Thanks Ashhar,?

Your right, one should face his fears ?, I'll give it a try (once I am back home).? I'll do it on both a RasPi-400 and a RasPI-Zero, I'll keep you posted when I succeed in building it from your sBITX repository.

According to my simplistic "guestimations" the SDR / DSP part of the sBITX software would need some 2 MFLOPS plus some for the sound driver.? According to what I can find here and there on the internet, the RasPi-Zero is capable of 350 MFLOPS, which should be adequate (the RasPi-4 does 13500 MFLOPS !).?
Beside the clock speed it is really the floating point coprocessor that does it, the humble RasPi-pico does barely 50 kFLOPS. ?

73 ON4JRT / Jean.

Op 5/07/2021 om 05:49 schreef Ashhar Farhan:

Jean,

If you can do Arduino, you can do RPI. I have extensively used arduino api (lile digitalRead(), etc. It is simple C. Check the repository.

On Mon 5 Jul, 2021, 2:26 AM jeanrenier2004, <jean.taeymans@...> wrote:

Many thanks Ashhar for the brilliant work so far, it is by far the most exciting hackable SDR (or at least partially SRD) project so far.
Take all the time you need, no pressure, I'll be patient ?. ?
If I can be of some assistance, let me know:? I have experience in hardware design (mostly with micro controllers), including PCB design (as it happens with the Si5351 and the WM8731) and in software design, mostly Arduino (AVR, ESP, some RP2040, I wish I was more fluent in Raspberry Pi).?
73 ON4JRT / Jean.

Op 4/07/2021 om 15:51 schreef Ashhar Farhan:

I have been hacking away at the small stuff for sbitx. There were ground loops between the digital board and the rf board. Then, there were oscillations in the PA. Finally, there was change to 40 Mhz filter, changing IF from 27 MHz to 40 MHz.?

I am working on the T/R switching with a reed relay, hopefully getting to full QSK. I hope to work one contest before posting it all.?

I am holding back because these are interrelatef issues and resolving each needs rework elsewhere. I want to be able to wrap it up asap. I am experiencing design fatigue. This should finish with a creative 'yay!' rather than an exhausted 'whew!'

On Sat 3 Jul, 2021, 9:03 PM Peter McCracken, <peter.mccracken@...> wrote:

Hi Jean,

In answer to one of your questions Ashar went over a couple of things in his video, one was the concept of a SDR contronolled uBitx which was basically bolting the Pi onto the front end of a uBitx and feeding CAT commands to the existing Raduino?over serial, he then moved on to to talk about the what was actually the sbitx which is substantially different from the existing uBitx and where the Raduino is scrapped and the transceiver is controlled directly by the RPi. All the diagrams and info are included in the docs/presentation that was shared above along with the video. The docs don't show the schematic for the sBitx hardware but you can screenshot them from the video. It's a shame this thread has went a bit quiet because this is quite exciting, tons for potential for experimenting with this.?

On Sat, 3 Jul 2021 at 15:36, jeanrenier2004 <jean.taeymans@...> wrote:

Greetings to all,

More or less the same question here, plus also a few remarks and questions for Asshar I suppose.
I sincerely hope that all is well in Hyderabad and that Asshar is able the pursue his work on the sBITX after his brilliant presentation last spring.
Many thanks to him for sharing this with us all.

Over the past years I experimented myself a bit with the quadrature direct conversion architecture, using various micro controllers: STM32F446, ESP32 and recently RP2040 in a set up with a WM8731 (running at 8ks/s), a FST3253 and a Si570.? Though all of this more or less worked, I never succeeded in having an adequate side band suppression (typically not more than 30dB).? This is of course related to the rather challenging hardware task of having a very precise 90° relation between the I and Q signals (of both the VFO and the audio) and having precisely the same gain on both I and Q amplifiers.

Asshar's double conversion architecture with one sharp IF filter at 27/40MHz in hardware and a second at 24kHz implemented in software is i.m.h.o. absolutely brilliant, as it circumvents all the above problems.? Also is the idea of performing filtering, demodulation, etc. in the frequency domain.
This comes at a price however, running the codec at 96ks/s and performing a 1024 bins floating point FFT every 10ms excludes the low end micro controllers, even the RP2040 (no FP coprocessor).? Hence the use of the Raspberry Pi as a computational platform.??

A few questions if I may.?
- Would the sBITX software run in a headless configuration (only CAT, no GUI) on a Raspberry Pi Zero ? (and obviously without the WSJT-X application).? I am particularly interested in using the sBITX in a remote battery operated (= low power) TRX set up.?
- The hardware setup diagram shows a Raduino board connected to the Raspberry Pi by USB.? Can the Si5351 not be controlled directly through the I2C bus of the Raspberry Pi ?? It is otherwise used for the WM8731.?
- The repository on is a bit confusing:? two libraries for bit banging the I2C bus, tough WiringPi includes the I2C functionality.? A library for the Si570 seems a bit lost in this context.?
- The WM8731 codec seems to be setup (sample rate, gain settings, etc.) by its driver build in Raspbian driver, which is fine I guess.? Can it be controlled (e.g. gain, microphone input, etc.) by alsa config ?
- A more elaborate guide to the installation of the various libraries and the building of the software would be greatly appreciated.? Is e.g. any flavor of Raspbian ok ??

73 ON4JRT / Jean.

Op 3/07/2021 om 00:15 schreef Peter McCracken:

Hi Niels,

Did you ever make any progress with this or get a board up and running ?

Also did you source some 27Mhz crystals, if so which ones did you go for ?

Thanks!