Gap in digi_utils.cpp for longer paths #251
Replies: 4 comments 11 replies
|
are you looking to use WIDE3-n and higher? |
|
Hi Ricardo - yes, I am experimenting with long and skinny nets. I have a node at the house configured as an iGate, then another several dozen meters away, then another a few dozen meters further, then one with a high gain antenna and my mobile can drive around and might hit any of those nodes. some routes may be 3 hops right now, as I grow it further may be more hops. |
|
yes, but I never encounter the need to use more than WIDE2-n I guess I will have to add this too |
|
I live on the side of a mountain, and it takes only a little distance to have paths that are no longer line of sight so intermediate nodes are valuable. |
|
here's how it's reporting on aprs.fi. -12 is the igate, -11 and -7 are some digipeaters. -8 is the mobile. It seems that callsign substitution is not consistent? 2025-02-13 15:30:07 MST: N7UV-8>APLRT1,N7UV-11*,WIDE2-1,qAR,N7UV-12:!/=SaL20sL>GjQ also, there are some what i'd consider incongruities - 15:32:56 -8 was 5 km from the house, and i don't know which it was heard by. No asterisk. -12 is in the bathroom on the floor on a dummy load. That seems odd for a direct path. |
|
Take care as already several times mentioned, a LORA transmission lasts 2-3 sec. At 7 hops one (!) station locks the freq for almost 30sec. Consider one good location, use this as split-freq digi and thats it. To reach distances >> 10km is no magic, we reach here distances of 60km and more at ideal locations with TX power of 100mW. |
I don´t think so. |
|
wow! you have quite the LoRa APRS network there! Very impressive!!! I see most stations are igates and a few digis, so in general the traffic doesn't need to take more than 2 hops to get to the Internet. Currently there's a mobile DK5RQ-9 making its way north into town and appears to be heard regularly. I'm in the 6th largest city in the US, and our 902-928 MHz band is a bit of a mess. Perhaps that's a reason why my performance seems to be so different. Sometimes, I get good tracking out to 5 km, other times it's hard to get beyond a km. Like I said, there are no other people involved with what I'm doing, at least not locally. My closest co-conspirator %^) is nearly 350 km away. I do prefer to have RF to RF in order to be able to (eventually) have others be able to put up a station and see local traffic. Two hops is fine. However, on 2 m APRS, we have some stations heard here in Phoenix that are over in New Mexico, over 350 km away. Even stations in California, Nevada, and Utah are heard infrequently. Honestly, there just isn't much traffic anymore. It's a bit sad as many of the mobiles I see are using their phones and TCP/IP. I hear you on the delivered antennas. I have 10 dBi panels, 14 dBi sectors, and some 6 dBi omnis. I also have a number of good mobile antennas that I can use. I'm far more interested right now in generating some enthusiasm by having traffic that folks can at least listen to. Available channel capacity is not a problem yet! |
|
Ah, I think the 900MHz might be the difference - we are using here 433.775 MHz. Regarding 2m APRS - this is completely different to LORA, therefore digipeating at all is not recommended and should be avoided if possible. As said it is much slower, less expensive and more robust from modultation point of view. If digipeating is neccessary the split one should be used. Do you have the possibility to use 433MHz Lora from regulation point of view in the US ? BR |
|
Yeah, in the US we hams we can use the 433 MHz units, we have 420-450 MHz available through most of the country. It's been a bit difficult to get the 430 MHz stuff, I got two T3_V1.6.1 units, and am still waiting for the Supreme units, which are being shipped from China. I'm also running another alternative modulation/coding thing called CATS (cats.radio) on 430.500 MHz. I have 3 of those units (one is a digi/igate) and it's been interesting so far, but like LoRa, no one is out there. I hear you on digipeating - however, perhaps you don't remember how vast and empty the US west is, where significant cities are 600-700 km apart. I have digipeaters that have no chance of Ethernet (due to cost, and in some cases there's not even cellular coverage), so digipeating is necessary and accepted. There are few or no people using 900 MHz LoRa out this way, I even created a groups.io group called https://groups.io/g/Ham-Radio-LoRa-APRS to generate conversations about it. In the US, 902-928 MHz has great potential but is a cesspool in major cities. The regulatory agency (FCC) allows all sorts of activity in that band, and of course we amateur radio operators have to deal with all that. I have ham friends who are big into voice repeaters on 900 MHz, and they face constant challenges in operation of their gear. Even the 420-450 MHz band is not without challenges, but generally it's quieter except near large military bases. There's also rules along the Canadian border! But the key is narrow bandwidth to protect the receiver from extraneous interference. When I get my 433 MHz stuff I already have prepared an outdoor elevated antenna for it. It will be fun to see what kind of range I can get. Thanks for the info and the assistance - I appreciate it! Cheers and 73 - Jon N7UV |
|
you are welcome ! |
|
@arizonajon1 N7UV-12 is the one injecting to APRS-IS (it should not have * ) |
|
thanks Ricardo for the feedback! Here's what's bugging me. I have the mobile -8 and three digis here at the house (-7, -11, and -12), only -12 is configured as an igate, -12 and -11 are on dummy loads, and only -7 has a real antenna (10 dBi panel pointed SW). But, when I look at the -8 track, I see this and the raw data looks like this 2025-02-15 14:04:11 MST: N7UV-8>APLRT1,N7UV-11*,WIDE2-1,qAR,N7UV-12:!/=R'Z213L>HZQ Which makes it seem like the one that is receiving best is -11, even though it's on a dummy load here at the house. I haven't sorted out the c++ code well enough to understand why it looks like -11 is the first to digipeat and -7 is so rare, even tho it has the outdoor antenna. I'd expect to see a lot more -7* digipeats when i'm many km away from the house. Does that make sense? Salud and 73 - Jon N7UV |
|
it does make senses as LORA boards has a "crazy" amount of sensitivity to receive. maybe you need to take them really apart of each other to do the test in the way you need to do. Also it seems you are sending with the tracker with "WIDE2-2" only as the path, right? |
Uh oh!
There was an error while loading. Please reload this page.
Point: I am the only person AFAIK here in the Phoenix AZ area running LoRa APRS, thus it is highly unlikely that any interference to others will occur.
Since I have many LilyGO devices, I have been setting up temporary networks (mostly long and skinny, but sometimes fat and small radius. Sometimes I'm forcing path by using physical separation and antennas, other times with less physical separation but using a dummy load on the digi. I was using in my mobile device WIDE2-2, but when I extended the network to force 3 hops to get back to the igate, something strange happened. The intermediate nodes no longer showed up in the path, and only the gateway node -12 is recorded at aprs.fi.
I'm not a C++ kind of guy, but I think there's something missing in digi_utils.cpp that causes unexpected behavior for a path of 3-* and greater.
` String buildPacket(const String& path, const String& packet, bool thirdParty, bool crossFreq) {
if (!crossFreq) {
String packetToRepeat = packet.substring(0, packet.indexOf(",") + 1);
String tempPath = path;
The above code appears to check for paths of WIDE2-2, WIDE2-*, WIDE1, but does not recognize paths with "WIDEn" larger than 2 in the string.
It would seem fair to modify this code to look for WIDEn-n, check if n<= 7, decrement by 1, reassemble the path, and transmit the path/packet again. I think that's the way the standard goes, I'm not sure though.
Is that something that can be improved?
Cheers - Jon N7UV
All reactions