This code generates a .wav file which emulates a fully-populated FT8 band! Usage This .wav file can then be played with rpitx on actual air (well on a dummy load) for receiver testing purposes. rm mixed.wav; sox -m *.wav mixed.wav sox mixed.wav -r 48000 sampleaudio.wav Now play sampleaudio.wav using rpitx. Testing I have tested rpitx on Raspberry Pi Zero 2W TX'ing @ 14.074 MHz. ...
Instead of punching islands or using commercial pads (islands) or making traditional pads myself, I simply make use of small pieces of single-sided PCBs to create the pads / islands easily for my boards. The Technique This technique has a huge advantage over the punched islands or scored and divided boards techniques as it keeps the bottom ground plane perfectly 100% intact! Also, it is almost a tool-free process to create the pads by breaking or cutting the single-sided PCBs. Single-sided prototyping boards also happen to be very reasonably priced. ...
Here is the FT8 WAV test file. On an AMD64 machine: 1 second using the fast decode settings. We were able to get the MSHV's FT8 engine to work on Android, thanks to the work done by https://github.com/sannysanoff/SDRPlusPlusBrown folks. Around ~4.5 seconds to decode the same file. This is why we can't really have full-blown realtime FT8 decoder(s) on Android just yet! Idea: But we can surely use this full-blown FT8 decoder in our FT8 Decoder app which has no realtime requirements! ...
OCXO's are awesome - this is what I learned while trying to get WSPR working reliably on the 2m band. Hardware Here is all you need for reliable WSPR'ing on the 2m band: A RAW OCXO "can" works fine too! Note: This modified Si5351 board image comes from SimonsDialogs. Essentially, we modify the CLK2 output port to become an input port that accepts the OCXO's output signal. ...
Our WSPR beacon design works fine even on the 4m band (~70 MHz) . The key is to use a 25 MHz high-quality TCXO, like the 25 MHz HCI 0.5ppm TCXO! Results Challenges 2m WSPR is still out of reach for now - see the following sporadic decodes. Of course, the next challenge is to get 2m WSPR beacons working 100% reliably at a low cost. ...
The usage of ready-made coupled inductors in RF designs was pioneered by Jim Veatch in his award-winning RF amplifier designs. Applications For Baby-QRO applications, we can use Wurth 744851330 coupled inductor. For 5W / 10W RF amplifier applications, we can use Bourns SRF1260A-4R7Y coupled inductor. Recently we were able to utilize very cost-effective YJYCOIN YPRH1207C-4R7M Chinese coupled inductor with excellent results. With our current Digital Amplifier designs, this Chinese coupled inductor works even at ~75 MHz. ...
Our 6m WSPR beacon design works fine now. The key is to use a 25 MHz high-quality TCXO, like the 25 MHz HCI 0.5ppm TCXO! 2m WSPR is still out of reach for this VFO design - see the following drifty screenshots: While WSPR is no good, 2m FT8 works pretty fine! ...
Tokmas CID9N65E3 rocks! Performance It is CRAZY how efficient this GaN FET is in TO-252 (DPAK) package. It generates 5W @ 28 MHz with 13.8V at drain without even warming up! Close to 4W @ 50 MHz with 12.0V at drain. 5W+ @ 50 MHz with 13.8V at drain. 11W+ @ 50 MHz with 20V at drain. It even produces ~3W at 70 MHz with 13V at drain - w00t! ...
I recently found some FET gate drivers which are quite fast! The Drivers Wuxi Maxinmicro MX1025D - Favorite! BD2311NVX-LBE2 from ROHM Semiconductor TI LMG1020 Tokmas LMG1020YFFR Usage They are perfect for driving the fast Tokmas GaN FETs that we recently discussed on this site. Our 'workhorse' gate driver for the recent experiments has been the Onsemi's NCP81074ADR2G in a friendly SOIC-8 package. And we have shifted to MX1025D gate driver in our latest amplifier designs. ...
This is an early design sketch of a 2m WSPR beacon. Key Points 25 MHz HCI 0.5ppm TCXO powering the Si5351 module instead of the 26 MHz TCXO. If this fails, we can use a 10 MHz OCXO instead (a bit out of spec but works fine). Tokmas CID10N65F GaN FET might work @ 144 MHz as the 'final'. If not, use Mitsubishi RD15HVF1-501 MOSFET as backup. 2SK3475 as the driver with variable DC bias at the gate The whole VCC to this amplifier will be PTT switched Ideally, we would want to keep the driver stage linear but the final stage as switched (Class C/D) Alternate: Perhaps by DC coupling the Si5351 input to 2SK3475, we can avoid the need for DC gate bias The "RFC" (bifilar coil still needed?) part will need to be figured out for VHF! Status Results: To be built and tested soon! ...