2023-12-22 - Reimagining Amateur Radio Repeaters in the MMDVM Era, Interesting Thread on digitalvoice Mailing List, Bob “Coop” Cooper K6EDX / ZL4AAA is a Muted Microphone
I agree with your article. Perhaps made a bad example. But I also think there are additional problems. 1. Lack of interest. Few are willing to roll up their sleeves and do this hard work, often for free. 2) poor documentation. I messed around with flexes apps, for example. They are doing a poor job cultivating a dev community.
Josh - Yes I am aware of that “installer” but apparently it was written for a previous version of SmartSDR and is not compatible with subsequent versions of SmartSDR.
It’s a mystery to me why FlexRadio doesn’t “simply” integrate FreeDV as a native mode as an example of the superiority of their software technology
Hi Steve - Maybe I’m just being dense but help me understand how high volume data networks would be used in ham radio. Still images when the need is maybe really for streaming video in emergency situations? Files of what kind of data? If the networks were there now, exactly how would they be used? Some examples that aren’t edge cases? Thanks! -Ed WA6AXX
Ed - We humans never seem to have enough communications capacity. We couldn't do VOIP over modem connections, but once we got broadband, we could do realtime voice AND then video. I don't think Amateur Radio will lack for things to "fill the pipe". As soon as we got above 20 Mbps, streaming entertainment was feasible.
One application I remain impressed with is "flamp - Amateur Multicast Protocol - file transfer program" - it basically streams blocks of files with checksums. If you miss a block, it will eventually be repeated. There are all kinds of files that could be streamed to you over Amateur Radio - DMR ID database updates, monthly repeater database updates, callsign database updates, Amateur Radio literature such as the fantastic Surrey ARC Communicator newsletterzine, etc. One compelling example I heard of is that APRS is a handy thing to have in an emergency, if the Internet is down, you better have your maps cached locally. And that's an easy thing to forget to do in advance. So why not "stream" the most recent maps via flamp to a Raspberry Pi in your shack, so that when you need the maps, just pull out the USB flash drive where they've been kept up to date, and go. And that's just one use of faster data.
The AREDN folks are demonstrating video, VOIP, monitoring realtime telemetry of repeater and network health, status maps, remote sensors, etc. Folks in Boulder, CO, San Diego, and elsewhere are having a blast with Digital ATV - basically a video intercom. Instead of doing that with a video-specific system, why build a network for DATA and you can then run your video AND other types of data on that network.
Absolutely yes, we could "shunt all of that over to Internet", but some of us (well, me, anyway) are involved in Amateur Radio because we (I) like to experiment with data communications over RADIO... just because it's cool and because we (I) learn things and get a lot of pleasure and fun from seeing my equipment doing so. One thing I can imagine is sending an I/Q stream from a remote receiver... I want to hear how my signal is being received at a receiver some distance from me. Etc.
But your larger point is well-taken - in advocating faster data, I should accompany that with examples of use cases for faster, better data communications in Amateur Radio.
Steve- Just started reading Zero Retries and it is very interesting with focus on things not talked much about in any Ham Radio magazines with a look to the future- very refreshing- Thank You
Every so often I try to repeat the origin story of Zero Retries, which is that I created Zero Retries solely out of frustration of knowing that there IS so much interesting stuff going on in Amateur Radio (and adjacent technologies / communities)... and it ISN'T getting discussed in any of the Amateur Radio media. We are an interesting bunch of techies doing interesting things that DESERVE to get mentioned.
And I keep finding more and more interesting stuff. Just an hour ago I spent 30 minutes talking to a very interesting fellow who is working on a very ambitious Amateur Radio project using technology beyond anything I've heard about to date. He might not be able to make it work... but it's really impressive that he's going to TRY to make it work - in Amateur Radio!
Steve said: I’m continually amazed that these systems are being developed, manufactured, sold, and supported… solo by WB6CXC.
-- And I am amazed at how easy it has become to do this stuff. The modern tools such as KiCad, and the (relatively) inexpensive scopes, signal generators, and spectrum analyzers, along with the extremely affordable PCB fab (and sometimes assembly) make rapid development and turnaround possible for us mere mortals. And of course the available software tools. And the CAD / NC milling tools for hardware prototypes. Yes, you need to understand how to design the circuits and systems, but that's what you get from years of working in the trenches. It's never been easier. Or more fun!
It also helps to hook up with people who have challenging problems that need solving. If I were doing this purely for my own amusement I would be much more likely to quit half-way through a project (I've got boxes full of half-baked ideas).
Boy, a lot to unpack here this week. I've been putting off building (rebuilding) the K0RV-DR repeater, mostly because when it went offline due to a wear-leveled microSD card a few years back I didn't get get it right back online right away, so some of the knowledge escaped to the aether. As things progressed the MMDVM-Pi hat was released along with a prebuilt Pi image of the Brandmeister firmware. Since I was basically the only ham club member with a D-Star radio I thought going to the Zum radio MMDVM Pi hat would be better for a club callsign repeater. However, there's registration across multiple networks needed, all the stuff that's specific to DMR, and generally not much in the way of a manual. Sure, there's plenty of "just click on this and enter that" how-to videos but not much that explains what's going on under the hood. So I get a little bent out of shape over not knowing what it is I'm actually doing, and go for a walk. Repeat every few months when the urge strikes. Don't get me wrong, I'm really happy that binary blobs exist, and the application is excellent. But without access to a good Elmer who can explain it at a level that goes beyond "click on the box" would be nice. And maybe get rid of the extra registrations and such. I have coordinated repeater with a license. Shouldn't that be good enough?
But, on a positive note, one of the big discoveries this year was ChatGPT 4's ability to write Python scripts, including GNU radio flow graphs. I've been using it for several simple tasks that I probably could write out on my own, but why? I still need to debug but it is much easier than staring at the blank screen wondering where to begin.
Eric - I think the ability of the Raspberry Pi 4 to optionally boot from a USB SSD was a big step forward for the reliability of a Raspberry Pi used in service such as repeaters. And now the Raspberry Pi 5 can (soon - we're waiting for the adapter cards) boot from SSDs via PCI-e (I think that's right). Another friend who I've been exchanging email with pointed me to another MMDVM repeater software base - https://w0chp.radio/wpsd/.
Thanks for the continued coverage of Project Yamhill! It's all been 'pencil whipping' at this point, but it should get a lot more interesting soon, and I'm pushing to get PCBs to fab in the very near future. Definitely want to beat Chinese New Year. I also just made a very large order of prototype components from Mouser, so things are in motion on that front.
The P3ST is interesting to me, as another modular type of project. Funny how multiple people will often come out with similar things at similar times.
Regarding the whole digital regulatory debate, I'm all for modernizing our regulations and easing up on the constraining regulatory environment for digital emissions, but I do feel strongly that legacy modes on HF such as CW and SSB need to be protected as well. Last I looked into it, the voice codecs used on HF that use the same bandwidth as SSB were still of inferior audio quality to SSB. Also, CW still plays a very important role as our premier low-bandwidth mode that can get through in marginal conditions and with simple gear. For the perhaps selfish reason that I believe there always needs to be a path to learning and deploying analog modulation in amateur radio (especially using home built gear). Also, I know that CW is very important for SOTA and POTA ops, for multiple reasons.
A long winded way of saying let's move forward without discarding the past that still is useful.
Jason - I don't take any pleasure in the observation that the HF bands are becoming less usable for low power operations and analog modes in general because low level interference has apparently been accepted by industry and regulators as the price we pay for technology such as inexpensive switching powers supplies, etc. Thus, I see the historical capabilities of the legacy modes fading a bit versus the capabilities of the newest data modes (which can overcome the noise issues). In some of the comparisons I've read, various WSJT-X (and other data modes) outperform CW for ability to get "data" through significantly below the noise floor. And ditto for voice communications using SSB versus FreeDV.
The Shannon Limit is the holy grail of communications theory (if you're not familiar with this, look it up, and the life of Claude Shannon), and many of the digital modes can get quite close to this. CW isn't even in the same ballpark.
But when talking about Signal to Noise Ratio (SNR) there is noise and there is noise. It's all different, and various modulation methods can compensate for different types of noise. The adaptive modulation modes and Forward Error Correction (FEC) are particularly good at getting through when things get gnarly. It's not clear to me if Spread Spectrum is allowed under the new bandwidth regulations (replacing baud rate limits), but if so this is another powerful technique that cuts through many types of noise. It's a fascinating area!
Of course you can send and receive CW using *extremely* simple gear, and that will always have an appeal (at least to me.)
Are you aware of
https://www.flexradio.com/documentation/smartsdr_freedv_waveform_installer-exe-2/
I agree with your article. Perhaps made a bad example. But I also think there are additional problems. 1. Lack of interest. Few are willing to roll up their sleeves and do this hard work, often for free. 2) poor documentation. I messed around with flexes apps, for example. They are doing a poor job cultivating a dev community.
Appreciate what you do. Keep speaking truth
Josh - Yes I am aware of that “installer” but apparently it was written for a previous version of SmartSDR and is not compatible with subsequent versions of SmartSDR.
It’s a mystery to me why FlexRadio doesn’t “simply” integrate FreeDV as a native mode as an example of the superiority of their software technology
Hi Steve - Maybe I’m just being dense but help me understand how high volume data networks would be used in ham radio. Still images when the need is maybe really for streaming video in emergency situations? Files of what kind of data? If the networks were there now, exactly how would they be used? Some examples that aren’t edge cases? Thanks! -Ed WA6AXX
Ed - We humans never seem to have enough communications capacity. We couldn't do VOIP over modem connections, but once we got broadband, we could do realtime voice AND then video. I don't think Amateur Radio will lack for things to "fill the pipe". As soon as we got above 20 Mbps, streaming entertainment was feasible.
One application I remain impressed with is "flamp - Amateur Multicast Protocol - file transfer program" - it basically streams blocks of files with checksums. If you miss a block, it will eventually be repeated. There are all kinds of files that could be streamed to you over Amateur Radio - DMR ID database updates, monthly repeater database updates, callsign database updates, Amateur Radio literature such as the fantastic Surrey ARC Communicator newsletterzine, etc. One compelling example I heard of is that APRS is a handy thing to have in an emergency, if the Internet is down, you better have your maps cached locally. And that's an easy thing to forget to do in advance. So why not "stream" the most recent maps via flamp to a Raspberry Pi in your shack, so that when you need the maps, just pull out the USB flash drive where they've been kept up to date, and go. And that's just one use of faster data.
The AREDN folks are demonstrating video, VOIP, monitoring realtime telemetry of repeater and network health, status maps, remote sensors, etc. Folks in Boulder, CO, San Diego, and elsewhere are having a blast with Digital ATV - basically a video intercom. Instead of doing that with a video-specific system, why build a network for DATA and you can then run your video AND other types of data on that network.
Absolutely yes, we could "shunt all of that over to Internet", but some of us (well, me, anyway) are involved in Amateur Radio because we (I) like to experiment with data communications over RADIO... just because it's cool and because we (I) learn things and get a lot of pleasure and fun from seeing my equipment doing so. One thing I can imagine is sending an I/Q stream from a remote receiver... I want to hear how my signal is being received at a receiver some distance from me. Etc.
But your larger point is well-taken - in advocating faster data, I should accompany that with examples of use cases for faster, better data communications in Amateur Radio.
Steve- Just started reading Zero Retries and it is very interesting with focus on things not talked much about in any Ham Radio magazines with a look to the future- very refreshing- Thank You
Tom - Thanks for the kind words.
Every so often I try to repeat the origin story of Zero Retries, which is that I created Zero Retries solely out of frustration of knowing that there IS so much interesting stuff going on in Amateur Radio (and adjacent technologies / communities)... and it ISN'T getting discussed in any of the Amateur Radio media. We are an interesting bunch of techies doing interesting things that DESERVE to get mentioned.
And I keep finding more and more interesting stuff. Just an hour ago I spent 30 minutes talking to a very interesting fellow who is working on a very ambitious Amateur Radio project using technology beyond anything I've heard about to date. He might not be able to make it work... but it's really impressive that he's going to TRY to make it work - in Amateur Radio!
Steve said: I’m continually amazed that these systems are being developed, manufactured, sold, and supported… solo by WB6CXC.
-- And I am amazed at how easy it has become to do this stuff. The modern tools such as KiCad, and the (relatively) inexpensive scopes, signal generators, and spectrum analyzers, along with the extremely affordable PCB fab (and sometimes assembly) make rapid development and turnaround possible for us mere mortals. And of course the available software tools. And the CAD / NC milling tools for hardware prototypes. Yes, you need to understand how to design the circuits and systems, but that's what you get from years of working in the trenches. It's never been easier. Or more fun!
It also helps to hook up with people who have challenging problems that need solving. If I were doing this purely for my own amusement I would be much more likely to quit half-way through a project (I've got boxes full of half-baked ideas).
Paul - Agreed that having a receptive, attentive audience for the things you do is a powerful motivation to keep doing it. 😀
Boy, a lot to unpack here this week. I've been putting off building (rebuilding) the K0RV-DR repeater, mostly because when it went offline due to a wear-leveled microSD card a few years back I didn't get get it right back online right away, so some of the knowledge escaped to the aether. As things progressed the MMDVM-Pi hat was released along with a prebuilt Pi image of the Brandmeister firmware. Since I was basically the only ham club member with a D-Star radio I thought going to the Zum radio MMDVM Pi hat would be better for a club callsign repeater. However, there's registration across multiple networks needed, all the stuff that's specific to DMR, and generally not much in the way of a manual. Sure, there's plenty of "just click on this and enter that" how-to videos but not much that explains what's going on under the hood. So I get a little bent out of shape over not knowing what it is I'm actually doing, and go for a walk. Repeat every few months when the urge strikes. Don't get me wrong, I'm really happy that binary blobs exist, and the application is excellent. But without access to a good Elmer who can explain it at a level that goes beyond "click on the box" would be nice. And maybe get rid of the extra registrations and such. I have coordinated repeater with a license. Shouldn't that be good enough?
But, on a positive note, one of the big discoveries this year was ChatGPT 4's ability to write Python scripts, including GNU radio flow graphs. I've been using it for several simple tasks that I probably could write out on my own, but why? I still need to debug but it is much easier than staring at the blank screen wondering where to begin.
Eric - I think the ability of the Raspberry Pi 4 to optionally boot from a USB SSD was a big step forward for the reliability of a Raspberry Pi used in service such as repeaters. And now the Raspberry Pi 5 can (soon - we're waiting for the adapter cards) boot from SSDs via PCI-e (I think that's right). Another friend who I've been exchanging email with pointed me to another MMDVM repeater software base - https://w0chp.radio/wpsd/.
Good luck with K0RV-DR!
Thanks for the continued coverage of Project Yamhill! It's all been 'pencil whipping' at this point, but it should get a lot more interesting soon, and I'm pushing to get PCBs to fab in the very near future. Definitely want to beat Chinese New Year. I also just made a very large order of prototype components from Mouser, so things are in motion on that front.
The P3ST is interesting to me, as another modular type of project. Funny how multiple people will often come out with similar things at similar times.
Regarding the whole digital regulatory debate, I'm all for modernizing our regulations and easing up on the constraining regulatory environment for digital emissions, but I do feel strongly that legacy modes on HF such as CW and SSB need to be protected as well. Last I looked into it, the voice codecs used on HF that use the same bandwidth as SSB were still of inferior audio quality to SSB. Also, CW still plays a very important role as our premier low-bandwidth mode that can get through in marginal conditions and with simple gear. For the perhaps selfish reason that I believe there always needs to be a path to learning and deploying analog modulation in amateur radio (especially using home built gear). Also, I know that CW is very important for SOTA and POTA ops, for multiple reasons.
A long winded way of saying let's move forward without discarding the past that still is useful.
Jason - I don't take any pleasure in the observation that the HF bands are becoming less usable for low power operations and analog modes in general because low level interference has apparently been accepted by industry and regulators as the price we pay for technology such as inexpensive switching powers supplies, etc. Thus, I see the historical capabilities of the legacy modes fading a bit versus the capabilities of the newest data modes (which can overcome the noise issues). In some of the comparisons I've read, various WSJT-X (and other data modes) outperform CW for ability to get "data" through significantly below the noise floor. And ditto for voice communications using SSB versus FreeDV.
The Shannon Limit is the holy grail of communications theory (if you're not familiar with this, look it up, and the life of Claude Shannon), and many of the digital modes can get quite close to this. CW isn't even in the same ballpark.
But when talking about Signal to Noise Ratio (SNR) there is noise and there is noise. It's all different, and various modulation methods can compensate for different types of noise. The adaptive modulation modes and Forward Error Correction (FEC) are particularly good at getting through when things get gnarly. It's not clear to me if Spread Spectrum is allowed under the new bandwidth regulations (replacing baud rate limits), but if so this is another powerful technique that cuts through many types of noise. It's a fascinating area!
Of course you can send and receive CW using *extremely* simple gear, and that will always have an appeal (at least to me.)