Whoops I didn't open Rene's attachment til just now and he's got things pretty well covered there.? ?Nice.
|
I've struggled with this exact problem and my advice would be to thoroughly?investigate the COM port settings and leave the uBITX as is.? ?Even now, when I have to restart the computer I sometimes have to restart WSJT-X a few times or disconnect/connect? the USB cable to get things to synch up.? ?Get your device manager open and display a window with the com port settings beside a window of the Settings/Radio screen in WSJT and check for any dissimilarities. For reference, my WIn10 just got a lengthy update (which completely knocked out my Bluetooth - ongoing) and here's my working driver info;  Scott
|
Re: Farhan to give sBITX talk at FDIM
Ashhar,
is your talk available online?
73
Dave
|
Re: Farhan to give sBITX talk at FDIM
As you can see in this (bad) photo:
we have some scaffolding code that monitors the temperature and load of the Teensy 4.1. This is pretty old code (you can see it's V072 and we're currently working on V107). As I recall, the T4.1 was running at 500mHz when this shot was taken and before we put a passive heat sink on the T4.1. When the PA's get added, there will be some forced air running over the system, too.
Jack, W8TEE
On Monday, May 24, 2021, 10:29:36 AM EDT, Shirley Dulcey KE1L <mark@...> wrote:
I'd guess that the sBITX isn't getting anywhere near 100% CPU or GPU utilization. Cooling the SoC on the board is only an issue if you do.
On Mon, May 24, 2021 at 10:20 AM Ashhar Farhan < farhanbox@...> wrote: I run the sbitx all the time. It keeps scrolling the wsjtx callsigns. What does get hot is the 2 dollar switching regulator. It doesn't have a heatsink. I am planning to build a proper one with a standard heatsink
On Mon 24 May, 2021, 7:45 PM Shirley Dulcey KE1L, < mark@...> wrote: He's using a Raspberry Pi 4. That's four Cortex-A72 cores clocked at 1.5 GHz maximum out of the box, and it can be overclocked to 2 GHz if need be. Plenty of processing power, and the multiple cores mean that the user interface won't become unresponsive because of the CPU demands of the DSP. That's one flaw that the popular inexpensive DMR HTs all share; the SDR processing, the digital voice codec, and the user interface are all done by an STM32F405 processor, and limitations like their slow band scanning are caused by the heavy demands on that CPU.
Sustaining 1.5 GHz, let alone 2 GHz, for extended periods of time on multiple cores requires the addition of some way to cool the CPU; either a heat sink plus fan or a LARGE heat sink that is essentially the entire case for the board.
Hi Trystan:
My problem was that at the last moment I change from my house to Al's for the presentation. My office is in the basement and the extender I'm using started causing the Internet connection to drop out. Since that would be a disaster during my talk, Al let me set things up at his house. Things got a little confusing, but we muddled through.
I really enjoyed Ashar's talk as we are also working on an SDT. A couple of years ago, Al and I did a DSP board that connected to the uBITX, but it never really caught on. I think we made it too complex to add it to the uBITX. The new SDT will have mutlple boards with lots of "board real estate" and test points for experimenting. We will also do a book where each chapter will discuss a section of the rig, from antenna to earphones. Another thing we learned from the JackAl experience was NOT to use a resistive touch screen. Most QRP rigs do not have enough mass to activate a button without using the other hand to stabilize the rig. Ashar's rig uses capacitive touch...smart move! Also, he is using a large display where us "fat-finger people" can actually poke an individual button! And the RP has a mega-munch of memory and is clocked at 750mHz (I think?). Anyway, it was interesting to see what he's done with the uBITX in the SDT sandbox. I also admired the "cleanliness" of his code...good stuff!
Jack, W8TEE
Jack, I really felt for you when you couldn't find your slides, then after you found them you couldn't share your screen. Then the summary table was all blacked out. Argh! Your talk was very good, thank you! Unfortunately I missed Asher's talk as it was on too late for me, but plan to see it when they release the recording. Trystan: G0KAY
--
Jack, W8TEE
--
Jack, W8TEE
|
Re: Farhan to give sBITX talk at FDIM
It was a great day.? I was only able to catch the beginning of Jack's talk but I will catch the rest when it gets uploaded in few weeks.? In typical Jack style it was more an interesting conversation then just a lecture.? I did see all of Farhan's presentation and really enjoyed it too.? I appreciated him staying up until the middle of the night in Hyderabad to accommodate us.
--
|
Re: Farhan to give sBITX talk at FDIM
SMD or through hole regular ?
toggle quoted message
Show quoted text
At 24/05/2021, you wrote: I run the sbitx all the time. It keeps scrolling the wsjtx callsigns. What does get hot is the 2 dollar switching regulator. It doesn't have a heatsink. I am planning to build a proper one with a standard heatsink
|
Re: Farhan to give sBITX talk at FDIM
Hi Shirley:
I haven't looked at the RP or its specs for a long time, and did not know it was running at that speed. I did notice during his talk that Farhan is using C for development, but there's a lot of Python code for the RP. Do you know if the C libraries automatically do threading for the UI?
Jack, W8TEE
On Monday, May 24, 2021, 10:15:55 AM EDT, Shirley Dulcey KE1L <mark@...> wrote:
He's using a Raspberry Pi 4. That's four Cortex-A72 cores clocked at 1.5 GHz maximum out of the box, and it can be overclocked to 2 GHz if need be. Plenty of processing power, and the multiple cores mean that the user interface won't become unresponsive because of the CPU demands of the DSP. That's one flaw that the popular inexpensive DMR HTs all share; the SDR processing, the digital voice codec, and the user interface are all done by an STM32F405 processor, and limitations like their slow band scanning are caused by the heavy demands on that CPU.
Sustaining 1.5 GHz, let alone 2 GHz, for extended periods of time on multiple cores requires the addition of some way to cool the CPU; either a heat sink plus fan or a LARGE heat sink that is essentially the entire case for the board.
Hi Trystan:
My problem was that at the last moment I change from my house to Al's for the presentation. My office is in the basement and the extender I'm using started causing the Internet connection to drop out. Since that would be a disaster during my talk, Al let me set things up at his house. Things got a little confusing, but we muddled through.
I really enjoyed Ashar's talk as we are also working on an SDT. A couple of years ago, Al and I did a DSP board that connected to the uBITX, but it never really caught on. I think we made it too complex to add it to the uBITX. The new SDT will have mutlple boards with lots of "board real estate" and test points for experimenting. We will also do a book where each chapter will discuss a section of the rig, from antenna to earphones. Another thing we learned from the JackAl experience was NOT to use a resistive touch screen. Most QRP rigs do not have enough mass to activate a button without using the other hand to stabilize the rig. Ashar's rig uses capacitive touch...smart move! Also, he is using a large display where us "fat-finger people" can actually poke an individual button! And the RP has a mega-munch of memory and is clocked at 750mHz (I think?). Anyway, it was interesting to see what he's done with the uBITX in the SDT sandbox. I also admired the "cleanliness" of his code...good stuff!
Jack, W8TEE
Jack, I really felt for you when you couldn't find your slides, then after you found them you couldn't share your screen. Then the summary table was all blacked out. Argh! Your talk was very good, thank you! Unfortunately I missed Asher's talk as it was on too late for me, but plan to see it when they release the recording. Trystan: G0KAY
--
Jack, W8TEE
--
Jack, W8TEE
|
Re: Farhan to give sBITX talk at FDIM
I'd guess that the sBITX isn't getting anywhere near 100% CPU or GPU utilization. Cooling the SoC on the board is only an issue if you do.
On Mon, May 24, 2021 at 10:20 AM Ashhar Farhan < farhanbox@...> wrote: I run the sbitx all the time. It keeps scrolling the wsjtx callsigns. What does get hot is the 2 dollar switching regulator. It doesn't have a heatsink. I am planning to build a proper one with a standard heatsink
On Mon 24 May, 2021, 7:45 PM Shirley Dulcey KE1L, < mark@...> wrote: He's using a Raspberry Pi 4. That's four Cortex-A72 cores clocked at 1.5 GHz maximum out of the box, and it can be overclocked to 2 GHz if need be. Plenty of processing power, and the multiple cores mean that the user interface won't become unresponsive because of the CPU demands of the DSP. That's one flaw that the popular inexpensive DMR HTs all share; the SDR processing, the digital voice codec, and the user interface are all done by an STM32F405 processor, and limitations like their slow band scanning are caused by the heavy demands on that CPU.
Sustaining 1.5 GHz, let alone 2 GHz, for extended periods of time on multiple cores requires the addition of some way to cool the CPU; either a heat sink plus fan or a LARGE heat sink that is essentially the entire case for the board.
Hi Trystan:
My problem was that at the last moment I change from my house to Al's for the presentation. My office is in the basement and the extender I'm using started causing the Internet connection to drop out. Since that would be a disaster during my talk, Al let me set things up at his house. Things got a little confusing, but we muddled through.
I really enjoyed Ashar's talk as we are also working on an SDT. A couple of years ago, Al and I did a DSP board that connected to the uBITX, but it never really caught on. I think we made it too complex to add it to the uBITX. The new SDT will have mutlple boards with lots of "board real estate" and test points for experimenting. We will also do a book where each chapter will discuss a section of the rig, from antenna to earphones. Another thing we learned from the JackAl experience was NOT to use a resistive touch screen. Most QRP rigs do not have enough mass to activate a button without using the other hand to stabilize the rig. Ashar's rig uses capacitive touch...smart move! Also, he is using a large display where us "fat-finger people" can actually poke an individual button! And the RP has a mega-munch of memory and is clocked at 750mHz (I think?). Anyway, it was interesting to see what he's done with the uBITX in the SDT sandbox. I also admired the "cleanliness" of his code...good stuff!
Jack, W8TEE
Jack, I really felt for you when you couldn't find your slides, then after you found them you couldn't share your screen. Then the summary table was all blacked out. Argh! Your talk was very good, thank you! Unfortunately I missed Asher's talk as it was on too late for me, but plan to see it when they release the recording. Trystan: G0KAY
--
Jack, W8TEE
|
Re: Farhan to give sBITX talk at FDIM
I run the sbitx all the time. It keeps scrolling the wsjtx callsigns. What does get hot is the 2 dollar switching regulator. It doesn't have a heatsink. I am planning to build a proper one with a standard heatsink
On Mon 24 May, 2021, 7:45 PM Shirley Dulcey KE1L, < mark@...> wrote: He's using a Raspberry Pi 4. That's four Cortex-A72 cores clocked at 1.5 GHz maximum out of the box, and it can be overclocked to 2 GHz if need be. Plenty of processing power, and the multiple cores mean that the user interface won't become unresponsive because of the CPU demands of the DSP. That's one flaw that the popular inexpensive DMR HTs all share; the SDR processing, the digital voice codec, and the user interface are all done by an STM32F405 processor, and limitations like their slow band scanning are caused by the heavy demands on that CPU.
Sustaining 1.5 GHz, let alone 2 GHz, for extended periods of time on multiple cores requires the addition of some way to cool the CPU; either a heat sink plus fan or a LARGE heat sink that is essentially the entire case for the board.
Hi Trystan:
My problem was that at the last moment I change from my house to Al's for the presentation. My office is in the basement and the extender I'm using started causing the Internet connection to drop out. Since that would be a disaster during my talk, Al let me set things up at his house. Things got a little confusing, but we muddled through.
I really enjoyed Ashar's talk as we are also working on an SDT. A couple of years ago, Al and I did a DSP board that connected to the uBITX, but it never really caught on. I think we made it too complex to add it to the uBITX. The new SDT will have mutlple boards with lots of "board real estate" and test points for experimenting. We will also do a book where each chapter will discuss a section of the rig, from antenna to earphones. Another thing we learned from the JackAl experience was NOT to use a resistive touch screen. Most QRP rigs do not have enough mass to activate a button without using the other hand to stabilize the rig. Ashar's rig uses capacitive touch...smart move! Also, he is using a large display where us "fat-finger people" can actually poke an individual button! And the RP has a mega-munch of memory and is clocked at 750mHz (I think?). Anyway, it was interesting to see what he's done with the uBITX in the SDT sandbox. I also admired the "cleanliness" of his code...good stuff!
Jack, W8TEE
Jack, I really felt for you when you couldn't find your slides, then after you found them you couldn't share your screen. Then the summary table was all blacked out. Argh! Your talk was very good, thank you! Unfortunately I missed Asher's talk as it was on too late for me, but plan to see it when they release the recording. Trystan: G0KAY
--
Jack, W8TEE
|
Re: Farhan to give sBITX talk at FDIM
He's using a Raspberry Pi 4. That's four Cortex-A72 cores clocked at 1.5 GHz maximum out of the box, and it can be overclocked to 2 GHz if need be. Plenty of processing power, and the multiple cores mean that the user interface won't become unresponsive because of the CPU demands of the DSP. That's one flaw that the popular inexpensive DMR HTs all share; the SDR processing, the digital voice codec, and the user interface are all done by an STM32F405 processor, and limitations like their slow band scanning are caused by the heavy demands on that CPU.
Sustaining 1.5 GHz, let alone 2 GHz, for extended periods of time on multiple cores requires the addition of some way to cool the CPU; either a heat sink plus fan or a LARGE heat sink that is essentially the entire case for the board.
Hi Trystan:
My problem was that at the last moment I change from my house to Al's for the presentation. My office is in the basement and the extender I'm using started causing the Internet connection to drop out. Since that would be a disaster during my talk, Al let me set things up at his house. Things got a little confusing, but we muddled through.
I really enjoyed Ashar's talk as we are also working on an SDT. A couple of years ago, Al and I did a DSP board that connected to the uBITX, but it never really caught on. I think we made it too complex to add it to the uBITX. The new SDT will have mutlple boards with lots of "board real estate" and test points for experimenting. We will also do a book where each chapter will discuss a section of the rig, from antenna to earphones. Another thing we learned from the JackAl experience was NOT to use a resistive touch screen. Most QRP rigs do not have enough mass to activate a button without using the other hand to stabilize the rig. Ashar's rig uses capacitive touch...smart move! Also, he is using a large display where us "fat-finger people" can actually poke an individual button! And the RP has a mega-munch of memory and is clocked at 750mHz (I think?). Anyway, it was interesting to see what he's done with the uBITX in the SDT sandbox. I also admired the "cleanliness" of his code...good stuff!
Jack, W8TEE
Jack, I really felt for you when you couldn't find your slides, then after you found them you couldn't share your screen. Then the summary table was all blacked out. Argh! Your talk was very good, thank you! Unfortunately I missed Asher's talk as it was on too late for me, but plan to see it when they release the recording. Trystan: G0KAY
--
Jack, W8TEE
|
Re: Farhan to give sBITX talk at FDIM
My understanding is that all of the presentations will be posted within the next couple of weeks.
Jack, W8TEE
On Monday, May 24, 2021, 9:59:23 AM EDT, Rafael Pinto [PU1OWL] <rafaelgcpp@...> wrote:
hi,
Does anyone have a recording of this talk?
I've been crazy busy lately and just found about it today.
Regards
Rafael, PU1OWL
-- Jack, W8TEE
|
Re: Farhan to give sBITX talk at FDIM
hi,
Does anyone have a recording of this talk?
I've been crazy busy lately and just found about it today.
Regards
Rafael, PU1OWL
|
Re: Farhan to give sBITX talk at FDIM
It was the closest that I have ever gotten to FDIM. I was able to watch all of the forums and can't say that there weren't any that I didn't enjoy. I learned a lot of new stuff and got a bunch of ideas that I plan on exploring.? -- '72
Aaron K5ATG
|
Re: Farhan to give sBITX talk at FDIM
Hi Trystan:
My problem was that at the last moment I change from my house to Al's for the presentation. My office is in the basement and the extender I'm using started causing the Internet connection to drop out. Since that would be a disaster during my talk, Al let me set things up at his house. Things got a little confusing, but we muddled through.
I really enjoyed Ashar's talk as we are also working on an SDT. A couple of years ago, Al and I did a DSP board that connected to the uBITX, but it never really caught on. I think we made it too complex to add it to the uBITX. The new SDT will have mutlple boards with lots of "board real estate" and test points for experimenting. We will also do a book where each chapter will discuss a section of the rig, from antenna to earphones. Another thing we learned from the JackAl experience was NOT to use a resistive touch screen. Most QRP rigs do not have enough mass to activate a button without using the other hand to stabilize the rig. Ashar's rig uses capacitive touch...smart move! Also, he is using a large display where us "fat-finger people" can actually poke an individual button! And the RP has a mega-munch of memory and is clocked at 750mHz (I think?). Anyway, it was interesting to see what he's done with the uBITX in the SDT sandbox. I also admired the "cleanliness" of his code...good stuff!
Jack, W8TEE
On Monday, May 24, 2021, 3:54:59 AM EDT, Trystan G0KAY <trystandavies+nodirect@...> wrote:
Jack, I really felt for you when you couldn't find your slides, then after you found them you couldn't share your screen. Then the summary table was all blacked out. Argh! Your talk was very good, thank you! Unfortunately I missed Asher's talk as it was on too late for me, but plan to see it when they release the recording. Trystan: G0KAY
-- Jack, W8TEE
|
Re: Suggestions for a switching regulator
Farhan,
Ultimately I'd go to some of the big HF rigs service manuals and check how they internally regulate the voltages.
The IC7300, for example, uses a TPS62110 from Texas Instruments, but that is a 1.5A part. The FT991A uses some 7809, 7805 and linear regulators.
Regards,
Rafael
|
Hello >Jerry, Thank you for your informationI also found a link in French that explains how to sort Quartz with spreadsheet.On your link, I also found this on one answer of the post:? ?The villa Grand Piton
Yeah BIG warning on this scheme.? Elecraft uses this in the K3¡ for the ¡°ON¡± switch by controlling the gate.? I had mine SHORT in the CLOSED/ON position.? Doesn¡¯t do much good when the diode shorts¡
?
?
Dr.?William J. Schmidt - K9HZ J68HZ 8P6HK ZF2HZ PJ4/K9HZ VP5/K9HZ PJ2/K9HZ
?
Owner - Operator
Big Signal Ranch ¨C K9ZC
Staunton, Illinois
?
Owner ¨C Operator
Villa Grand Piton ¨C J68HZ
Soufriere, St. Lucia W.I.
Rent it:?
Like us on Facebook!?
WHAOOWWW? if i beat the lottery, I'm willing to go one month.? LOLcdt
|
Re: Suggestions for a switching regulator
A number of years ago I worked for a Microwave module company. We used to use a spread spectrum technique for our switchmode local supplies. We also used to put DC modules in a shielded box and used thunderline filter feedthroughs to route
DC power in and out.
?
Good luck in the path you chose.
?
toggle quoted message
Show quoted text
From: [email protected] < [email protected]> On Behalf Of
Ashhar Farhan via groups.io
Sent: Monday, May 24, 2021 1:22 AM
To: [email protected]
Subject: Re: [BITX20] Suggestions for a switching regulator
?
It was suggested to use a lower switching frequency so that the harmonics in the HF are weaker. I will look at some 50 KHz parts. The transients need to b snubbed on the output and the input needs to be cleaned. I am wondering what kind
of inductors will help. I will try with the FT37-43 (use what I have) and see if that works.
?
On Mon, May 24, 2021 at 10:44 AM Ashhar Farhan via <farhanbox=[email protected]> wrote:
Jerry, Terry,
All good suggestions. Let me play with these.
?
I've built a lot of boards with switchers, but never
anything with radio gear on it.? It is done all the time though,
successfully.? I'm sure any cell phone has a few.
Here's an ap note from TI on this:
If you are building your own switcher, buy a chip from a mainstream?
manufacturer.? Being in the US, I would tend to look first at TI and Analog Devices,
though there certainly are other good sources.
A well documented part will have suggestions on board layout and
the proper choice of capacitors and inductors.? Read it carefully.
You might try corresponding with the manufacturer, asking for advice.
Your volumes will be relatively low, but the application engineers?
generally are happy to help.
For a radio, you want a switcher that operates at a constant frequency,
and at a frequency of your choosing.? A pin into the part can be supplied
with a clock of perhaps 500khz, telling the part when to switch.
The word "synchronous" in the world of switcher design means something?
else, and is not a requirement in your case.? It means that there are two FET's in your
buck mode switcher feeding the inductor.? Sometimes the high side FET
is on, connecting the power source to the inductor.? Other times the bottom side
FET is turned on, sourcing current from ground into the inductor.? In a "non-synchronous"?
switcher, that low side FET is typically just a schottky diode.? The diode is slightly
less efficient because it has a roughly 0.3v forward voltage drop, but that won't
make much difference for a 5v supply,
You definitely want a complete ground plane under it, follow the docuementation
on part placement and how to stitch up the ground plane.? Be aware of how
the switched currents are flowing.
You may want that ground plane to be on a different board than the radio.
Keep things physically small and the layout tight to reduce radiation.
For a buck mode switcher, it is switching the current coming in.
The current going out is fairly smooth if you have s sufficiently large inductor.
So pay special attention to selecting the capacitors at the input to the switcher,
and make sure they are placed very close to the FET switch.
Jerry, KE7ER
|
Also test if the uBITX works 'stand alone' without the USB cable plugged in.
Is this a situation of everything was untouched and just not powered, or 'picked up the bench' and detached cables?
Scenario: You don't have the uBITX powered through the 12V.? 5V backfeeds? ~3.6V to the uBITX main board and is enough to power RX only.
& (You are using a power only USB cable | you have a bad USB cable connection |? Windows update possibly hosed you | using a different computer)
Rgds,
Gary
|
Re: Suggestions for a switching regulator
LM2576S-5.0 does 3A, robust and well proven....52 kHz Fixed Frequency Internal Oscillator
|
Re: Farhan to give sBITX talk at FDIM
Jack,
I really felt for you when you couldn't find your slides, then after you found them you couldn't share your screen. Then the summary table was all blacked out. Argh! Your talk was very good, thank you!
Unfortunately I missed Asher's talk as it was on too late for me, but plan to see it when they release the recording.
Trystan G0KAY
|
_Dave_ AD0B
Raj vu2zap
Shirley Dulcey KE1L
Rafael Pinto [PU1OWL]
Aaron K5ATG
Gerard
Trystan G0KAY