2025-06-13 — What’s New at DLARC - June 2025, M17 - N8GNJ’s Perspective as of June 2025, Followup On Single Frequency Repeaters, New Packet Commander IOS App, Jeff Geerling in QST, IP400 News
Now that I've received my second CS7000-M17 HT from Jerry and it's club night for the Clark County ARC (Vancouver, WA), I'm going to be able to have my first direct radio-to-radio M17 QSO. I'm also going to use this opportunity to evangelize M17, and maybe get some other hams interested in building an MMDVM UHF repeater for the club. I've got all the parts except for two things: a location for the repeater, and clear enough bench to build it on. -w7com
Uncle - Yeah, I hear you. I gotta get my Zero Retries Interesting demo capability built back up. First up, IP400, then MMDVM / MMDVM-TNC, NinoTNC (faster packet modes using IL2P), etc. M17's on the list too.
Tom - I gotta get my CS7000 M17 PLUS on order. Just need to wait a bit for the hobby funding to recover. But yeah, being able to chat with you on M17 over Internet would be very cool.
There's a theme in this issue that surfaces the desperate need for marketing and PR assistance to help take these incredible projects to the next level. Future project planners should budget equal parts development time to marketing time. With the current media environment being so fragmented, it's nearly impossible for one person to wear both hats. (I face the same issue with my own projects.)
Completely unrelated, I'm curious to hear more about how the machine learning piece enters RADE. I was under the impression that ML was used to help develop the technology, but this issue suggests it might be an active component in the encoding and decoding of each transmission. Or in other words, is the RADE software actively running some form of ML in real time?
Cale - It's my understanding (perhaps flawed) that the Machine Learning element of RADE is embedded (local) and operating in realtime as one speaks. That's why it requires the compute power of a modern Windows PC, or with what was demonstrated at Hamvention, a Raspberry Pi 5.
The VP-Digi is certainly interesting! I would be very interested if it had a 1 watt transmitter integrated. I've been wondering for a while about "hidden in plain sight" ax.25 digipeaters powered by small solar panels and Lipo battery packs. They could be easily placed on mountaintops to make use of the 1w. I suppose that's the use case for IP400 haha.
As it stands right now I think I'd just use a Pi Zero running Direwolf with a digirig lite or similar.
Raspberry Pis are great for experimenting and local use as appliances, but they really need some extraordinary measures to be fully reliable in remote use where you can't get to them easily. Off the top of my head...
* Mechanical - those little micro connectors suck and are fragile.
* Environmental - not really qualified for high heat or bitter cold or high humidity.
* Use of uSD cards for storage. With some creativity and knowledge, you can get them to boot from a USB flash drive, which is probably a bit more reliable.
* Power needs to be really really clean and sufficient power so transient loads don't drop the power supply below the threshold to keep it operating.
* For real reliability there should be a watchdog timer that if the OS or the application locks up it will reboot with no issues.
* Remote access is required sometimes, but that's problematic if it doesn't come up cleanly and get a good IP address, so need to set it for static IP address, etc.
Easy things that can be done:
* Use an external KVM that's much more reliable.
* Have a way to remotely cycle power.
* Do what is done with the DigiPi - boot, copy the OS and apps into memory and execute from there - no high usage of the SD card for the temp directory, logging, etc.
That's why for a relatively simple use case like a digi, the VP-Digi might be a great unit.
Eh in my experience for recent Pi Zeros this is a bit overblown. The full size Pis are indeed very sensitive to clean power at the right voltage, but I've run a pi zero 2w in a hot car handling Android Auto and it's shockingly reliable:
I haven't corrupted an SD card in a Pi since my first-gen Pi which did it with frequency. It is a good idea to minimize writes. Given neither setup here has an integrated radio, I'd probably have that handle transmit timeouts. Having a full Linux distro available means you could send APRS beacons for status.
I think if the entire setup is in a weathertight box it would do better than you think in most climates even outdoors.
All that being said it'd be awesome to have an integrated unit that did the 1w transmit! This is all borne out of me living near mountains that are primarily state parks haha
Re: new zBitx: check. It's scheduled for delivery :-)
Re: stamps...the owner of the W7RAT IRLP repeater in Portland is big into philately.
Re: M17 killer feature... For me, it is because it is open source. We are all slaves to vendor lock-in. I look forward to the day when open source, free-to-use codecs and modes reach critical mass and help to democratize amateur radio.
And once again, you and Kay have produced an amazing edition of Zero Retries. Do you sleep? Eat? I don't see when you could do those things, given the exceptional coverage you provide in ZR!
Well, Kay has his day job with DLARC and contributes his column monthly. But yeah, I put a lot of my waking hours into bigger issues like this one. The info just kept coming.
As I hinted, I'm trying to force myself to do more walking the Amateur Radio talk than just talking the talk as I've been mostly doing of late. Your work with Random Wire is much more balanced than ZR because you're reporting out from walking the talk... and I need to do more of that.
Now that I've received my second CS7000-M17 HT from Jerry and it's club night for the Clark County ARC (Vancouver, WA), I'm going to be able to have my first direct radio-to-radio M17 QSO. I'm also going to use this opportunity to evangelize M17, and maybe get some other hams interested in building an MMDVM UHF repeater for the club. I've got all the parts except for two things: a location for the repeater, and clear enough bench to build it on. -w7com
Uncle - Yeah, I hear you. I gotta get my Zero Retries Interesting demo capability built back up. First up, IP400, then MMDVM / MMDVM-TNC, NinoTNC (faster packet modes using IL2P), etc. M17's on the list too.
I'd love to try an M17 QSO in the evening, almost any day. We could also test my M17 reflector: M17-PNW. I am KJ7T.
Tom - I gotta get my CS7000 M17 PLUS on order. Just need to wait a bit for the hobby funding to recover. But yeah, being able to chat with you on M17 over Internet would be very cool.
There's a theme in this issue that surfaces the desperate need for marketing and PR assistance to help take these incredible projects to the next level. Future project planners should budget equal parts development time to marketing time. With the current media environment being so fragmented, it's nearly impossible for one person to wear both hats. (I face the same issue with my own projects.)
Completely unrelated, I'm curious to hear more about how the machine learning piece enters RADE. I was under the impression that ML was used to help develop the technology, but this issue suggests it might be an active component in the encoding and decoding of each transmission. Or in other words, is the RADE software actively running some form of ML in real time?
Cale - It's my understanding (perhaps flawed) that the Machine Learning element of RADE is embedded (local) and operating in realtime as one speaks. That's why it requires the compute power of a modern Windows PC, or with what was demonstrated at Hamvention, a Raspberry Pi 5.
The VP-Digi is certainly interesting! I would be very interested if it had a 1 watt transmitter integrated. I've been wondering for a while about "hidden in plain sight" ax.25 digipeaters powered by small solar panels and Lipo battery packs. They could be easily placed on mountaintops to make use of the 1w. I suppose that's the use case for IP400 haha.
As it stands right now I think I'd just use a Pi Zero running Direwolf with a digirig lite or similar.
Raspberry Pis are great for experimenting and local use as appliances, but they really need some extraordinary measures to be fully reliable in remote use where you can't get to them easily. Off the top of my head...
* Mechanical - those little micro connectors suck and are fragile.
* Environmental - not really qualified for high heat or bitter cold or high humidity.
* Use of uSD cards for storage. With some creativity and knowledge, you can get them to boot from a USB flash drive, which is probably a bit more reliable.
* Power needs to be really really clean and sufficient power so transient loads don't drop the power supply below the threshold to keep it operating.
* For real reliability there should be a watchdog timer that if the OS or the application locks up it will reboot with no issues.
* Remote access is required sometimes, but that's problematic if it doesn't come up cleanly and get a good IP address, so need to set it for static IP address, etc.
Easy things that can be done:
* Use an external KVM that's much more reliable.
* Have a way to remotely cycle power.
* Do what is done with the DigiPi - boot, copy the OS and apps into memory and execute from there - no high usage of the SD card for the temp directory, logging, etc.
That's why for a relatively simple use case like a digi, the VP-Digi might be a great unit.
Eh in my experience for recent Pi Zeros this is a bit overblown. The full size Pis are indeed very sensitive to clean power at the right voltage, but I've run a pi zero 2w in a hot car handling Android Auto and it's shockingly reliable:
https://github.com/nisargjhaveri/WirelessAndroidAutoDongle
I haven't corrupted an SD card in a Pi since my first-gen Pi which did it with frequency. It is a good idea to minimize writes. Given neither setup here has an integrated radio, I'd probably have that handle transmit timeouts. Having a full Linux distro available means you could send APRS beacons for status.
I think if the entire setup is in a weathertight box it would do better than you think in most climates even outdoors.
All that being said it'd be awesome to have an integrated unit that did the 1w transmit! This is all borne out of me living near mountains that are primarily state parks haha
Re: new zBitx: check. It's scheduled for delivery :-)
Re: stamps...the owner of the W7RAT IRLP repeater in Portland is big into philately.
Re: M17 killer feature... For me, it is because it is open source. We are all slaves to vendor lock-in. I look forward to the day when open source, free-to-use codecs and modes reach critical mass and help to democratize amateur radio.
And once again, you and Kay have produced an amazing edition of Zero Retries. Do you sleep? Eat? I don't see when you could do those things, given the exceptional coverage you provide in ZR!
Thanks Tom!
Well, Kay has his day job with DLARC and contributes his column monthly. But yeah, I put a lot of my waking hours into bigger issues like this one. The info just kept coming.
As I hinted, I'm trying to force myself to do more walking the Amateur Radio talk than just talking the talk as I've been mostly doing of late. Your work with Random Wire is much more balanced than ZR because you're reporting out from walking the talk... and I need to do more of that.