Hacker News Reader

Comment by edrxty | Hacker News Reader

The sdr situation is rough right now. The typical people who are hams and are buying the bulk of the radios are not the kind of people who are great with computers at anywhere near the level necessary to benefit from SDR. These same people also fight back against regulatory changes that would benefit the strengths of sdr (usually anything bandwidth related).

There are products out there but they generally require SWE level expertise to use unless all you want to do is listen to AM/FM. The only one that really has gotten anywhere is the IC-7300 but it's software locked to behave like an old superhet radio but with a waterfall and it famously doesn't allow IQ output because the legacy manufacturers are assholes. Everyone is market segmenting heavily off of firmware features instead of allowing the hardware to be used for custom applications so very few makers are doing interesting things with sdr beyond the rtl-sdr.

View on HN

It's such a sad state of things, really where ham radio should be going but instead it feels like things just get stuck in the past(unless you're talking FT-8 but even that is less about experimentation and more just another mode to get contacts on).

Elecraft gear has pretty good output/inputs but it's spendy, it's a hard ask for someone to fork out $1k for a 10-15W radio.

The whole bandwidth/symbol rate thing is also just incredibly short sighted, a number of other counties have started to change things but here in the US we can't even seem to get something as simple as that updated.

If you want to play with the intersection of radio and software, get an SDR dongle kit, and put GNU Radio on your laptop to play with it. You'll learn a ton about DSP and the things are fairly amazing, considering the price.

Also know that GNU radio can do audio I/O... so you can hear what you output, if you just want to play with sine generators, etc.

I managed to build a VOR receiver (flowchart) that correctly gets the bearing (heading?) to my local beacon. It was fun.

This has huge appeal to me, but I have no idea what I'd be missing or what weaknesses such a setup would have versus spending hundreds of dollars on a splashy portable radio.

The main weakness, in terms of use, is that you're lugging around a laptop with a dongle sticking out of it... which could break. The CPU and I/O over USB speed of modern laptops is amazing.

The main technical weakness is that $30 RTLSDR dongles only have 8 Bit ADCs, but then you're down-sampling the heck out of things, so that really doesn't matter, either.

I carry around laptops anyways so that's less of a concern. I also have played around with using my phone's USB port for USB audio & video recording, and hope for more growth over time in that segment; even smaller an option, with similarly huge CPU & I/O. In my fantasy world, I could plug a SDR into any device & load a webapp with webusb, that under the hood uses webgpu for acceleration.

Im going to try to list a couple more limitations. Rtl-sdr notably is Rx only; price goes up a lot I think with tx too. I'm not sure what kind of amplifiers ham radios have (tx only? Rx too?) but at least having a couple W output would be good. And maybe a Rx amplifier, especially if we are talking low bit depth. Other elements in the analog path of ham gear might be filters; I don't know what radios tend to have builtin versus what hams tend to add themselves here, but that seems like another factor.

I bought & have barely used a pretty fancy LimeSDR USB device a long time ago (~$300 then, wish they still made it!). I keep meaning to come back & even finally bought a range of antennas for it. I look at stuff like the $1300 LimeRFE front end, & radio products like this submission & wonder what analog stuff is missing, wonder what else a good ham-grade SDR might have that the dongles & higher end SDR might not have. I like the idea of going software based a lot, it seems appealing, but I think I'm missing an understanding of what hardware is important, is still super good to have.

A friend got me an HF upconverter and better RTLSDR dongle, and we were able to receive signals at 0.1 µV, which is pretty much as good as it gets, there's always about that much noise from the universe. The 8bit limitation generally isn't a huge issue, unless there are very strong local signals in the same capture bandwidth, in which case they'll likely fall well below the quantization limit.

You're taking 2,000,000 samples and down sampling by a factor of at least 40 in most cases, which gets you 5 more bits, before demodulating it.

Bear in mind that there are very stiff penalties for transmitting without a license, up to a year in prison and $10,000, for each offense. The FCC doesn't take kindly to pirate radio.

I'd like to see this conversation have evolved towards more specific & more technical.

Yes, a lot of hams are just crufty & computer averse, and that has kept back SDR. But also I think SDR so far leaves out a lot of the bigger picture needs, and I've been desperate to steer this thread away from "it's great" anec-data & into something more enriching & technically distinguishing.

This submission is so interesting in that there are plenty of radio devices here that have kept semi-conventional ham-ish form factors & capabilities, but which have SDR under the hood. My feeling is that less than 50% of that is software; that a lot of these radios are integrating analog signal chain systems normal ham gear would have like filters and amplifiers, that one has to pick & choose & build out - often with subideal results - to successfully replace ham gear with SDR.

This topic deserves better consideration. Maybe effective but depth is sufficient replacement to all these other typical analog doodads & gizmos; that could for sure be true. But I'd like a deeper exploration than 'we got great results & resolved a weak signal'. I think the $1300 LimeRFE exists for a good reason & purports to do good & valuable things we've want from good radio solutions. Rather than leave the conversation at some basic simple level like receiving a quiet signal, I think a deeper explanation & comparison would inform & help.

Quite a few of the things that distinguish "high end" radios these days are in terms of user interface, and connections to accessories. It used to be that crystal filters, the ability to select bandwidth, noise suppression, and other dedicated circuits made the radios better. With SDR, if you have sufficient bandwidth, bit depth, and dynamic range, you can do things that were previously $10,000 or more.

I was just trying to make the point that the cheapest, entry level, SDR (based on a test mode of the receiver chip) can do stuff that used to be expensive. The fact that a laptop can easily do a gigaflop makes it all possible.

Of course there are reasons to spend the money on a good radio. Lower noise figures, better clock stability, internal signal processing and UI. Dedicated knobs for functions are amazingly undervalued in terms of usability.

Getting an actually FCC Type approved amateur transceiver, gives good assurance of spur free output, and costs quite a bit to produce.

One modestly-priced and commercially-available SDR transceiver with open sourced hardware I know of is the trUSDX (Manuel, the hardware guy, is in one of the clubs in my region!) https://dl2man.de/

Use the "where to buy" links to see suppliers he has tested demo units from, both for kits and for ready-to-use devices. There's no guarantee from him about the quality (he's open-sourced the design and isn't paid for quality control), but it's a starting point. Check the forums to find out about good and bad examples users have encountered.

It's designed to use the 80m, 60m, 40m, 30m and 20m amateur bands, but as he says, there are ways to modify the filters if you're building the kit yourself to get other bands.

Manuel told me that he's mostly interested in CW (Morse) operation, but that it can be used for some digital modes without an additional soundcard - otherwise, regular PC soundcard connections will do.

Edited to add caveat: the firmware is closed source and available on request after showing evidence of a successful hardware build. They did this because they were fed up with all the out-of-spec clones on the market and those deceived buyers then complaining to them about it not working, and that is why it's so hard to find an SDR with openly-available firmware.

If you want something that is still open, look for their previous design, the uSDX, but with a serious caveat emptor on pre-built units.

I have a fun little addin board for my 7300 - https://www.radioanalog.com/ptrx-7300/ - Install was pretty straight forward, maybe 10 minutes.

It taps the internal wiring to grab IQ output, suitable to then feed into a SDR (RSPDuo for me), then into a computer from that.

It's receive only, but works pretty well. And you can always hook up the (natively supported) USB connection on the radio for transmit.

Reply on news.ycombinator.com