So pleased to have you back Charlie and with such a great idea for a new build! Really enjoyed this and looking forward to seeing how it all comes together. 73, Nick M0NTV
Digital vox would be easy. Always receive and digitize on the pc->radio line. When you detect a signal above threshold you detect frequency. Since you're then twiddling the si5351 to send you just buffer the symbol stream while switching to transmit. When it's fully switched you send the symbols. You'll only need a few ms at most so you won't need to buffer much
Hello Charlie! This would be a nice project. 1) Those cheap USB interfaces are stereo output, MONO input IIRC, 2) You could think about using an ESP32. Very cheap and way more features and processing power. Regards!
Thanks. It's should be a fun project. Just need to make the software work. I have to measure audio freq from to 0.01Hz every 0.6S (for WSPR). I hope I can make it work.
Enjoyed this video. I look forward to seeing how your project develops. I think if you only use the I output of the Tayloe detector, it's effectively a direct conversion receiver, and the I output alone will contain frequency components from both sides of the local oscillator frequency. To properly suppress the lower sideband and leave only the upper sideband (for data modes) it will be necessary to further process the I and Q signals. For example, you could use the Phasing Method, with a Hilbert Transform (-90° shift) of the Q output summed with the I component. Good luck with the project.
Hey your back!!!! Glad to see a video from you. Now make sure you use one of those neat food containers. The best so far was the tuner fs meter one. But there all great.
Welcome back Charlie. Re filtering for the two bands 40 and 20m, you could use separate relay switched LPFs for those bands, route the Rx signal thru one of these, then into a Fixed HPF with cutoff below 7MHz. The combination is effectively a BPF on 40 and a wide (7MHz) BPF on 20. So only one pair of filters to switch. I've tried this and it worked FB. Of course it depends on the local environment. If you live near a big transmitter on 8 to 13MHz, you might think again. Most of the energy coming down my coax is MW BC stations so it worked for me. 🤓
That's a great idea Paul. I've jotted that down. I think two BPFs should work, as we're only interested in a tiny portion of each band. Getting the BPF passband to cover that section should be easy. I'll have to double check, but I'm sure Codan did that with their 7727 at least. Both RX and TX went through it.
@@CharlieMorrisZL2CTM Oh No, that was a lot of time into it I bet. This project is very interesting. I thought of doing something similar with voice also and run the whole thing from the PC. I've been so busy with work and things around the house I've been to exhausted to get to my hobbies. but winter is around the corner and should free some time up. Good day, Charlie. I love listening to your videos.
The computer has stereo out as well. Software allows you to send data on one channel and a constant tone on the other. Use the tone to activate the VOX. This eliminates drop out caused by data interruptions. And I think hang time is adjustable in software as well.
@@CharlieMorrisZL2CTM Not sure about all. Have only used WSJT. It’s been awhile and don’t remember the menu setting but it’s in there. Let’s say data is on right ch and tone on left. The left ch can be amplified separate and gain controlled for VOX sensitivity. To make it software independent you could put a jumper between L and R channels, making it mono.
Hi Charlie ! Great project. I have downloaded the code from your blog and am now breadboarding the transmitter side of things. When do you think you will have the receiver side in shape for testing ?
Nice to have you back with another video. I take it your looking at the fst3253 for the mux. I'm sure there's a library out there that will read the frequency of an input pin. Sounds like an interesting little project, can't wait to see how it goes. Brian. M1CEM
Thanks Brian. Yes it was the 3253. I hope I can still buy them. The software will be interesting as I need to (from what I can see) measure the audio frequency down to 0.01Hz every 0.6S for WSPR at least. JS8 and FT8 are more stringent on timing (0.19S). I'll have another play today.
Those timing tolerances might be a bit much for a nano, well it certainly would be fo my programming skills. Maybe look at the stm32 or esp32. The esp32 is fast enough to do some quite reasonable DSP.
Your description of the Tayloe detector, particularly your description of the quadrature LO inputs, doesn't sound right. It sounds like you're saying that one of the LOs is just an inverted version of the other, which may appear to work, but this isn't quadrature. Quadrature is a 90 degree phase difference between the two, and I think you can get the SI5351 to produce outputs that are 90 degrees out of phase, but it's important to make this distinction, or your detector won't work well. That, or I misunderstood what you were saying, in which case, never mind!
And now that I've watched the rest of the video, if you're not using the Q output, then it doesn't matter that your LO inputs to the Tayloe detector are in quadrature, anyway. In fact, if you're only using the I output, you don't even need a quadrature input to the mux. If you look at the multiplexer in a Tayloe detector, one of those LO inputs feeds the mux that produces the I output, and the other feeds the mux that produces the Q output, so it doesn't do anything if you're not using the Q output, and you can leave the second mux out as well. And honestly, the main reason to use a Tayloe detector is to reject the unwanted sideband if you're using direct conversion, i.e., zero IF, since you can't use an IF filter for that. But since you're using a superheterodyne configuration, it should be perfectly fine to use a single LO and mux, producing only the I signal to your audio amp. You're basically using a mux as a balanced mixer.
That's good to know. I found the old code which dates back to a 2018 SDR radio. In that case i did need the I and Q, but as you quite rightly point out, I'll only need one channel in this application. That will simply things. Thanks again.
