On Nov 3, 2023, at 13:09, Steve Beckman <ssteven3sb@...> wrote:


I have back-to-back Zeners from the final FET Gates to Ground on my DE. (No series resistors) Before installing them I tested how well they would clamp at RF frequencies, due to concerns I had about their speed. My tests using a scope and RF at 30 MHz showed that the 5 Watt Zeners I selected did in fact clamp at their design voltage.

On Nov 3, 2023, at 1:29 PM, Robert D. Bowers <n4fbz@...> wrote:

?If I remember right, there are also avalanche breakdown diodes that aren't zeners and have a very quick reaction time, which might do the job.? I also remember a circuit from years ago that limited RF (rather like a zener for DC circuits), but I don't remember where I read about the circuit - I think it was meant to help control the RF going into a big amp.? I remember it vaguely - from back when I first got my ticket (1980), but since I didn't have an amp, it was just something interesting.? It was in one of the magazines from the 70s or maybe early 80s.

I may still have information on that breakdown diode (it was used as a trigger for laser flashtube circuits and more).? Once I find a little of that rare commodity, free time, I could look through my files.

I hope that might prove helpful!

Would reducing the power out help to reduce frying the finals in the radio?? I remember having customers (CBers) who fussed about getting every watt possible out of their radios - they'd even find and cut the ALC diode because they were told that would hugely increase their power out on a meter.? They wouldn't understand that there was little difference between 3 3/4 watts out and 4 watts out - and that cutting the ALC diode only increased splatter and interference.? I could show them on a scope what actually happened, but they insisted that a wide bandwidth scope was less revealing than a meter.

3db down doesn't make that much difference in being heard - conditions and mode are far more important in 'reaching out'.? 3db is a reduction of half the power if I remember my math correctly. I've made contacts using less than a milliwatt (400 miles if I remember right) on HF and used to be on 10 meters a long time ago, regularly making contacts on 5 watts or less (to Australia and New Zealand).? Some people find that hard to believe - that maximized power out doesn't make that much of a difference, unless you can do it in big amounts (like x10 or more).

Bob
N4FBZ


On 11/3/23 11:35, Ashhar Farhan wrote:

Gordon,

This is in response to your idea of adding zeners to protect the gates. It doesn't work.

I added a zener 9v diode and 1k resistor (in series). I was expecting this to clip between 9v and -0.6v. Nothing of the sort happened.

I kept decreasing the tx frequency. The clipping sets in below 1 MHz.

The zener is not a fast diode!

The other challenge is that the zener also behaves as a varactor. I measured the reverse capacitance to be in the whereabouts of 50 pf. This will decrease the effective drive to the gates even more.

Notes to myself, RTFM! I discovered that some datasheets that do characterise it, quote the switching speed of zeners as 300 KHz or so.

- f

Thanks, David...the kind words are sincerely appreciated.

Jack, W8TEE

On Friday, November 3, 2023 at 12:11:10 PM EDT, david todd via groups.io <ac9xh@...> wrote:


Ashhar,

Well said ole Master of electronics!!!!!!. Jack ,Well said to u too ole Master of electronics too.

Remember folks, this is totally open sourced, and allowed to be ,odded and hacked.
following the forum, you will get a sense of the development of the Sbitx. Following that, gives u an idea of what MAY be expected from a Ham buying it.

Anyone, buying Ashhars transceivers from the bitx40 when he produced them thru the varoius versions of the ubitx thru v 6 and now the Sbitx, knows that there will be a slight learning curve on the units. its open sourced. Of course there will be mods, hardware changes upgrades and such. thats only natural.

And to get this level of communication and support from Ashhar himself PROVES that he is always dedicated to make things right with the hardware, software.

This is why this forum is here. it lists about every problem that has come up with ANY transceiver Ashhar has created.I agree with Jack. Ashhar and others who develop amazing products for the Amateur Radio Community COULD decide to just ignore everyone who bought their products, but they DONT. they work endlessly to solve even the small problems. This is what makes a good developer. CARING about their creation and taking pride in it and able to listen to anyone in regards criticism.

I for one applaud these creators today.

TO ASHHAR AND JACK AND OTHERS WHO DEVELOP NEW KITS AND SUCH:

3 CHEERS TO THEM:

HIP HIP HURRAH!!!
HIP HIP HURRAH!!!
HIP HIP HURRAH!!!!


73
DAVID
AC9XH

On Friday, November 3, 2023 at 10:45:08 AM EDT, Gordon Gibby <docvacuumtubes@...> wrote:


As I’ve said many times, this is an amazing development spearheaded by one extraordinarily gifted person.

People who don’t see what has been accomplished are too quick to make complaints

I do have concerns about the cleanness of the signal. ?My own testing may be faulty, it has certainly improved, but my particular sBitx V2 (Ppurchased second hand) ?does not look like an ICOM 718 in terms of cleanness.

I have now demonstrated that a reasonably clean signal is very possible without any improvement at all of the amplifier chain.

I think this is a foundational problem that must be solved. I don’t know how many units it affects nor where is the problem. ?My suspicions are that there is variability in the components utilized. ?Leading to variability in the production run devices. I would welcome others trying to replicate my studies. ? My next moves are to try to add, piece by piece, additional portions of the bandpass filtering systems, trying to add the most innocuous portions first. ? So far I’ve managed to show that the HV voltage doubling circuit does not seem to be a major issue at all

Following suggestions by Evan and others, I think I will add diode input or output ?to the non-used filter sections next — to see whether by some unknown method the back biased diode‘s are adding harmonics.

Then try adding the diode input to the filter in use. ?I suspect that may add some problems

Then lastly add diode output from the filter in use, (which I have already shown produces a harmonic that appears grossly out of spec).


If we can show that it is just certain 1N4007 diode‘s, then it is a relatively simple process to make some sort of a jig to test for “good” versus “bad” diodes, and put only good ones in the switching circuits. Or if we can show some variation of the circuit solves the problem then that would be the route

I think the fact that I’ve been able to take a stock unit (as purchased second hand) and by simple circuit deletions, ?change it from noncompliance (bad) to generally compliant (good!) Oon the worst of the bands studied, demonstrates that the issue not only exists (In an unknown percentage of units and on some bands) ?but is can be caused by some certain portion of the schematic — thereby, fixable!

I’m guessing that not many people have the capabilities of making these exquisite measurements. ? I hope I’m making them properly. It has been a learning curve for me. My only confidence comes from being able to demonstrate success not only for commercial icons but for the sbitx ?after changes. ? This radio has fantastic potential and in my mind is one of the most amazing achievements in ham radio in the recent years. ? To bring some thing like this to market in large quantity and available with so much open source is simply amazing. ?But I think we have to solve the manufacturing issues completely.

And it looks pretty simple to me to add Zener diode’s to the gates of the existing finals. ?I haven’t done that yet because I’m trying to keep this unit as stock as possible while I work on the harmonics. ?But once I get that solved I’ll simply add the zeners (Just ?as I did with the development unit) ?and I’m perfectly happy so far with the final MOSFETs. ?So others may wish to go with the 510s but I’m happy with what we’ve got

If it turns out that my measurement system has some mess up, I’ll be **delighted** (and of course red-faced!) ?for someone to show me where I have made it! ? That’s one reason I produce pictures so everybody can see what I’m doing

Gordon Kx4z

On Nov 3, 2023, at 10:27, Gerald Sherman <ve4gks@...> wrote:

?

This is with regard to Jack's comments about those who know what it is like to take a bunch of components and turn them into that first contact.? I did that over half a century ago.? The parts came from what was scavenged from everything from a war surplus TVI generator to an old automotive air filter (that last provided the perforated cover for the 1625 PA tube "doghouse").? My programming experience is largely with the RS274 G-code used in machine tool control - not easily transferred to C, C++, Python, etc.? Unfortunately, for some of us, our years catch up with us, and we have to pass the tools to the next generation.? My eyes aren't what they used to be, and the hands aren't very steady any more.? My abilities are largely confined to something like assembling a uBIT 6 and yes, I have a couple.? One is assembled as it was shipped, the other one is a much-modified unit from Howard Zehr.

Gerry Sherman

Sent by the Windows 11 Thunderbird

On 2023-11-03 08:57, Jack, W8TEE via groups.io wrote:
All:

I'm pretty sure that Hans agrees with me that some people here have been far too unkind to Farhan and have ignored all that he has brought to QRP operators all over the world. Farhan, Hans, and myself, are trying to cater to a different type of operator than someone who is "an appliance operator". We want to market to the operator who enjoys not only operating their radios, but also the journey of taking a bunch of components, piecing them together, and feeling the thrill of making that first contact with something that was just a pile of parts only a few days ago.

Open Source is a partnership where others join the originator to extend and improve the project. Those who see it differently are probably better off buying a commercial radio. Anyone who knows Farhan or Hans also know that they strive to make their customers happy and are deeply saddened when that's not the case. They should also know that Farhan will work tirelessly to "make it right". Given all that he has brought to our tables, we need to show patience while he does just that.

Jack, W8TEE

On Friday, November 3, 2023 at 09:26:36 AM EDT, Ashhar Farhan <farhanbox@...> <mailto:farhanbox@...> wrote:


Dear All,

I must apologize for the late reply to this thread. I am getting two threads on my inbox so I am not sure which one to reply to, I will probably copy paste this into both the threads.
Before I address the problems themselves, I must explain my own behaviour of being quiet in all these emails.

To begin with, I have written about 95% of the code on my own. I borrowed some code from Phil Karn, KA9Q's sdr receiver and some libraries like the FFTW and the Mongoose. All in all, as of today, there are approximately 30,000 lines of code. The industry average productivity of a systems level programmer is 50 lines of production grade code per day. This amounts to about 600 full working days of work. I work erratically, so on good days I manage even 300 lines of code and on other days I am just staring at some unexplained html bug for hours. All this code is not just open source but all free. It can be used by anyone to produce their own commercial radios, it can be modified, extended, and contributed to. I have also taken great care to keep the code simple and understandable.? The code belongs to the community.

I have also always acknowledged and credited the software ideas, code and the circuit origins to where credit is due. Hams build radios, learning from others. However, I have to keep track of what can be easily replicated by others as well in the bitx radio designs. Though HF SIgnals does produce bitx radios, neither their name (bitx) nor is the circuitowned by anyone in a proprietary way. We would like to actively see more people adopt the architecture and build radios around them. Earlier, Hendricks kits and now Inkits produce these designs and their modifications. We never had any commercial dealings with any of them. I do meet Bob Hendricks on the sidelines of Dayton who was kind enough to present me with one of the original?BITX20A kits, I have never met Sunil of Inkits. (The only kits that I do enjoy and build are from QRP Labs!).

I am just mentioning all this to put into perspective that we are not just a "Vendor", we are _also_ a vendor of open source radios. In the last 9 months, I have released three updates to the hardware and three updates to the software. The software was, of course, free. The hardware changes needed (for instance, the one that made DE use diode T/R instead of the relays) was posted as an upgrade of parts and PCB to all the existing customers. However, from sbitx v2 onwards, the spurs are under the FCC limit (in all our own testing, however, Gordon Gibby's tests are worrying). The main hardware challenge has been to prevent blow out of the PA. I think we have a fix. I am just nervous to lead everyone through yet another change. However, it is a simple fix, swap the IRFZ24Ns with the trusty IRF510s. They will give 25 watts on the lower bands and more than 10 watts even on 10 meters. A few changes of resistors are in order with the predriver. I will publish the mod after a few more days. Three friends extensively used this radio during WW SSB contest this weekend. The results have been heartening.

A new version of software is ready as well. I am cleaning up the HTML UI. My aim with this release is to make it usable even on a raspberry zero, this meant that the native UI should work as well (Chrome doesn't open on zero due to low memory). I am working to tweak the HTML UI? in addition to making the native GTK have complete functionality of the HTML UI as well.? So, here is where the new version (V3) is:

1. Circuit Changes: There are no changes, except that sooner, we will swap the IRFZ24Ns with IRF510s. The maximum power will be reduced from 40 watts to 25 watts. This is far less of a consequence than we think. On the oscilloscope, the peak voltage is 100v instead of 125v. However, the IRF510s have 100 V rating instead of the IRFZ24Ns that are rated for 60V. Some resistor values are tweaked to keep all the spurs down to within the FCC limits (-43 db below the peak output). Additionally, the higher bands like 10M and 12M will have higher power output of 10 watts.

The new software will work with all previous versions from sBitx DE onwards.

2. FT8: The FT8 has been fixed. I have been running FT8 on native UI as well as HTML. The timing issue has been fixed, some bugs that confused the sequencing of the messages is also fixed.

3. CW: I wrote a new CW decoder that works very well. I have been using it to Run as well as Search & Pounce. The new logger, logbook, macros and the decoder works pretty well. I have gone back to being a QRP CW operator with this.

4. Apart from this, the Hamlib's NET Ctrl allows all external programs like the WSJT-X, FLDigi can be used if the native CW/FT8 are not good enough. The issues with interoperability has been fixed.

I am going to host this version as a different fork so as to not confuse the regular users. A few of us will try this out before declaring it done.

This update has taken far more time than I thought (the software takes more time you think it would, even if you have taken this into account). There were many niggling hardware troubles including some problems with our? SMD assembler, they have been sorted out too though for a few days we were despondent enough to consider ordering SMD assembly from China instead of our regular SMD assembly shop (the moment passed).

I am grateful to the community that supports all of us who develop sbitx. I request you all to be kinder to criticism of the radio. There is no excuse as a programmer to ship out buggy code and I personally take the entire blame. Perhaps I have been ambitious to pull this off as a single contributor. We must collectively identify and fix the bugs and keep adding features. Our endeavour will be to make the basic radio as stable as we can, and then make it extendible in easy and simple ways. This is a radio that can grow younger with age, that is the magic of free software.

73, and cu on 7015 soon.
de f.









On Sat, Oct 28, 2023 at 3:23?AM Gordon Gibby <docvacuumtubes@... <mailto:docvacuumtubes@...>> wrote:

Jack, I am with you. I have learned so much— so so much! from this
group and from these radios. I am a little bit concerned about
spurious signals ?in the manufacturing production line. This is
such a fantastic advance that it deserves to be a very clean radio
and I’m concerned that we’ve got an issue with the filtering and
bleed around. The filters them selves look fantastic!? I
understand this had happened to other firms’ radios as well and so
it’s not unheard of with new devices. I’m pretty convinced that we
will get it fixed and then I’m looking forward to moving on to
working on software, I’m especially fascinated with getting back
into digital signal processing



But I’m just swamped with the high school students I’m teaching
right now. Today we had them do their first CW contacts with a
roundabout technique using a web SDR. They’re doing around six
words per minute. I need them up to 15 by winter field day.

Gordon

On Oct 27, 2023, at 11:56, John Terrell, N6LN <N6LN@...
<mailto:N6LN@...>> wrote:

?

I have been lucky enough to work with my two sBitx rigs (the DE
and also the V2) and the uBitx V6 before that. I have learned far
more along the way about transmitter and receiver design than I
could have any other way. If I want an easy contact I can simply
switch on my Icom or Kenwood or Yaesu, but that’s not what I
wanted. I have benefited beyond measure from the community of
engineers and fellow tinkerers who have helped me. Hopefully I
have helped others through my posts as they have walked me
through my fits and stops. It has at times been frustrating but
all in all I haven’t had this much pleasure in ham radio since I
was a teenager going to TV repair shops, taking apart junked tube
TV sets and building homebrew rigs. It’s like hiking for days in
the mountains. There’s no hot water or wifi, but that’s not the
reason to make the trip.

Enjoy the trip.

73,
Jack, N6LN

--
Jack, W8TEE

--
Jack, W8TEE

Gordon,
This is in response to your idea of adding zeners to protect the gates. It doesn't work.
I added a zener 9v diode and 1k resistor (in series). I was expecting this to clip between 9v and -0.6v. Nothing of the sort happened.
I kept decreasing the tx frequency. The clipping sets in below 1 MHz.
The zener is not a fast diode!
The other challenge is that the zener also behaves as a varactor. I measured the reverse capacitance to be in the whereabouts of 50 pf. This will decrease the effective drive to the gates even more.
Notes to myself, RTFM! I discovered that some datasheets that do characterise it, quote the switching speed of zeners as 300 KHz or so.
- f

On Nov 3, 2023, at 1:09 PM, Steve Beckman <ssteven3sb@...> wrote:

?Farhan, Gordon;

I have back-to-back Zeners from the final FET Gates to Ground on my DE. (No series resistors) Before installing them I tested how well they would clamp at RF frequencies, due to concerns I had about their speed. My tests using a scope and RF at 30 MHz showed that the 5 Watt Zeners I selected did in fact clamp at their design voltage.

The idea is one recommended by W1FB and at least one MACOM spec sheet (MRF275G). I have also come across sources that claim this is a bad idea.

At the time, we were all looking for the cause of final FET failures in the DE. We found low frequency spurious oscillations in the V1, caused by feedback paths in the TR switching circuit, TR Timing problems, excessive drive capability, and excessive low frequency gain. Those issues have since been addressed, reducing if not eliminating the potential for an over-voltage situation at the gates of the final FETs.

I've left the Zeners installed in my DE. As we make progress towards reducing harmonics, spurs, IMD, etc, they may end up getting removed.

73; Steve, N3SB

On Nov 3, 2023, at 11:36, Ashhar Farhan <farhanbox@...> wrote:

?

Gordon,

This is in response to your idea of adding zeners to protect the gates. It doesn't work.

I added a zener 9v diode and 1k resistor (in series). I was expecting this to clip between 9v and -0.6v. Nothing of the sort happened.?

I kept decreasing the tx frequency. The clipping sets in below 1 MHz.?

The zener is not a fast diode!

The other challenge is that the zener also behaves as a varactor. I measured the reverse capacitance to be in the whereabouts of 50 pf. This will decrease the effective drive to the gates even more.

Notes to myself, RTFM! I discovered that some datasheets that do characterise it, quote the switching speed of zeners as 300 KHz or so.?

- f

On Nov 3, 2023, at 10:27, Gerald Sherman <ve4gks@...> wrote:

?

This is with regard to Jack's comments about those who know what it is like to take a bunch of components and turn them into that first contact.? I did that over half a century ago.? The parts came from what was scavenged from everything from a war surplus TVI generator to an old automotive air filter (that last provided the perforated cover for the 1625 PA tube "doghouse").? My programming experience is largely with the RS274 G-code used in machine tool control - not easily transferred to C, C++, Python, etc.? Unfortunately, for some of us, our years catch up with us, and we have to pass the tools to the next generation.? My eyes aren't what they used to be, and the hands aren't very steady any more.? My abilities are largely confined to something like assembling a uBIT 6 and yes, I have a couple.? One is assembled as it was shipped, the other one is a much-modified unit from Howard Zehr.

Gerry Sherman

Sent by the Windows 11 Thunderbird

On 2023-11-03 08:57, Jack, W8TEE via groups.io wrote:

All:

I'm pretty sure that Hans agrees with me that some people here have been far too unkind to Farhan and have ignored all that he has brought to QRP operators all over the world. Farhan, Hans, and myself, are trying to cater to a different type of operator than someone who is "an appliance operator". We want to market to the operator who enjoys not only operating their radios, but also the journey of taking a bunch of components, piecing them together, and feeling the thrill of making that first contact with something that was just a pile of parts only a few days ago.

Open Source is a partnership where others join the originator to extend and improve the project. Those who see it differently are probably better off buying a commercial radio. Anyone who knows Farhan or Hans also know that they strive to make their customers happy and are deeply saddened when that's not the case. They should also know that Farhan will work tirelessly to "make it right". Given all that he has brought to our tables, we need to show patience while he does just that.

Jack, W8TEE

On Friday, November 3, 2023 at 09:26:36 AM EDT, Ashhar Farhan <farhanbox@...> wrote:

Dear All,

I must apologize for the late reply to this thread. I am getting two threads on my inbox so I am not sure which one to reply to, I will probably copy paste this into both the threads.?

Before I address the problems themselves, I must explain my own behaviour of being quiet in all these emails.?

To begin with, I have written about 95% of the code on my own. I borrowed some code from Phil Karn, KA9Q's sdr receiver and some libraries like the FFTW and the Mongoose. All in all, as of today, there are approximately 30,000 lines of code. The industry average productivity of a systems level programmer is 50 lines of production grade code per day. This amounts to about 600 full working days of work. I work erratically, so on good days I manage even 300 lines of code and on other days I am just staring at some unexplained html bug for hours. All this code is not just open source but all free. It can be used by anyone to produce their own commercial radios, it can be modified, extended, and contributed to. I have also taken great care to keep the code simple and understandable.? The code belongs to the community.

I have also always acknowledged and credited the software ideas, code and the circuit origins to where credit is due. Hams build radios, learning from others. However, I have to keep track of what can be easily replicated by others as well in the bitx radio designs. Though HF SIgnals does produce bitx radios, neither their name (bitx) nor is the circuitowned by anyone in a proprietary way. We would like to actively see more people adopt the architecture and build radios around them. Earlier, Hendricks kits and now Inkits produce these designs and their modifications. We never had any commercial dealings with any of them. I do meet Bob Hendricks on the sidelines of Dayton who was kind enough to present me with one of the original?BITX20A kits, I have never met Sunil of Inkits. (The only kits that I do enjoy and build are from QRP Labs!).

I am just mentioning all this to put into perspective that we are not just a "Vendor", we are _also_ a vendor of open source radios. In the last 9 months, I have released three updates to the hardware and three updates to the software. The software was, of course, free. The hardware changes needed (for instance, the one that made DE use diode T/R instead of the relays) was posted as an upgrade of parts and PCB to all the existing customers. However, from sbitx v2 onwards, the spurs are under the FCC limit (in all our own testing, however, Gordon Gibby's tests are worrying). The main hardware challenge has been to prevent blow out of the PA. I think we have a fix. I am just nervous to lead everyone through yet another change. However, it is a simple fix, swap the IRFZ24Ns with the trusty IRF510s. They will give 25 watts on the lower bands and more than 10 watts even on 10 meters. A few changes of resistors are in order with the predriver. I will publish the mod after a few more days. Three friends extensively used this radio during WW SSB contest this weekend. The results have been heartening.

A new version of software is ready as well. I am cleaning up the HTML UI. My aim with this release is to make it usable even on a raspberry zero, this meant that the native UI should work as well (Chrome doesn't open on zero due to low memory). I am working to tweak the HTML UI? in addition to making the native GTK have complete functionality of the HTML UI as well.? So, here is where the new version (V3) is:

1. Circuit Changes: There are no changes, except that sooner, we will swap the IRFZ24Ns with IRF510s. The maximum power will be reduced from 40 watts to 25 watts. This is far less of a consequence than we think. On the oscilloscope, the peak voltage is 100v instead of 125v. However, the IRF510s have 100 V rating instead of the IRFZ24Ns that are rated for 60V. Some resistor values are tweaked to keep all the spurs down to within the FCC limits (-43 db below the peak output). Additionally, the higher bands like 10M and 12M will have higher power output of 10 watts.

The new software will work with all previous versions from sBitx DE onwards.

2. FT8: The FT8 has been fixed. I have been running FT8 on native UI as well as HTML. The timing issue has been fixed, some bugs that confused the sequencing of the messages is also fixed.

3. CW: I wrote a new CW decoder that works very well. I have been using it to Run as well as Search & Pounce. The new logger, logbook, macros and the decoder works pretty well. I have gone back to being a QRP CW operator with this.

4. Apart from this, the Hamlib's NET Ctrl allows all external programs like the WSJT-X, FLDigi can be used if the native CW/FT8 are not good enough. The issues with interoperability has been fixed.

I am going to host this version as a different fork so as to not confuse the regular users. A few of us will try this out before declaring it done.

This update has taken far more time than I thought (the software takes more time you think it would, even if you have taken this into account). There were many niggling hardware troubles including some problems with our? SMD assembler, they have been sorted out too though for a few days we were despondent enough to consider ordering SMD assembly from China instead of our regular SMD assembly shop (the moment passed).

I am grateful to the community that supports all of us who develop sbitx. I request you all to be kinder to criticism of the radio. There is no excuse as a programmer to ship out buggy code and I personally take the entire blame. Perhaps I have been ambitious to pull this off as a single contributor. We must collectively identify and fix the bugs and keep adding features. Our endeavour will be to make the basic radio as stable as we can, and then make it extendible in easy and simple ways. This is a radio that can grow younger with age, that is the magic of free software.

73, and cu on 7015 soon.

de f.

?

On Sat, Oct 28, 2023 at 3:23?AM Gordon Gibby <docvacuumtubes@...> wrote:

Jack, I am with you. I have learned so much— so so much! from this group and from these radios. I am a little bit concerned about spurious signals ?in the manufacturing production line. This is such a fantastic advance that it deserves to be a very clean radio and I’m concerned that we’ve got an issue with the filtering and bleed around. The filters them selves look fantastic!? I understand this had happened to other firms’ radios as well and so it’s not unheard of with new devices. I’m pretty convinced that we will get it fixed and then I’m looking forward to moving on to working on software, I’m especially fascinated with getting back into digital signal processing?

But I’m just swamped with the high school students I’m teaching right now. Today we had them do their first CW contacts with a roundabout technique using a web SDR.? They’re doing around six words per minute. I need them up to 15 by winter field day.?

Gordon

On Oct 27, 2023, at 11:56, John Terrell, N6LN <N6LN@...> wrote:

?

I have been lucky enough to work with my two sBitx rigs (the DE and also the V2) and the uBitx V6 before that. I have learned far more along the way about transmitter and receiver design than I could have any other way. If I want an easy contact I can simply switch on my Icom or Kenwood or Yaesu, but that’s not what I wanted. I have benefited beyond measure from the community of engineers and fellow tinkerers who have helped me. Hopefully I have helped others through my posts as they have walked me through my fits and stops. It has at times been frustrating but all in all I haven’t had this much pleasure in ham radio since I was a teenager going to TV repair shops, taking apart junked tube TV sets and building homebrew rigs. It’s like hiking for days in the mountains. There’s no hot water or wifi, but that’s not the reason to make the trip.?

Enjoy the trip.

73,
Jack, N6LN

--
Jack, W8TEE

All:

I'm pretty sure that Hans agrees with me that some people here have been far too unkind to Farhan and have ignored all that he has brought to QRP operators all over the world. Farhan, Hans, and myself, are trying to cater to a different type of operator than someone who is "an appliance operator". We want to market to the operator who enjoys not only operating their radios, but also the journey of taking a bunch of components, piecing them together, and feeling the thrill of making that first contact with something that was just a pile of parts only a few days ago.

Open Source is a partnership where others join the originator to extend and improve the project. Those who see it differently are probably better off buying a commercial radio. Anyone who knows Farhan or Hans also know that they strive to make their customers happy and are deeply saddened when that's not the case. They should also know that Farhan will work tirelessly to "make it right". Given all that he has brought to our tables, we need to show patience while he does just that.

Jack, W8TEE

On Friday, November 3, 2023 at 09:26:36 AM EDT, Ashhar Farhan <farhanbox@...> wrote:

Dear All,

I must apologize for the late reply to this thread. I am getting two threads on my inbox so I am not sure which one to reply to, I will probably copy paste this into both the threads.?

Before I address the problems themselves, I must explain my own behaviour of being quiet in all these emails.?

To begin with, I have written about 95% of the code on my own. I borrowed some code from Phil Karn, KA9Q's sdr receiver and some libraries like the FFTW and the Mongoose. All in all, as of today, there are approximately 30,000 lines of code. The industry average productivity of a systems level programmer is 50 lines of production grade code per day. This amounts to about 600 full working days of work. I work erratically, so on good days I manage even 300 lines of code and on other days I am just staring at some unexplained html bug for hours. All this code is not just open source but all free. It can be used by anyone to produce their own commercial radios, it can be modified, extended, and contributed to. I have also taken great care to keep the code simple and understandable.? The code belongs to the community.

I have also always acknowledged and credited the software ideas, code and the circuit origins to where credit is due. Hams build radios, learning from others. However, I have to keep track of what can be easily replicated by others as well in the bitx radio designs. Though HF SIgnals does produce bitx radios, neither their name (bitx) nor is the circuitowned by anyone in a proprietary way. We would like to actively see more people adopt the architecture and build radios around them. Earlier, Hendricks kits and now Inkits produce these designs and their modifications. We never had any commercial dealings with any of them. I do meet Bob Hendricks on the sidelines of Dayton who was kind enough to present me with one of the original?BITX20A kits, I have never met Sunil of Inkits. (The only kits that I do enjoy and build are from QRP Labs!).

I am just mentioning all this to put into perspective that we are not just a "Vendor", we are _also_ a vendor of open source radios. In the last 9 months, I have released three updates to the hardware and three updates to the software. The software was, of course, free. The hardware changes needed (for instance, the one that made DE use diode T/R instead of the relays) was posted as an upgrade of parts and PCB to all the existing customers. However, from sbitx v2 onwards, the spurs are under the FCC limit (in all our own testing, however, Gordon Gibby's tests are worrying). The main hardware challenge has been to prevent blow out of the PA. I think we have a fix. I am just nervous to lead everyone through yet another change. However, it is a simple fix, swap the IRFZ24Ns with the trusty IRF510s. They will give 25 watts on the lower bands and more than 10 watts even on 10 meters. A few changes of resistors are in order with the predriver. I will publish the mod after a few more days. Three friends extensively used this radio during WW SSB contest this weekend. The results have been heartening.

A new version of software is ready as well. I am cleaning up the HTML UI. My aim with this release is to make it usable even on a raspberry zero, this meant that the native UI should work as well (Chrome doesn't open on zero due to low memory). I am working to tweak the HTML UI? in addition to making the native GTK have complete functionality of the HTML UI as well.? So, here is where the new version (V3) is:

1. Circuit Changes: There are no changes, except that sooner, we will swap the IRFZ24Ns with IRF510s. The maximum power will be reduced from 40 watts to 25 watts. This is far less of a consequence than we think. On the oscilloscope, the peak voltage is 100v instead of 125v. However, the IRF510s have 100 V rating instead of the IRFZ24Ns that are rated for 60V. Some resistor values are tweaked to keep all the spurs down to within the FCC limits (-43 db below the peak output). Additionally, the higher bands like 10M and 12M will have higher power output of 10 watts.

The new software will work with all previous versions from sBitx DE onwards.

2. FT8: The FT8 has been fixed. I have been running FT8 on native UI as well as HTML. The timing issue has been fixed, some bugs that confused the sequencing of the messages is also fixed.

3. CW: I wrote a new CW decoder that works very well. I have been using it to Run as well as Search & Pounce. The new logger, logbook, macros and the decoder works pretty well. I have gone back to being a QRP CW operator with this.

4. Apart from this, the Hamlib's NET Ctrl allows all external programs like the WSJT-X, FLDigi can be used if the native CW/FT8 are not good enough. The issues with interoperability has been fixed.

I am going to host this version as a different fork so as to not confuse the regular users. A few of us will try this out before declaring it done.

This update has taken far more time than I thought (the software takes more time you think it would, even if you have taken this into account). There were many niggling hardware troubles including some problems with our? SMD assembler, they have been sorted out too though for a few days we were despondent enough to consider ordering SMD assembly from China instead of our regular SMD assembly shop (the moment passed).

I am grateful to the community that supports all of us who develop sbitx. I request you all to be kinder to criticism of the radio. There is no excuse as a programmer to ship out buggy code and I personally take the entire blame. Perhaps I have been ambitious to pull this off as a single contributor. We must collectively identify and fix the bugs and keep adding features. Our endeavour will be to make the basic radio as stable as we can, and then make it extendible in easy and simple ways. This is a radio that can grow younger with age, that is the magic of free software.

73, and cu on 7015 soon.

de f.

?

On Sat, Oct 28, 2023 at 3:23?AM Gordon Gibby <docvacuumtubes@...> wrote:

Jack, I am with you. I have learned so much— so so much! from this group and from these radios. I am a little bit concerned about spurious signals ?in the manufacturing production line. This is such a fantastic advance that it deserves to be a very clean radio and I’m concerned that we’ve got an issue with the filtering and bleed around. The filters them selves look fantastic!? I understand this had happened to other firms’ radios as well and so it’s not unheard of with new devices. I’m pretty convinced that we will get it fixed and then I’m looking forward to moving on to working on software, I’m especially fascinated with getting back into digital signal processing?

But I’m just swamped with the high school students I’m teaching right now. Today we had them do their first CW contacts with a roundabout technique using a web SDR.? They’re doing around six words per minute. I need them up to 15 by winter field day.?

Gordon

On Oct 27, 2023, at 11:56, John Terrell, N6LN <N6LN@...> wrote:

?

I have been lucky enough to work with my two sBitx rigs (the DE and also the V2) and the uBitx V6 before that. I have learned far more along the way about transmitter and receiver design than I could have any other way. If I want an easy contact I can simply switch on my Icom or Kenwood or Yaesu, but that’s not what I wanted. I have benefited beyond measure from the community of engineers and fellow tinkerers who have helped me. Hopefully I have helped others through my posts as they have walked me through my fits and stops. It has at times been frustrating but all in all I haven’t had this much pleasure in ham radio since I was a teenager going to TV repair shops, taking apart junked tube TV sets and building homebrew rigs. It’s like hiking for days in the mountains. There’s no hot water or wifi, but that’s not the reason to make the trip.?

Enjoy the trip.

73,
Jack, N6LN

--
Jack, W8TEE

I am grateful to the community that supports all of us who develop sbitx.

I am grateful to you and the rest of the sbitx developers! I am enjoying my 'tinkering'! I'm sorry some discover late that the radio may not be what they thought it was. I think you've done a great job explaining what the radio is and what it is not.

I am looking forward to the software update. I like that you have done hamlib net control. I think that will make it even easier for us to tinker. I am willing/able to try the software as early as you want to release it. If you want to reach me, I can be reached via the email address in qrz.com.

-- Regards, Dave, N1AI