Amateur APRS over rNode #40
Replies: 9 comments 34 replies
-
|
Not knowing what you are doing is the mother of all invention!
Hope that all makes sense and answers your questions! |
Beta Was this translation helpful? Give feedback.
-
|
Thanks for the prompt and informative reply, Mark. Will work on this more today! |
Beta Was this translation helpful? Give feedback.
-
|
Cool, my pleasure. Let me know if you get this up and running, would also really love to read/see it if you post about it somewhere. |
Beta Was this translation helpful? Give feedback.
-
|
Am getting this error on occasion: "pi@raspberrypi:~ $ rnodeconf /dev/ttyUSB0 -T --freq 915000000 --bw 31250 --txp 5 --sf 7 --cr 5 The only recovery is to unplug and replug the rNode into the USB port, wait a few seconds, then it's fine. "pi@raspberrypi:~ $ rnodeconf /dev/ttyUSB0 -T --freq 915000000 --bw 31250 --txp 5 --sf 7 --cr 5 Any ideas? Cheers and 73 - Jon N7UV |
Beta Was this translation helpful? Give feedback.
-
|
Holy cats! I think it's working!!! Both rNodes are set to 902025000 Hz (very bottom of 902-928 MHz band), 31.25 kHz BW, and 17 dBm output. (Yes, I know I'm prolly on some repeater's input, but just for a little while, and only transmitting every minute or so for a second...) I've got the one rNode in the truck, cabled to a 900 MHz gain omni atop the roof of the truck. That rNode is usb-plugged into the Win10 PC in the truck, and I'm running APRSIS32. The rNode showed up as COM10 on that PC. No magic required (by me)! Here in the house, I'm running the other rNode connected to a tri-band Larsen 150-450-900 MHz antenna mounted in the window of the house. From there, I have a view that goes for 150 km or so (my house is ~60 m above average terrain). In the house, it's a Win7 box running PinPoint, another APRS client. The rNode there plugged in via USB, took a few min for Win7 to figure it out, and told me for it to work right I needed a reboot. So I did. The screenshot is from the house PC, in the PinPoint APRS program. I just sent a message to the truck, And here's the screengrab from the truck PC: I think this counts as a success. Mark, this is amazing! It worked exactly as you said. Now to see what kind of range I can get. It's time for a beer to celebrate, so the pub is a few km away. Let's see! Cheers and 73 - Jon N7UV |
Beta Was this translation helpful? Give feedback.
-
|
so far the range sux. Like 30 m. I will sweep the frequency and see if there's something not right there. |
Beta Was this translation helpful? Give feedback.
-
|
I've checked both antennas and the feedlines. While I can't guarantee the SWR, I suspect that the antennas are at least good enough for covering the neighborhood. So far that's not the case. As soon as I'm in a 30-40 m radius of the house, it works. Will keep experimenting. Mark, is there a way to get a SNR or some other form of RSSI from the unit? I think at least some of them had a value from 0-100 or something like that. Cheers - Jon N7UV |
Beta Was this translation helpful? Give feedback.
-
|
GM Mark - |
Beta Was this translation helpful? Give feedback.
-
|
That is very interesting to hear, thanks for sharing all the details! Also very much looking forward to hearing how increasing the bandwidth will affect results! I've had similar experiences with interference when placed close to clusters of computer equipment, and especially monitors. |
Beta Was this translation helpful? Give feedback.
-
|
did another drive test, this time with 125 kHz BW, SF 9, CR 8. 1.10 kb/s. Down at 902.080 MHz. Same antennas, same setups. This time I got at least 6.7 km range, NLOS generally, unless really close to the house. Encouraging. Darn! Time to go back to work. %^) Cheers and 73 - Jon N7UV |
Beta Was this translation helpful? Give feedback.
-
|
And here's the picture of the 900 MHz base station antenna taped in the front window... |
Beta Was this translation helpful? Give feedback.
-
|
Haha, that is perfect mounting! |
Beta Was this translation helpful? Give feedback.
-
|
I have found 900 MHz has a lot of noise in a metro area. Meter reading etc.
I had much better luck with 440mhz.
Steve N0FPF
On Wed, Apr 6, 2022 at 7:18 AM arizonajon ***@***.***> wrote:
GM Mark -
Thanks for the response.
The suggestion that the house end is suffering significant receive
interference is a very reasonable idea. Since it sees half of the city, the
noise floor must be substantially higher than the mobile end.
I have modified the setup at the house somewhat. I put a real 900 MHz base
antenna taped into the window facing west. It's about 60 cm long, and i
realize now it's subject to potential interference from my entire computer
setup with multiple monitors only a m distant. The original antenna was
similarly placed, but not as good. The mobile side has had no changes.
Drive test: Max mobile-receive range now is over 3 km, and moving vs not
moving seems to be no longer a big issue. The mobile is hearing the base
beacon much more reliably and regularly. As well, base-receive range
improved about 4x, so there is likely something to antenna placement and
local interference.
While one may think of PHX as a desert, it's actually pretty green, with
lots of water-sucking trees and shrubberies. Much of the paths between
mobile and base were very much NLOS.
A minor issue is that I can set the base station to transmit a beacon no
faster than once every 30 seconds, so my sample rate is somewhat low. The
mobile side is set for 15 seconds.
I will do as you say - increase the operating bandwidth to 125 kHz, and of
course move the center frequency up to 902.080 MHz to be fully in-band.
I am much more heartened by the results this morning.
Cheers and 73 - Jon N7UV
—
Reply to this email directly, view it on GitHub
<#40 (reply in thread)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/APRMG4VA462DEZ26EDND673VDWMKHANCNFSM5SRPD2AQ>
.
You are receiving this because you are subscribed to this thread.Message
ID: ***@***.***>
--
Pardon the brevity, sent from a mobile device. So there.
|
Beta Was this translation helpful? Give feedback.
-
|
Hi Steve @stevewa2066 What's your setup? Mark, is it possible to get 400 MHz rNodes? Cheers and 73 - Jon |
Beta Was this translation helpful? Give feedback.
-
|
I just have a few Rnodes to experiment with on 440 MHz. I’m Connected to a
outdoor antenna which helps a lot. Quite a few people doing APRS Lora as
well. Do a search.
In my 900 MHz days , even a window would make a difference because lots of
modern windows have stuff in them to give them a R value which also
restricts RF.
The FCC rules are archaic as far as digital, but that is another
discussion. Since I haven’t looked at the rules for awhile , I think we can
do 19.2k in 440 MHz. But the SS rules are different too. Time to read the
rules again!
If your just doing APRS , 1200 or 9600. Plus the slower the speed, the
greater the greater the distance you can go.
On Wed, Apr 6, 2022 at 10:02 AM arizonajon ***@***.***> wrote:
And here's the picture of the 900 MHz base station antenna taped in the
front window...
[image: IMG_5569]
<https://user-images.githubusercontent.com/16637451/162028611-742364dc-d18e-4c0c-a2ad-1165b307e78a.jpg>
Very fancy mounting. Only the best Scotch 3m painters tape. %^)
—
Reply to this email directly, view it on GitHub
<#40 (reply in thread)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/APRMG4UR6EWRKHTBFUKTNHLVDW7TFANCNFSM5SRPD2AQ>
.
You are receiving this because you were mentioned.Message ID:
***@***.***>
--
Pardon the brevity, sent from a mobile device. So there.
|
Beta Was this translation helpful? Give feedback.
-
|
At these EIRP levels, I wouldn't think twice about using a bit more than the 100KHz digital mode allowed bandwidth. As you say, those guidelines are ancient. I'm not in the US of course, and the "consensus on the street" here may be a bit different, but people are doing all kinds of whacky stuff in the UHF amateur bands that don't follow the equally ancient band guidelines we have some places. The point of amateur radio is experimentation. Either way, I unfortunately don't know of any software that will take a translate a KISS serial port to a KISS TCP/IP port, even though it seems quite weird if such a program shouldn't exist. You could cobble one together by just scissoring stuff from the Reticulum KISS and TCP interfaces, but I don't have time to do it myself right now. An alternative might be to just run the listening software directly on the Pi in the attic. |
Beta Was this translation helpful? Give feedback.
-
|
Also, I know this might be quite a stupid question, but you are sure that those are 900 MHz RNodes, right? Do they show up as such if you run But yeah, as @stevewa2066 said, another (and probably more plausible) possibility is simply interference from noise source in the near-field of the transceiver (which you wouldn't see on the discone), or possibly interactions from objects you wouldn't think would form part of the antenna system, but do. From the (admittedly not incredibly large) experience I've had with it, 900MHz is weird. From my wireless ISP days, I heard stories about 900 MHz links being so damned strange you would hardly believe it. Point the highly directional antenna straight towards the tower? No sire, hardly enough signal for a stable handshake. How about turning it 60 degrees off-direction and pointing it halway into the ground? Fine, we now have perfect signal quality and a stable link... Strange stuff I never really wrapped my head around. |
Beta Was this translation helpful? Give feedback.
-
|
Hi Mark - First, when I run the infocommand on the nodes, they do tell me they're "850 to 1020 MHz", or something like that. Don't have one on Linux in front of me right now. But, to your point, I know it's all the same chipset and that chipset goes from dc to daylight. I am somewhat sensitive about following the rules. And yes, the 420-450 MHz band here in the US has some extremely antiquated rules, written long before real digital, and still liberally sprinkled with "RTTY" whenever it needs to refer to "digital modes". We share it with the US Military, and they in theory could pull our access. (They have done that in places where the USAF set up some pretty damned cool phased-array radar systems.) I've seen the go-ins with ATVers and others when something causes their precious NTSC TV signals (with that 50 dB SNR need) to get the least frizzle on the picture of the dead indian (Native American). And then don't bring up the ARRL, which is a backwards bunch of geezers who still dote on CW on HF. Anyway, 900 MHz in principle is just fine. I have a fair amount of experience with the bands around 902-928 MHz for the past two decades. Foliage is way more important. Laser-clear shots are great, but not mandatory. The mobile reception I'm getting in my experiments are a testament to excellent propagation. But, as you surmised earlier this week, it does appear that the problem is at the base station end. I think I've swapped the transceivers just to be sure. Paths are reciprocal, as the physics notes, and there aren't any magic circulators or isolators out there that keep the signals from traveling back to the house. During the week I work so have less time to play. Sorry to hear that you can't think of a quick and dirty way to reroute the connection. I was hoping the miracle of netcat or similar would make that work. My *nix-fu is not great. I'll figure something out. I'd just not have to drag a whole terminal up to the roof to set up a quick and dirty base station. I have no attic, I'm either inside the house or outside. More like a ski chalet in that way. BTW, here's the view from the roof.
Cheers - Jon N7UV |
Beta Was this translation helpful? Give feedback.
-
|
Not good. It appears that one of the two radios has either a RX or TX issue. I thought I'd already checked by swapping them, but I did that explicitly today and definitely one works 30 dB better than the other. Will put each on a SpecA later and see if they both have the correct output power. |
Beta Was this translation helpful? Give feedback.
-
|
I don't know, Mark. I'm pretty sure I bought two on the same band. How would I know the difference? Examining the tiny tiny bandpass filter on the antenna line? |
Beta Was this translation helpful? Give feedback.
-
|
rNode #1 I've installed LoRaMon on both RPis. I'm rummaging around to find my step attenuator, so I can connect the two and see what kind of path loss they can tolerate. |
Beta Was this translation helpful? Give feedback.
-
|
Ain't pretty. 50 dB of attenuation. @markqvist How to send constant packets? |
Beta Was this translation helpful? Give feedback.
-
|
That is an incredible view you've got there. Everytime I see something like that I just wanna pack the bags and leave. Really want to go see Arizona some day. Is that Tucson in the distance? Or Phoenix? Ok, good they should be for the correct band then. Nah, wasn't because I thought you'd have ordered two different units and then forgotten about it, more that there might have been an error on my side when they were packed and shipped. If you want to be sure, you can also check the small print on the backside of the transceiver module, next to the antenna port. They should both say "9B", not "4". But it really should be ok, if both are reporting that via rnodeconf. Hmm, that definitely does not sound good. Very strange. You should be able to put at least 140db of attenuation on there and still have perfectly find SNR for decode. You can just slightly modify this Python script to loop sending packets: https://github.com/markqvist/RNode_Firmware/blob/master/Libraries/Example.py Just remove the part that takes Hope you get to the bottom of this. If there's an hardware issue with any of them, I'll send you a replacement :) |
Beta Was this translation helpful? Give feedback.
-
|
Alrighty - I didn't do anything fancy. I just pulled Example.py and RNode.py into a directory on each machine. I am running Example.py as-is, with all the settings that it natively had, including the +2 dBm TX power, except that I customized the text messages in each. And my test setup is as shown in the previous picture, two RNodes separated by 50 dB fixed attenuation. Given some additional losses in the test setup, probably a little more than 50 dB attenuation. So, 2 dBm should show up as around -50 dBm at the other receiver. SNR should be beaucoup. Radio serial # 1a:00:00:97 Radio serial # 1a:00:00:93 What I'm seeing does not make a lot of sense to me. SNR is wacky, unless it maxes out at 13-14 dB. And one radio is measuring RSSI around 14 dB higher than the other, and higher than the attenuators should allow. Ideas? Cheers - Jon N7UV p.s. Yes, downtown PHX is 11 km distant in the left-of-center background, where the two more distant "mountain" ranges overlap. Sky Harbor International airport (KPHX) control tower is behind the one tall palm tree at the far left. This is the view I always wanted, thanks! Had to get divorced to be able to have it %^) |
Beta Was this translation helpful? Give feedback.
-
|
I think I've solved the mystery of RSSI numbers that don't make sense. The script in Example.py where the operating frequency, tx power level, spreading codes, etc., does not get executed. Haven't spent time to figure out why. By using rnodeconf -i /dev/ttyUSB0 I discovered that the 2 rnodes are set up exactly the way I had them when i did the TNC testing. They are, as a refresher:
The important thing is that both claim to be set up identically, with 17 dBm output power. I brought out the old SpecA, put a 20 dB attenuator on the input, and ran a bunch of packets from each radio. Set to max hold. remember there's the 20 dB attenuator on the input, so the measured tx power on both are nearly identical at around -3 dBm, or +17 dBm through the attenuator. Both transmitters appear good. Here's the receive performance. Crude, had each transmit 3 packets at each attenuator step. Kinda sorta took the eyeball average of the 3 RSSI and SNR values, and recorded them in the table below: I didn't go beyond 100 dB attenuation since the leakage in the setup started to predominate. Howwever, it does appear that sn 93 has a receive issue. s/n 97 appears to step down pretty consistently, and the SNR holds steady. My conclusion? S/N 93 has a receiver problem. Comments or questions? Cheers - Jon N7UV |
Beta Was this translation helpful? Give feedback.
-
|
Just got back from a hiking trip, so haven't seen all your posts yet, but just a few quick, but important nodes:
Why one of your receivers are seeing a higher RSSI than it should (with the attenuator in there), is a complete mystery to me right now though! |
Beta Was this translation helpful? Give feedback.
-
|
The receiver seeing a higher RSSI than it should is an artifact of the TNC mode issue you pointed out. In TNC mode, I had set the tx levels to +17 dBm, so using +17 as the output, 50 dB attenuation brings that down to -33 dBm. Since I'd assumed incorrectly that the script did set the tx level, and the script reported back that the radio reported +2 dBm tx level, 50 dB attenuation would make that -48 dBm, which then makes it seem that one of the receivers is behaving in an impossible manner. |
Beta Was this translation helpful? Give feedback.
-
|
Just a note to say that I did purchase a couple of LilyGO paxcounter units. Each I was able to flash with the rnodeconf software, although I had some issues with github occasionally. Both of them do not appear to put out more than +2 dBm transmit, and very intermittently will communicate with the one rnode that I have that works well. The other rnode, like I've said before, has a poor performing receiver. |
Beta Was this translation helpful? Give feedback.
-
|
Oh, and another thing: The rnodes always seem to attach as ttyUSB0, while the LilyGO nodes connect as ttyACM0. Another mystery to me. Cheers and 73 - Jon N7UV |
Beta Was this translation helpful? Give feedback.
-
|
Okee-dokee. This afternoon I found my 900 MHz cavities, two 15 cm single-cans, one 4-can multi-coupler for 2 receivers. Chose to use the multicoupler, just one side. So it's tuned to 902.020 MHz. It's got about 1.5 dB insertion loss, tho I could prolly tweak it a bit. 50 cm LMR240 cable to the antenna, still taped in the window. That's the house-side (base) setup. (Reminder: I have about 200'/55 m clear over the level of the valley immediately below.) The house is the base. The windows box to which the LoRA transceiver is connected via USB is running APRSIS32. It uses the LilyGO 2.1 v1.6 unit, the mobile the known-good RNode (OR). Both are set for 902.020 MHz, with 31.25 kHz bandwidth. The LilyGO (I need help on this, @markqvist ) has only +2 dBm output, no matter what rnodeconf command I send. The OR is the full +17 dBm. The truck side is running APRSIS32, with only the OR attached, as well as GPS (of course). The antenna is roof-mounted gain 900 MHz on an NMO mount, with LMR240 coax (5 m or so). Here's the latest plot: If there was only a way to fix the LilyGO 2.1 rnodeconf issue to get the tx power up to +17 dBm... (@markqvist, I am looking at you.) I even thought of taking the second RNode, the one with the crappy receiver, and putting both that one and the LilyGO in the truck, with different calls, so I could get some kind of symmetry. Too much work. I think this is pretty impressive. At the house end, the whole shebang needs to go up on the roof (antenna, cans, radio, and an RPi) but I don't know what to do to remote the KISS function so I can communicate with the RPi via the network. Any ideas? Cheers and 73 - Jon N7UV |
Beta Was this translation helpful? Give feedback.
-
|
Very nice Jon. Thanks for all your detailed reports on this, I really enjoyed seeing this. Sorry for being a bit absent regarding the driver issue. I've been on the road for some days with not much time to check in on github. I really thought I had noted somewhere that those LilyGO units need driver installs on a lot of operating systems, but appartently I didn't do it visibly enough, if at all. I'll put up a better notice about it, can be quite confusing. Pretty strange that those units are stuck at +2dBm. That is definitly NOT the intention. Should be an easy fix in the firmware though, I will get to it ASAP. Am on the road right now, but will find a fix as soon as I get to a location where I can work on it. All in all, that is great work Jon. Really cool to see that plot and the range you are getting. I'll get back with the firmware as soon as I have a chance! |
Beta Was this translation helpful? Give feedback.
-
|
Hi Mark - |
Beta Was this translation helpful? Give feedback.
-
|
Thanks for sharing that @arizonajon! That IS pretty good for just +2dBm! I had to focus on some of the outstanding tasks in Reticulum for a bit, but now that RNS 0.3.7 is almost ready for release, the output power issue is at the top of my list! I've had a LilyGO unit sitting on my workdesk for a week now, just staring at me ;) I will have a fix out this coming week. |
Beta Was this translation helpful? Give feedback.
-
|
And it is out (v1.30). It took longer than I had anticipated, but at least it is here now. |
Beta Was this translation helpful? Give feedback.
-
|
Okay, it took longer than expected, but the TX power issue should be fixed now. You can simply update the RNodes with |
Beta Was this translation helpful? Give feedback.
-
|
Just for clarity, the fix is in version 1.30. |
Beta Was this translation helpful? Give feedback.
-
Hi Mark et al -
I finally got my two 2-yr-old rNodes up and running. I am certainly sure I don’t know what I’m doing mostly, but at least I’ve updated firmware on both to 1.27 from 1.10! Yeay!
I want to replicate your APRS test you did back in 2018. A couple of questions:
I hope y’all can help. I’m really interested in seeing what kind of range I can squeeze out, at a baud rate similar to 1200 baud APRS.
Cheers and 73 Jon N7UV
Beta Was this translation helpful? Give feedback.
All reactions