the USB interface should not damage the Si5351A.
The Si5351A uses the 3V3 which is sourced from the USB interface chip, although this is not ideal because the regulators are not very well specified.
The issue with the 5V from the USB interface is that it tries to back-feed to the 12V supply . This can be fixed by fitting a diode on the feed to the regulator on the Raduino.
For completeness, I also added a separate 3V3 regulator for the Si5351A.
/g/BITX20/message/104139Possible causes of failure of the Si5351A may be ESD or sending 5V signals to its IO pins. This may occur if the Arduino is programmed to output 5V signals on the I2C interfaces prior to programming it with the target firmware.
73, Dave
?
|
Re: uBitX V6 Serial Interface Issue
I cut the red cable in the USB extension cable to keep 5V from the host USB computer from getting to the Arduino. I did this a couple of years ago and haven't notice any ill effects. It allows the power switch to kill the power to the Arduino allowing a reboot if necessary.
--
73,
Mark, N8ME
|
uBitX V6 Serial Interface Issue
As part of my deep dive into the V6 hardware, I noticed an issue with the serial connection on the back of the radio that "could" explain a lot of the Raduino failures I've been reading about. The connection exposes itself to the 5 volts applied to the Arduino Nano. If the Nano is exposed to more than one 5v source at a time, there is potential for damage to it and the Si5351. If you want to use the rear serial connector for computer control, perhaps it would make sense to cut the 5v trace on the connector board so that potential is removed and you can safely connect it to your computer. If you want to use the rear serial connector for flashing the Nano firmware, DON'T...Instead, remove the Nano completely from the radio and flash it standalone. Then just plug the Nano back in to the Raduino. Really, how many times are you going to flash the firmware anyway?
These are my thoughts...Any other thoughts, ideas or comments? Jim WB2LHP
|
Re: uBitxV6 SSB transmit has RF out with no audio input, voice distorted when present
Dave,
Are you sure that the BFO is set correctly?? Did you use the BFO Tuining Aid web page?
73
Evan
AC9TU
|
Re: uBitxV6 SSB transmit has RF out with no audio input, voice distorted when present
Dave,
I have seen a few boards with BAT54 legs not making good contact. The moment I resoldered them
the sensitivity went and other problems vanished. Start by resoldering D5.
Raj
toggle quoted message
Show quoted text
On 05/11/2023 5:58 AM, DaveW, KL7HJF wrote: I thought my problems had been ironed out, after fixing a Raduino problem (bad si5351a), and performing calibration several times.
The receiver is on frequency with good quality reception of SSB. But I was told my transmit audio is distorted.
The factory supplied mic was suspect. I replaced the unshielded coil cord with a shielded straight cord, to minimize RF pickup.
I checked the mic on the computer, recording into Audacity, watching scope display and listening to playback. The voice quality is not great, and is competing with a substantial AC hum. Tried another electret mic element (from junk box), with similar results in Audacity. My computer headset mic has no hum in Audacity.
But then, I noticed that even with mic disconnected from the radio, when the transmitter is activated in LSB or USB (by closing PTT at mic jack), it puts out about 3 watts. Using a 50uf cap, I squashed any residual modulation audio at TP20, with minimal effect. So, using a 0.05uf cap, I worked backward through the xmtr chain, connecting cap between ground and TP1, TP13, TP16, TP17 (one at a time, in that order), and also between the R114/R115 junction and ground. With cap in place, there is no xmtr RF output.
This seems to indicate an imbalance in the BFO mixer. DC voltage at TP20 is zero in receive and xmt. Therefore diode D1 (BAT54SL) was suspect. I pulled it out, and it tests as it should with an ohm meter. But, just in case, I replaced it with a pair of 1N4148 diodes. But the problem remains: there is RF output with no voice input.
I have a second Raduino board (one with Arduino Nano, the other RPi-Pico, both with KD8CEC software for a Nextion display, which must be moved between boards, as needed). The problem is the same with either board.
My test equipment is limited. But using an Antuino, Clock #0 is about 11.050 Mhz.
T7 looks okay with an ohmmeter (but I haven't lifted the board, or taken the transformer apart. By prodding it a bit, no intermittency appears.)
Okay, enough from me. What would you look at next?
|
Re: uBitxV6 SSB transmit has RF out with no audio input, voice distorted when present
Addition notes to the above post:
1. The 3 watt output is on 80M (on that band CW output is about 10 watts). The symptom occurs on all other bands tested, but with lower output.
2. As noted above, SSB reception is clear. However, there is not much difference between LSB and USB. An LSB station sounds good, but switching the receiver to USB makes little difference. The CEC software has an IF shift function. LSB and USB sound about the same with this function centered. LSB remains intelligible when the IF is shifted left (but not when to right). The USB remains intelligible when IF is shifted right (but not to left).
|
uBitxV6 SSB transmit has RF out with no audio input, voice distorted when present
I thought my problems had been ironed out, after fixing a Raduino problem (bad si5351a), and performing calibration several times.
The receiver is on frequency with good quality reception of SSB. But I was told my transmit audio is distorted.
The factory supplied mic was suspect. I replaced the unshielded coil cord with a shielded straight cord, to minimize RF pickup.
I checked the mic on the computer, recording into Audacity, watching scope display and listening to playback. The voice quality is not great, and is competing with a substantial AC hum. Tried another electret mic element (from junk box), with similar results in Audacity. My computer headset mic has no hum in Audacity.
But then, I noticed that even with mic disconnected from the radio, when the transmitter is activated in LSB or USB (by closing PTT at mic jack), it puts out about 3 watts. Using a 50uf cap, I squashed any residual modulation audio at TP20, with minimal effect. So, using a 0.05uf cap, I worked backward through the xmtr chain, connecting cap between ground and TP1, TP13, TP16, TP17 (one at a time, in that order), and also between the R114/R115 junction and ground. With cap in place, there is no xmtr RF output.
This seems to indicate an imbalance in the BFO mixer. DC voltage at TP20 is zero in receive and xmt. Therefore diode D1 (BAT54SL) was suspect. I pulled it out, and it tests as it should with an ohm meter. But, just in case, I replaced it with a pair of 1N4148 diodes. But the problem remains: there is RF output with no voice input.
I have a second Raduino board (one with Arduino Nano, the other RPi-Pico, both with KD8CEC software for a Nextion display, which must be moved between boards, as needed). The problem is the same with either board.
My test equipment is limited. But using an Antuino, Clock #0 is about 11.050 Mhz.
T7 looks okay with an ohmmeter (but I haven't lifted the board, or taken the transformer apart. By prodding it a bit, no intermittency appears.)
Okay, enough from me. What would you look at next?
|
Re: Adding zener to protect mosfets
No, that would be extremely bad engineering judgement.? Using 3.3V-12.6V tranzorbs might be more appropriate depending upon the rest of the circuit design. ? ? Dr. William J. Schmidt - K9HZ J68HZ 8P6HK ZF2HZ PJ4/K9HZ VP5/K9HZ PJ2/K9HZ VP2EHZ ? Owner - Operator Big Signal Ranch – K9ZC Staunton, Illinois ? Owner – Operator Villa Grand Piton – J68HZ Soufriere, St. Lucia W.I. Rent it: ? Moderator: North American QRO Group at Groups.IO. Moderator: Amateur Radio Builders Group at Groups.IO. ? email:? bill@... ? ?
toggle quoted message
Show quoted text
From: BITX20@groups.io <BITX20@groups.io> On Behalf Of Jim Strohm Sent: Saturday, November 4, 2023 7:05 PM To: BITX20@groups.io Subject: Re: [BITX20] Adding zener to protect mosfets ? Are you suggesting Tranzorbs or other suppressor diodes like TVSes in the 45 to 50 volt range for BS170s? Like Vishay Tranzorb?SMBJ48D ?
|
Re: Adding zener to protect mosfets
Are you suggesting Tranzorbs or other suppressor diodes like TVSes in the 45 to 50 volt range for BS170s?
Like Vishay Tranzorb?SMBJ48D ?
73 Jim N6OTQ
|
Re: Adding zener to protect mosfets
Just be aware that there are single direction and bidirectional TVS diodes - you want bidirectional for this application.
As for hams not having them in their junk box - I actually do have half a dozen in mine and as mentioned above, they are not expensive nor hard to get a hold of.
--
|
Re: Adding zener to protect mosfets
Are these the same as TVS diodes? ?I have used those before. I thought they were zeners. ?Live and learn !!
toggle quoted message
Show quoted text
On Nov 4, 2023, at 14:17, HA3HZ <gyula@...> wrote:
?
--
Gyula HA3HZ
|
Evan and Jack,
Thank you both for your reply. The links from Evan provided interesting reading. The first one included contained a post from Ashhar regarding his reason for providing CW1 and CW2, to facilitate adding a narrow bandwidth filter (at about 1.8 Mhz) in parrallel with the existing 11 Mhz filter. Filter election would be accomplished by software changing clock frequencies as appropriate. The other posts, in the thread Evan mentioned, highlight the problems involved in implementing an additional crystal filter.
I have been considering the Hi-per-mite audio filter. It seems to have very positive reviews.
|
Dave, I searched "ubitx cw crystal filter" using Google and found the following: /g/BITX20/topic/23176080Most hams (including myself) use audio filters for CW.? They are easier to implement and do not require software changes.? Here are details on that approach by K4LXY: You can also search in Groups.io directly, but I have found that interface challenging to use. Note that I am NOT a CW operator.? The mods I made were for other hams that operate CW.? All have been happy with the results. Raj may have better information than I do. 73 Evan AC9TU
|
IK0XBC
Mark Erbaugh
Raj vu2zap
K9HZ