How to compare radios

Foundations of Amateur Radio One of the topics I've been talking about lately is the idea that we might be able to measure the performance of your radio in some meaningful way using equipment that can be either obtained by any amateur, or by introducing a process that allows results to be compared, even if they have been generated differently. Recently I came up with a tool that automatically generates a spectrogram of an audio recording. That on its own isn't particularly interesting, but it's step one in the processing of an audio signal. In addition to the spectrogram, I also created a tool that generates a tone frequency sweep, think of it as a tone that changes frequency over time, let's call it a sweep. If you combine the two, you can generate a spectrogram of the sweep to give you a starting point or baseline for comparison. You can build on that by using your radio to transmit that sweep and record the result using a receiver. In my initial experiments, I used an RTLSDR dongle to receive the audio with some success and a boatload of spectacular harmonics, but I wanted to find a better, more accessible way to do this and during the week I realised that my Yaesu FT-857d that's sitting in my shack, is connected to a perfectly functional antenna and with a few settings it could do the job perfectly. One of the biggest issues with my RTLSDR setup was squelch. That is the difference between what is a legitimate transmission and what is noise. Set it too high and you hear nothing, set it too low and you hear everything, including background noise. Since the VHF or 2m noise levels are quite high at my location, or QTH, I normally have the squelch completely closed. This is fine if you're normally using a strong repeater, but if you're attempting to receive a weak hand-held, that's never going to work. As any self-respecting amateur I was dragged down the path of last resort to read my user manual where I discovered that in addition to CTCSS, a way to transmit a tone to open a repeater, there's also a setting called Tone Squelch or on my radio TSQ, which will keep my radio squelch closed, unless it hears the CTCSS tone from another radio. Truth be told, I had to read a different user manual to discover how to actually set the CTCSS tone on my handheld to test, but that's just adding insult to injury. It has been a while since I read any manual, even though I try to get to it once a year or so. I blame it on the lack of field-day camping. That's my story and I'm sticking to it. So, combining all this, the spectrogram generator, the sweep, CTCSS, and adding a Raspberry Pi with some website magic, if you're interested, an AWS S3 bucket, I now have a service that listens on a local frequency, opens the squelch if it hears the correct CTCSS tone, records the incoming signal until it stops, then generates a spectrogram from that audio and uploads it to a web site. None of this is particularly complicated, though I did have some bugs to work through. I've published the code as a branch to my existing frequency-response project on github and I've asked my local community to experiment with what I have on-air before I start doing more far reaching experiments. For example. If I were to tune my radio to a local repeater output frequency, rather than the simplex one I'm currently on, I'd be able to record and generate spectrograms for each transmission coming from that repeater. If that repeater was connected to the internet, using AllStar, IRLP, Echolink, DMR or Brandmeister, or even all of them, the global community could send their audio to my recorder and it could generate a spectrogram on the spot. If using that repeater, you played a sweep into your microphone, or used your digital audio interface to play the sound, you could then compare your signal path against others and against the baseline response. One of the issues with doing this is that much of the audio that travels across the internet is pretty munched, that is, it's compressed, frequencies are cut-off, there's all manner of interesting harmonics and the value of the comparison appears limited at best. Once I have my multi-band HF antenna, which I'm told is still being built, I intend to set this contraption up on HF where we can do point-to-point recordings and we end up having a direct comparison between two stations who transmit into my frequency-response software. I should add some disclaimers here too. At the moment I'm only using FM. The intent is to get this to a point where I can compare any mode, but when I move to HF, I'll likely start with Single Side Band and go on from there. One other annoyance is that any user needs to configure CTCSS to make this work, which is yet another hurdle to overcome, not insurmountable, but I like to keep things simple when you're starting to learn. Also, the harmonics still show, even on an analogue radio, so there's plenty more to discover. In the meantime, what kinds of things can you think of to use this for? I'm Onno VK6FLAB