#us - NYC metro
1 messages · Page 7 of 1
I need pictures of your nodes, because I can't help but picture them sitting inside faraday boxes... or you're running them with original antennas which would explain a lot lol
Nah I have a e1000 which works somewhat well within a few blocks radius, and a WisMesh pocket v2 that ive tried various antennas on
Muzi, an AliExpress, now that one mentioned recently.... TE Connectivity Linx 915Mhz Antenna for Meshtastic LoRa SMA(M) ANT-916-CW-HW-SMA
the stock whip it comes with is tiny and I took it off immediately
at home I have known pretty aggressive window treatments that block RF but not sure about work here
916? oof. I'd recommend this one:
https://www.aliexpress.us/item/3256807293036956.html
16.7CM 906mhz
I have one of every size, in multiples lol
starting to wonder if my wismesh could be broken tbh. yesterday purposefully went to a 20 floor hotel bar at delancey bowery on day off and no ACKs up there. lots of RX.tho
huh, its possible - that def doesn't sound right if it's having trouble with TX/ACKs somewhere you're getting good RX
have you tried watching the logs over serial at all? other thing is what does the channel/airtime metrics look like in the app? could also be you're somewhere with >25% usage which the firmware won't send in
Same situation at the 7th (topmost) floor in Briarwood, NY. Every message I send has the cloud with the strike through it, whenever my 2nd node is powered off.
Well, I'm using my Seeed Studio SenseCap Card Tracker T1000-E, but similar enough.
i've sent a dozen or so messages on the default LongFast channel. I get the cloud with the check. And I can see the mesage on my second node. But get no response on the channel. Could it just not be getting to any nodes?
I started to DM people as soon as they send a meassage in the default channel. Asking about their equipment etc and get no response
i see a cloud with checkmark on the DMs
and I can message back and forth to my second node / phone
Yeah, I've noticed that the local mesh is very inconsistent in whether messages get through.
i'm showing 75 nodes
It cant be that no one responds when they see messages. I mean what is the point. I must be doing something wrong
Am I in your NodeDB?
Can you see any of @gardener2549's nodes? Most of them have a region in NYC and then four hexadecimal digits, with the latter matching the short name.
Such as this one:
Yes. I didn't want to ping them, but I did mention their username.
Ok I'll DM them
If that's ok
still dont know the etiquette but damnit i want to chat 😂
one can only send so many messages to their second node before it gets weird
The nodes are on rooftops around the city, specifically co-located NYC Mesh's infrastructure.
my wife is like .... "who are you texting" I have to explain i'm texting myself lol
Nope 😦
https://map.nycmesh.net/
BTW, not every location on that map is a Meshtastic node.
but NYC Mesh doesnt have anything to do with LoRa / Meshtastic, right?
Correct. They just happen to have access to ideal locations to put Meshtastic nodes at, with a power setup already in place.
ok got it
sent them a DM. Hopefully I'll hear back 🤞
That's me, Carroll Gardens MT
From my understanding, they only see the DMs whenever they climb onto the roof and connect to the node itself. I do remember gardener2549 (someone on this Discord) saying that he setup a way to remotely control and reboot those nodes, using a Raspberry Pi. Though I'm not sure how often he checks on it.
makes sense that they wouldn't get the DM unless his phone is connected via bluetooth or a serial cable to PC etc.
Can you do a teacerout to his node? see how many hops you are away?
here was a screenshot from yesterday on 20 th floor. Muzi whip
I've never seen channel use 25 pct
do you have a node called ZLLA1 ? Showing up on my node
yes. zlla1 is a t1000
zlla is the rokland
zlla1 seems to work at close range. I can hit porks nodes if I am down on Grand St(BK) etc.
Assuming you still have that setup in which you can remotely access the nodes using a Raspberry Pi, one way we might be able to do a "range test", without involving physical activity (it's cold outside), is to look at the NodeDB map, but filter for nodes accessible in exactly zero hops.
Don't bank on people checking any of the infra-nodes DM's. THey could get rebooted at any time, and likely aren't monitored.
Which reminds me, we should probably consider adding something to the long name for infra nodes to indicate that they aren't monitored. And if anything is perma-powered, we should consider using those nodes to host ping-bots.
now or some other time?
Which also reminds me, @brazen spear - I know your free time is limited, but i think we need to discuss a full node donation. aka a boomer (1W).
I'd give you the node 100% completed, but it's gotta be in a REALLY prime location. Preferably powered so we can put a luckfox up there, but if not, I we can do an MCU based one with Solar.
It's showing me it's CLIENT_MUTE. Also is the power off? Is that what the battery indicator means? I sent a DM but didn't go through.
it's on and next to me right now. what's your nodeid?
at work I keep the t1000 client_mute on me and put the WisMesh in the window.
Above is @toxic warren's Node ID.
oh you're Grand St nice. I'm facing away from you
No, the topmost node listed is his. GrandSt is gardener2549's
oh right ok. I see you in the nodedb on ZLLA from 21h ago and not at all on ZLLA1
@abstract iron try this one. It's my base station.
This is my other one:
Anyone, everyone feel free to DM me
can't see BGC no reply CGM
Let's wait a bit and see if they connect. I'll keep working on it! I just power cycled BGC.
all this talk of no acks got me wondering if anyone is hearing me. just did a quick test and someone immediately ack’d 😂
I ack'd, but only because you said please
oh your message didn’t reach my node. interesting
how much did Ray pay you to ignore me 😄
it's just the mesh being the mesh really, i def have probs getting msgs out even though they get an ack 🤷
This is what my node saw, in Briarwood, NY:
this is pretty much what i’m seeing too
@normal osprey that was me
I'm BGC Base_GC
I can message you on the LongFast channel but you dont show up in the list of nodes 🙄
i don’t have you in my nodedb either but i always hit the limit and have nodes roll off. i also switch devices a lot so the list gets cleared from my phone
i just turned this node back on a few minutes ago
i have 2 nodes here. I get a couple of your messages on one node and a couple on the other.
how do you clear a nodedb?
are both of those nodes running CLIENT role, or is one set to CLIENT_MUTE?
both client
AHHH, try setting one to CLIENT_MUTE as that role will stop it from retransmitting anything it hears, with both as CLIENT they're quite likely causing some confusion as they both in theory are hearing and acknowledging each other vs someone else hearing you and returning an ack
So set the "base station" to CLIENT_MUTE?
leave that one CLIENT the other one CLIENT_MUTE
it's a very common setup to have your "inside" node CLIENT_MUTE to not retransmit, and the main / window / base station as CLIENT since it will retransmit what it hears from others on the mesh
both being that close to *each other is also not really great, the ALFA is probably the better of the two antennas so leave that taped up there and move the t-echo away from the window
I got the T-echo for traveling / around town etc. I don't mean to use it at home. But also wanted to use the T-echo to test my WisBlock
So I don't feel so lonely 😂
haha, most of us have been there - got a 2nd node to confirm the first is actually working
thats a solid plan though, the t-echo is very compact and more discrete to carry around vs giant honker antenna
I'll put the t-echo on CLIENT_MUTE. Or just turn it off til I leave home
i gotta get back to work but we can try couple other things to confirm / optimize your setup later
No.
my t1000 working well downtown facing got ACKs.
Would be cool to put up a 1watt'r with a good antenna. Would have to set it to amateur which I'm not sure the software limitations of. Would it still retransmit p2p encrypted messages?
Does this connect to SouthSlope?
Anyone having trouble getting an ack should bump their hop limits to 5.
See above for someone discussing the unlicensed power limits.
Oh great!!!
Super happy to bump some extra wattage. Just need to have the ears for a balanced equation.
What is the power consumption of the boomer @proven grove ? Would it be possible to have a battery backup for it? Or not really feasible?
I was talking with folks in the Carolinas who had meshtastic set up, they were incredibly frustrated to find half their nodes down after the storm and their network in shambles.
boomer is the WashTastic, there's a few other 1W designs like Ikoka Nano and Donktastic
i just sent my spare PCBA Ikoka Nano v3 to buddy earlier which had my last E22-900M30S 1W radio on it, on the upside (2) more along with (2) 33S modules are somewhere in the US waiting for ali to deliver and I have a few more Donktastic PCBs
The power consumption is stated in the README: https://github.com/valzzu/meshtastic-pcbs/tree/main/WashTastic#power-consumption
I'd... use that as a low estimate, IIRC Iris doesn't have a lot of nodes around. And the 1W module is def hungry, wants 5V 650mA during tx though idle is typical for SX1262 modules
Yes both my nodes connect to SouthSlope cdef
Just sent them a DM
Donktastic on the bench psu, those spikes are TX
@jaunty harness thank you for covering for me lol, in the middle of cooking dinner.
If it's semi-permanent location, I'd go with a femtofox with an M30S, then using an external charger for the battery backup.
@brazen spear any thoughts on location? Because if it's pretty high up I'd likely plan for significantly colder temps and go with either a LiFePo4 battery, LTO or a Sodium. Then again, I've been toying around with the idea of a standard lipo with a small warming coil in the enclosure, but I'd likely only do that if it was on semi-permanent power, with battery backup. Like throw a 10aH pack in there and the coil with the luckfox. Otherwise if it's any of the exotic chemistries I'd need to get my hands on one of Keith's boards (MCSBC). I have panels and the whole nine, so I can easily outfit it with a great setup. I'm waiting on PCB's and parts for a babblefox which also has 4G for internet. So that'd be my prime pick to go in. It's using an LR1121 radio so not a boomer, but with the LTE and the PA (E21) in there, it should be even more powerful.
Heh I got the PCBs but haven’t ordered the parts yet
battery chem is def a good point, could maybe do some the temp cutoff fuses ... Golf IIRC uses to cutoff charging below/above certain temps
other concern with deploying a *fox right now is foxbuntu being so in-flux, but you can flash whatever version on it and let it ride since it's not like apt doesn't work (just... painfully slow!)
this is the batt pack Austin Mesh recommends, I didn't realize it also had temp cutoff fuses builtin (as well as a solar input and always-on usb out) https://voltaicsystems.com/v25/
kinda pricey but also .. perfect
oh HAH I also never realized they're in Brooklyn!
I have multiple panels from them and a pack. I don't really love usb packs though for nodes, but that's just me.
I've got a v2 faketec out back with one of the tiniest panels from voltaic, paired with Keith's board and 3x 18650 LTO's. Thing hasn't gone down once in 4 months.
you can def build a whole setup for just the cost of that pack but it still seems perfect for dead simple solar node - panel to batt, batt to node's USB and done - but of course WHERE IS THE FUN IN THAT?!?!?
Did you know that Adafruit (https://www.adafruit.com/product/5035) is also located in Brooklyn?
yeah, been a customer since they were in Astoria, and their USPS shipping is 2 days max within the city! also used to know a guy who was in college with Limor and no matter what project I had going on he'd always be like "Oh you should see what my friend built" and it was always something awesome like the x0xb0x
According to the UPS Ground map, shipping from Mouser's location is three days.
oh nice, digikey i usually get 5-7 days USPS, but usually there's a weekend/sunday in there
Yeah, USPS is likely longer.
I miss the days when you could put in an order with Jameco on mon night and it'd show up usually thurs, fri latest via USPS priority
Rokland's $5 FedEx is pretty good, but generally i'll skimp on the shipping / prefer USPS cause they sure need the money and for the longest time were the only ones who could access my building to deliver packages
(and yes I have PTSD from having gone to the hell that is the Williamsburg PO multiple times)
hi guys
been a few weeks since I've talked in here 😛
happy to see some new usernames, welcome 🙂
I think you missed the coordination of putting Meshtastic nodes at existing NYC Mesh locations (https://map.nycmesh.net/).
Hey pork, what is this view from?
Meshsense - https://affirmatech.com/meshsense
Here at Affirmatech Inc., we create and provide affordable, fun, easy to use software and services with courteous, friendly support.
if you run it be sure and go to the settings and disable the automatic traceroutes or set them to a more reasonable value for a busy mesh like 120mins
yeah that software is a ddos machine otherwise
if your traceroutes actually work 😄
new makerspace coming in bushwick (via adafruit blog) https://brooklynspark.com/
traceroute 🤦♂️ heh. I've got confirmed checkmark off S Slope down here but still no trace
oh hey it worked
hah! yeah it works like 10-15% of the time
what's that weird second hop 0xffff ?
It means there's either a REPEATER or CLIENT_HIDDEN in the mix.
weird. the trace back makes sense. window pocket mesh with big antenna, zlla1 t1000
The trace towards would make sense if someone nearby set their node to REPEATER, as it would rebroadcast it before your window node would get a chance, thus inhibiting your window node from making a rebroadcast of its own.
i think older <2.4 firmware also comes back as ffff (somethingsomethingpacketdecryptionsomething)
The NodeDB stopped updating four hours ago on my node, strange.
However, my two nodes can still communicate with one another.
Wait no, nevermind. It's updating again.
haha, maybe just didn't see any new nodes in that time window?
That's a pretty inprecise measurement. Would probably be fine with 4 18650s. I'm really not too concerned with the battery composition since they're kept fully charged they should last many many years before degrading. Would probably need a bigger solar panel than what I've been using tho, if we wanted these to stay up through a week-long power outage.
I'm using 18650s since I have a big stash of free ones from a recovered ebike battery.
I think the composition matters more when it's a solar only node and goes through regular discharges.
there's some temp factors, like LiPo you don't wanna charge/discharge below 32F but LTO is good down to like -40F/C (where they converge)
but then you're also paying more for LTO cells and something that can properly deal with LTO vs LiPo (which is like $0.56 for a PCB off ali)
The solar/battery nodes we built are on their second winter without issue running on a single 18650. Haven't seen any significant decline in max voltage. They only ever discharge to like 90%. But these are neighborhood nodes, not central repeater locations with tons of traffic.
mmm, last winter wasn't too bad but def had a bunch of 20F days... guessing inside the box it's staying above freezing so they're fine just the usual LiPo cold weather behavior or some capacity loss and not charging em at fast enough to matter
they're kept fully charged
If this is only for power outages, maybe lithium metal batteries might be better? Their operating temperature goes down to -40° (source: https://energizer.com/wp-content/uploads/2023/11/123.pdf).
I tried but I RX better than TX
What's your username?
that was from pork d446 / PRK1
I sent you one Animal. I would assume it handed off to the s slope...
You aren't in my nodedb 🤦♂️
You sent me those tests? I never got them!
yes. hilarious that traceroute works but not msg in this case
I'm verified check to S Slope
you're only visible in my winmesh nodedb not t1000
i'm at 5 hops so my packets should be relayed enough from williamsburg but also only send out position/telemetry/nodeinfo on regular intervals .. can also check for my userid !865ad446 as you might have only heard a position/telemetry packet
zlla1 is my t1000(very indoors, intended to bounce off the ZLLA in window)
that makes sense (not the asymmetry but the location)
the asymmetry is pretty normal, i'd be more surprised if it was symmetrical with our* density
ok
So turning off position for a channel setting doesn't work 🤦♂️
I feel .... exposed 😂
do you also have it enabled for the private channel? it might be preventing you from disabling on the default channel maybe?
Animal you try my zlla1?
I do but the way they explained it it should be per channel settings
Will do now
yeah i'm just not 100%, i ever have it fully off or on with slightly adjusted coords + enhanced precision as we all get piled ontop of each at grand/union up here 😄
I'm getting the confirmed-check back slowly for some of the messages..I sent you a few
the person-check
k the wild thing here is that zlla1 is a t1000 inside. it's skipping zlla in the window
that's the way the mesh works - if something else also hears t1000e clearly, however faintly, it will then retransmit it. since it heard with a lower signal it retransmits it quicker than zlla in the window which heard it really good so has a longer delay
it helps propagation / decreases density problems... when it works
ohhhhhh I see
there's some other stuff going on, like a REPEATER/ROUTER/ROUTER_CLIENT (but not ROUTER_LATE) would hear your packet and try and repeat it pretty much immediately / before a client would, which is why having too many / poorly placed ones can hurt the mesh as they'll do their thing before clients which are actually designed to also be repeaters on their own
yeah its a common misconception... like those people running infrastructure roles are doing it for the right reasons (wanting to help) but not realizing actually hurting
that ffff cropping up
it's probably older firmware, there's still a few out there
the new S Slope node is a monster... that's blocks and blocks from me and to an indoor node
re: the return path second last hop
yeah Gardener's access and willingness have been an incredibly help, i've thanked him many times and will continue to do so and donate hw to help him out 8)
alright just nuked + updated PRK1 to 2.6.0 ... let's see how this goes
hehe 20 nodes in under 10 mins
Can't believe I finally made a DM contact today after 4 days of trying lol
pretty cool. esp the interesting routes
message completion % still a little shaky tho
Yes definitely shaky
gardener - i just went through the --setlat/--setlon funkiness you had, doing just --setlat --setlon seems to be what actually sticks and adding anything else along with those was triggering a reboot, and it had the location before reboot (in --export-config) but would be gone after reboot.. this is def a bug vs being able to also set fixed_position and changing channel precision in one single command which I used in earlier 2.5.1x builds
Their actually located on varick st. Unless they moved
pork you run cli tools which connect over... wifi to the mesh nodes? serial?
too bad traceroute doesn't work partially. it seems like with these new nodes we are so close to more south Brooklyn to north Brooklyn connecting path
yeah like 4?5? months ago they moved to Industry City - and still no Will Call/pickup
I use the meshtastic python cli tool to do a lot of stuff like backup/restore configs and tweak values here and there, but also to watch the logs over serial connection - I have an laptop by the window node and just VNC/ssh as needed to mess with it
you can manage most of a node over RF with adminkeys, but like i did the whole shutdown + nuke fs + install 2.6.0 via ssh and then needed to mess with it sorting out why it wouldn't take the location coords and drop precise precision on the default channel
May have an opportunity to stick a solar node up high on a bulkhead of a huge building on the UES. Would that help the mesh at all? Would it best serve set up as “client” or “repeater”? Also- how would I go about accessing it remotely without hooking it up to a cell or wifi network?
wow adafruit is at IC??
use the CLIENT role, it will act as a repeater on it's own - there's no real place for infrastructure roles in a high density mesh like NYC where clients are already repeating packets and it becomes detrimental as the infra role stymies CLIENTs from repeating. you can administer it remotely over RF/the mesh as long as it's reachable, and the remote node has the publickey from your usual node that you'd administer from added to it's adminKeys
Is "N1UGK Portable" in here?
don't think so... maybe in the NJ specific channel?
Maybe just wondering cuz I just saw the node pop up.
I have a mini pc coming in where I'm gonna run some VMs in them. maybe one of them can become a dedicated meshtastic VM or something. Plug one of my heltecs into it or something.
haven't really looked into it much, but I've bought some hardware for another project soooo.
Or maybe host a private MQTT server, and connect gardener2549's nodes to it? That way we can broadcast PSAs to use the CLIENT family of roles throughout the majority of the city, simultaneously.
Hmmm.. interesting...
Ideas, Topics, and Experiences in the Amateur Radio Hobby
yea, I snooped that already.
checked their qrz page and they link to that
N1UGK also has a base setup as ROUTER.
Or other useful information broadcasts, but that was my first idea. If we do that, I think it would be best with a hop count of zero, given that there's already around a half dozen nodes that would theoretically broadcast it at once.
yeah zero hop is probbaly best for multi node broadcasts
but.. no reason for a simulcast.
Its probably better for a few nodes or all nodes to send the message at different time slots.
if you let the mesh "take care of it" you'll get multiple broadcasts out.. probably use less channel utilitzation too?
rather than having nodes all over the city blast as message at one time?
assuming the nodes will be in "key location" too.. so they will have a pretty load broadcast. I guess it's just an interruption for a short time?
will probably need to think about that a bit.. but it's not a bad idea either.
IMO, there's too many ROUTER-like nodes that'll attempt to rebroadcast at once and cause a major collision.
Ah.. and I guess the zero hop will prevent that.. not a bad idea.
Anyone want to go to https://n1ugk.com/contact-me/, and send them this link: https://meshtastic.org/docs/configuration/tips/#avoid-routerand-repeater
Should we invite them to the discord since their node is reaching into NYC?
could mean one less node in ROUTER mode when it should not be. or maybe it should be.
In that case, as long as its on and powered, does it need to be connected to meshtastic via bluetooth, wifi or cell service? Or once set up- will it operate on its own?
It will contribute to the Meshtastic network on its own, assuming it's powered.
if nobody else is gungho I'll contact them. maybe we would get them over here more quickly to chat if I use my call sign on the form. shrug?
I would, but I’m out of town at the moment. As an aside, it seems they have a few nodes.
let's just invite them no matter what?
Interesting. He's in deep. haha
It’s a slippery slope. 😂
heh I have like 5 900 ISM whips already and I joined 1 month or something?
I invited them with their contact form
btw my AliExpress 915mhz is working great down here at my gfs in S Slope BK. can easily hit cdef from 10+ blocks. really think there are some serious RF blocking properties to some windows (my apt, work).
Muzi whip for the win 😅
Where's the best spot in Midtown Manhattan for access to the wider Meshtastic network, specifically outdoors? Last time I went to Bryant Park, I saw no nodes.
not sure, there's so much concrete and steel around just messing with RF
like in theory 🦒 would be amazing pointed north... for CLIENTs to RX but they'd still have probs TX
I also think N1UGK is far out enough that ROUTER is less detrimental to the city other than being an occassional relay for the 7 hoppers in North PA, but I don't really know what else is going on in Jersey aside from the random packets that make it to me in Brooklyn
Being on the Jersey side, this is what I’ve seen from the Jersey side.
I’m in the cluster under RabR
Granted, I drive all over the dang place.
hah! i've yet to see nodes from that part of Jersey but I also have everything that pops up on Staten Island ignored between the jerks like DA64, random ATAK users and everyone set to 7 hops
There's a location with a view to central park that would be cool to install
Blocking DA64 seems to have improved things a ton for me.
feeb was in SI a month or so ago nearbye and investigated the node, seemed to be spamming every like 5 secs which far as I know is something you'd need to modify the firmware to be able to do
Right. So they suck, but they’re clever…
good evening fellow meshtastic enthusiasts
hey hey, glad you could join us!
How high up is that node?
If it's not super high, I'll bet you don't see a difference in performance in the area and maybe see an improvement.
im up 350 feet
Above sea level or ground level?
That's a good height. Near the water?
Above sea level, not near water, but views of NYC, SI and BKLN
Def see your nodes here in Queens. Maybe not always direct but I see them.
You're making a trip.
Yeah seen them here in Williamsburg as well - and ty for being willing to change the role to CLIENT!
current node list (RF only no MQTT)
let me know if that helps, or if you want me to try switching it back. I believe it was set to “router & client”
If you're on firmware version 2.3.15 or newer, I believe ROUTER_CLIENT does the same thing as CLIENT.
I don't think it was impacting us too much here in the city but we also have multiple ROUTER_CLIENTs and hidden REPEATER or two in Brooklyn making things funky since they suppress the CLIENTs
thats my understanding as well
there's also the new ROUTER_LATE role you may be interested in, it's basically a CLIENT but will also ensure retransmission AFTER the normal CLIENT retransmission window
Interesting, that could be helpful
yeah, it sounds like you have a pretty great location if it's 350ft up and could still have infra use
I could try router late, and switch it off if it causes issues
to me personally - it's 100% up to you, your equipment and shared* public airwaves but does seem like a better fit to allow CLIENTs to do their thing but also still guarantee retransmissions
Make sure you're on the latest beta firmware or newer, as older firmware versions either have a bug with ROUTER_LATE or the role wasn't implemented yet.
That node is 2.5.18
Yeah, you'll have to update it, if you want to try ROUTER_LATE.
Also the 2.5.18 is before they fixed the storage issue.
So it may be a good idea to wipe that node and put 2.5.2x on.
After backing up the config.
i have cheap heltec node that very often is in the reboot loop. I had to roll back the firmware to the previous release
Ahh I have a bunch of the DIY builds and many had been set aside from LFS corruption making misbehave but behaving themselves on 2.5.20+ after nuking and fresh new firmware and restoring the cfg (the private/public keypair being the most important)
i usually fully reset them after upgrades and just set everything back directly rather than backup/restore config
whatever works for you!
RAKs are the way to go for set it and forget it nodes it seems.
The RAK has been good so far, and the tbeam. I have a station G2 which I have yet to set up
T-Beams (the non-supremes) have issues with newest Semtech radios the LR1110 (IIRC, still not 100% on the exact #) on LongFast, basically can't hear each other but as long as there's an SX1262 client in the mix they can hear the packet
do you know KD2WQI @radiant cairn
We’ve exchanged an email or two
I see his node sometimes also. First time I saw his node was when he was working or riding the 5 boro bike tour?
His node said bike tour or something.
I did that tour for about 15 years in a row, from 2000 to 2015
Moving bike tour mesh this year 🙏
Anyone know KD2WWU?
never heard that call before
mine also starts with KD2
i need to put my node back outside. gps died and i took it inside to troubleshoot but just got distracted by other stuff
a solar rak
It’s freezing cold out there tonight. My rak is solar also. How do you update firmware without wrangling a usbc cable in there? I’ve never been able to do it via Bluetooth
supposedly you can do it with the Nordic Connect app, trick I see people use is set the # of packets/sec to 5 and then... patience (like 10mins)
Thanks, I’ll give that a go.
I'm not clear on how you get the remote node into DFU mode for the fw update but I would guess something like remote admin and --enter-dfu would do it
so, basically the cli client to pass that - but again, I'm not 100% on that part of it, I just cable because it's within reach
I've tried the nodic app
It's not even consistent with the node sitting next to me
I don't trust it at all.
It worked for me ONCE
So it "works"
But failed to work in the other 3 attempts
25 percent success rate sucks.
Yes; the only issue for me on that RAK is taking the weatherproof case apart, then lifting the top layer board a little to squeak the usb c cable in there. It’s a PIA.
Yea I have the same issue with my node that's on top of a building. I decided to just build a new one and swap them.
that's better than our record talking over RF 😄
Not gonna bother with firmware update etc up there.
that sucks if it's not realiable, i'll try with the RAK Board 1 this weekend and see how it works for me (in an extremely 2.4GHz hostile env)
this is hardly a solution but I open the window and stick myself out haflway and grab the node off the fire escape railing and bring it in, open it up to warm and dry out, and then update
I use 3M command 'velcro' (its actually something similar) to attach the node tot he railing and then i just have the solar panel laying flat on the fire escape floor
hardly ideal but it works and takes up minimal space
I'll have to take some photos now that I got a new camera. been real busy lately and I'm just starting to catch up on my hobbies again 😄
I'm in fort hamilton btw (the neighborhood)
Whatever works - as long as it doesn’t fall off or a pigeon decides to grab it for lunch
Or poop all over the panel.
@proven grove saw your question in #1302945239689990185 for a node's firmware version and owner/short... --device-metadata for the former (though --get owner / --get user.owner don't work)
I already wrote a function for it lol, but if they change the --info format I gotta update the function.
¯_(ツ)_/¯
Haha k, the device-metadata for firmware I only noticed recently but added to my backup scripts comments out to not break the yaml
welp i got ~50% thru OTA update and then got distracted and screen timed out and then it just got wedged until I yanked power, trying again now
it's neat how the app puts it into DFU mode, little funky to d/l the meshtastic zip and then get the -ota.zip and get that on the phone but also probably extra iOS fun
7m35s from 1ft away @1.5kbps
ooo even though it aborted and I had to recopy 2.5.20 on it still kept the settings cause they're on LFS so 2.5.22 OTA update just picked those up, though seem to be missing serial connection after it autorebooted
and serial back after reboot.. Ray did you also lower the Number of Packets -> 5?
or hrmm, I guess Android uses nRF Connect App but iOS uses the nRF DFU app - that's another variable
I am using nRF DFU on android to do my upgrades and it works. IIRC you have to disable Request high MTU in the settings, and make sure you disconnect from the device in the meshtastic app and nRF DFU works.
You know you can just dump the uf2 to the node via USB-C from the files app, right?
Yes, but in a weatherproof box, it’s a PIA to remove the cover and then lift the board up to get a usb c cable in there each time.
Didn't realize it was in an enclosure. And up until ~3 months sago, I didn't know it was possible, so just sharing that.
Miss you guys
nice part of Mexico tho!
yeah I hadn't even considered leveraging Files and mounting directly, though at that point I'd likely use one of my raspi powered handhelds and have meshtastic cli in the mix as well - mostly wanted to try OTA update since hadn't before
find the highest point you can and...yeah you're still not reaching us without MQTT
very good to know! the meshtastic docs mentioned DFU on Android was no-go past a certain version and to use nRF Connect instead
What hops are most people at? Been chilling at 4.
hey welcome! i've been running 5 lately trying to get bit more reliable transmission from Williamsburg -> Astoria, before was 3 for window CLIENT and 4 for CLIENT_MUTE on the desk node
my node is back online
just tweaking the config and it's going back on the railing outside
I run 4 on my muted clients I expect to have a local hop with first and 3 on stationary clients.
I'd up it but congestion is so bad now with the mesh more connected.
Even for a local message it can take almost a dozen seconds to find a moment of free airtime.
But if I start seeing 7 hop counts again I'm going back to aggressive blocks. Super unnecessary when it's hard enough to even get 1 short message out.
be mindful of which 7 hoppers you're ignoring, I noticed SouthSlope cdef and Prospect a031 seem to be set to 7 right now but they're only sending only telemetry (and likely that high so Gardener can admin them over RF?)
Yeah I set the new remote nodes to 7, vernon grand and lilybx are also at 7.
But they only send telemetry every 3 hrs or so
I might have set a couple to one hour when i was having trouble with their positions. I'll take a look at them and make sure they're all only every three hours.
What was the idea of linking them as an alert system?
Something like notifyNYC?
My idea for the use-case was to persuade people to switch their nodes to either CLIENT_MUTE for handheld/portable nodes, or CLIENT for stationary nodes. It would be good to send them as zero-hop messages, so all the ROUTER(_CLIENT)s that'll likely be in range won't rebroadcast simultaneously and cause a major packet collision.
Someone else (I forget who) mentioned weather reports.
Yeah. I don't block any at the moment. I did back when SI nodes were spammy but that seems slightly better. I just don't think nodes need to set their own hop high if they aren't sending anything but telemetry. Their own hop count won't matter and it's fine if traces don't have node info sometimes.
Big agree! Just wanted to be sure you didn’t inadvertently block those well placed infra-like CLIENTs since they have a legit reason to 7 hop for admin over RF.
Telemetry and position is 2s of every hour, if they detect being below the usual 25% CLIENT Tx threshold - that’s why I kinda Ok with Gobo and Harlem Relay being ROUTER_CLIENT on older fw, it’s just no really spammy and far as Ive seen were around before OOP and Brooklyn Solar Node which showed start of 2025
I switched my two nodes to CLIENT_MUTE. I think it's too saturated where I am. Not helping much rebroadcasting
On CLIENT, it won't rebroadcast if another node has done so already. There's a random delay in which it first listens for other rebroadcasts, and then chooses whether or not to rebroadcast at that moment. I believe the delay is skewed in a certain direction depending on the initial broadcast's SNR.
yeah but there's also all those big rf blocking steel and concrete structures
late to the party - nice! I see the femtofox there, what are the other boards?
hah, other ones are Ikoka Nano v3 https://github.com/ndoo/ikoka-nano-meshtastic-device
E22-900M30S 1W module + XIAO nRF52840 on double sided PCB, v2 was simpler design of just those two parts + capacitor but powered the E22 directly off battery so didn't have full 5V/650mA the E22 will max out but still stronger (and more power hungry) than usual SX1262 radios
a.li3n mentioned the v3 might still not be getting enough power due to the LTO but I haven't tried to measure/confirm, have few other 1W designs (like the Femtofox and a Donktastic) as much as I adore the Ikoka's size/efficiency
Washtastic is the only one I'm seeing consistent power delivery from. I was redesigning a variant of the Xiao 1W, but since there's so many out there right now I put it on the back burner in favor of a portable LR1121 eink.
I haven't tried donktastic and I'm still not even sure I have all the components for the babelfox because Fire's BOM is vague AF, so im gonna have to import it EasyEDA and pick out all the components myself to make sure of it. (Damn Kicad users)
nice I'll check all these out
@radiant cairn KD2WQI here, great to see you in the channel
I'm headed to the NYC Resistor radio club meeting this wed if anyone else is going, first time for me - seems like a great bunch
caught up now on the channel - I gotta keep a closer eye on this. relieved to see my node isn't causing active problems
ooo those both sound dope! i'm getting consistent power draw on the Donktastic with Adafruit 5V boost board (had an INA219 inline between it and the E22 and it's def hitting 5V/650mA) but Femto just has so much more going on than being a "simple" node
theres also the #1194757507013427250 rabbit hole
and ... which node is yours? we've def been encouraging people to run CLIENT instead of ROUTER/ROUTER_CLIENT/REPEATER and there's def some benefits to running on new firmware starting at 2.5.20 as there was some LFS corruption in earlier 2.5.x which... wasn't fun to sort out
I'm KD2WQI TRACK-1 / 54e1, should be on CLIENT_MUTE - will check firmware version tonight but should be pretty fresh! I have a T-ECHO I sometimes take portable that is also CLIENT_MUTE but 98% powered off
oh ahhh!! I'm the pork - something / PRK# person with too many nodes 😄
I just started leaving mine as the default names. Saves me from judgement lol
Seeing you in NJ, South of SI
anyone got a suggestion/preference for that tape you use to wrap antennas etc for outdoor use
You did he 5 boro last year with the meshtastic device right? How did it do? Were you just riding around to see ho wany nodes you could pick up? Or were you using it for comms?
finally updating my tbeam 1-1
sheesh one wet though no problem now the other two are killning me
fighting you getting into bootloader or just being funky in general?
I thought the whole point of having a fiberglass antenna (like this one: https://store.rokland.com/collections/all-helium-antennnas/products/low-profile-6-dbi-n-female-omni-outdoor-915-mhz-antenna-for-helium-rak-miner-2-nebra-indoor-syncrob-it-bobcat) is that the fiberglass is the environmental protection. Please correct me if I'm wrong.
hmmm, maybe he means silicone sealing tape for the connector?
first one went smooth and easy
second one hung up a few times, then finally got in, but was stuck at 10% for like 30 minutes
Now seems totally bricked.
third one cant get a CLI either.
@toxic warren picked ya up in Jersey yesterday
On a side off shoot, my t-beam1-1 that got flashed can see my sensecap
sent a message to it direct message
No way! Where in NJ? And which one of my nodes?
got the cloud with the checkmark
but.. don't see it in the sensecap
esptool.py v4.8.1
Found 1 serial ports
Serial port COM4
COM4 failed to connect: Could not open COM4, the port is busy or doesn't exist.
(Cannot configure port, something went wrong. Original message: PermissionError(13, 'A device attached to the system is not functioning.', None, 31))
A fatal error occurred: Could not connect to an Espressif device on any of the 1 available serial ports.
Wow I can see so many more nodes than the last time I checked
I got that same one in Laurence Harbor, NJ (a particular corner of old bridge)
Thanks for the info. That is my T-Echo. That's quite the reach. I wonder who it hopped through to get to you guys
Good thing we have discord or we'd never know who gets our messages 😂
ugh, yeah i've had "fun" getting my esp32 devices into bootloader mode, and yet to find a reliable method. the t-beam 1.1 was usually one or two tries, last night t-deck fought me for like 30mins doing the same procedure that worked perfect previously for a local build before I gave up and just used the web flasher's 1200bps reset + flash which "just worked"
I finally got it to flash.. that was painful
it reminds me of abbeyway's products
its still flashing
Im doing a full erase
these are identical hardwares
why one works and the other two gives me incredibably hard times
🤷
should be ok if it made it that far, otherwise the usb cable / port would be highly suspect
its usually just getting into the right bootloader mode that's the pain but once its good and esp32tool talks to it should work, or should say never had it not work at point
i'd give it a few mins but worst case: erase + flash again
I mean saying "it can take a while..." is like... teeth biting
back to the sensecap- so the tbeam can see it
can also try a diff cable / usb port / machine / browser - sometimes it really is something so silly
and send stuf to it
and it gets confirmation
hopefully the update ont he sensecap will fix it
as long as there's an SX1262 inbetween, as i understand Semtech basically said "meh" to fixing the LR1110<->SX127x on LongFast because the latter is so old and basically that low level of a bug
@still fjord are you E53C?
I'm making a "bird feeder" to stick to my exterior window
nope, 519d
that accidentally will have a solar node
oh dope! you've seen the Austin Mesh "Birdhouse" build I take it?
that's why I might want to tape up my solar
no????
Am I on your node list?
I just saw a neighbor with an actual external bird feeder and was like oh wait a minute
perfect way to get around my faraday window
yeah https://www.austinmesh.org/devices/#solar down at the bottom
@toxic warren yes and no. no hardware info or long name. but 84ce is on my node list
also someone check on boneyard . i think he left his node in the oven
never exchanged keys i forgot to mention
I'm out in S slope doing errands if anyone wants to try to ping zlla1
Cant you just tap on someone's name and tap Direct Message?
hhha ping to cdef is like 0.2 second return
Does that create an encryption key?
acked by different node
I think I'm searching the nodeDB incorrectly. What do I search for your node?
pretty sure it was just a lucky message hopping the right way cuz i dont have any of ur node info
not even sure if my hops are turned up enough to make it to you
not sure what you mean
For example, this is my node's info:
It shows all the names. So I can better search for yours on my nodeDB
It's confusing that the node gets basically 4 names.
- the initials (short name?)
- the long name
- the user ID
- the last 4 digits of user ID
haha it's a temp sensor inside a case inside but yeah I noticed the sensor started going crazy couple days ago
And to make matters worse the way you search for someone on iOS is different than on Android 🙄
Yes! Rode the 5boro, tossed my current main node into a fanny pack. I was just seeing how it did and seeing what nodes were out there - that was right when I was first testing meshtastic in the city. Node did great despite getting soaked so it was a good general stress test.
@jaunty harness which one are you?
Both 😃
I dropped my gf off cuz she did the ride. I picked up your node first when I dropped her off in the LES. I continued to get your node a few times as you came into Brooklyn and left Queens.
Sent a DM to pork femtofox
I use silicone self fusing tape around radio connectors, generously wound, and then cover that with a few layers of electrical tape as the silicone stuff isn't particularly UV resistant. Looks like Rokland sells rolls of silicone sealing tape, you can get it on amazon too. I don't have a ton of outdoor infra deployed but can say that this has kept my setups healthy
I used silicone tape on the antenna connections of my outdoor nodes also. I didn't know that fact about the UV. I'll add electrical tape on future builds.
@jaunty harness also sent to D446. Got an Cloud with checkmark for Pork Femtofox and a "Max Retransmission Reached" for D446.
How copy?
definitely fact check me on that but that's what I was led to believe by other radio people - can't hurt! I imagine heat shrink would do even better
Cant hurt.
sorry, fox is a bit wonkier to check - no ack on either 😢
sent a DM, fox last saw CGM 14hrs ago but PRK1 saw it like 5mins ago
Alright shoot. Well I just updated both node names to make it easier to recognize
Sorry for the cheese
looking for you in the node list ...
All things considered, there's a lot of sx127x units out there
Its interesting because I see the sensecap joining the meshtastic on the sx127x screen
ok, here's an even WEIRDER one
I have an un updated one
and it can see messages from the sensecap
its receives it but its marked as cloud with a line
sensecap (I cant tell if its receiving or not) as it doesn't have a screen). and it doesnt show in app
now that I think of it... this could be why I also don't see much propegation of my home nodes. THeir all sx127x based. Either the t-beam or the t-lora
@abstract iron DMed you from Spock 🖖
It’s certainly not helping but should only be the LR1110 and SX1276 communicating, maybe not getting heard by a SX1262 to propagate
got it, it's still called CGM, I replied, did you get it?
my link to cdef seems shakier today
Not yet
Delete the message and let's make a new one. Could be a conflict from the name change
woah wtf.
my other nodes shwoed up
my techo in the car and my tlora v2-1-1p6
though my theory is now that it was tlora->t-echo-sensecap
because I dont see my other node.s. lol
Ok resent
I responded
I see ya there did you get my responses?
also got Rx over in NJ!
check out that union city node
makes sense if it was 1 or few hops I was close to water
We don't have a good sent to received ratio 😂 But seriously why does it say it went through with the check mark if it didn't ?
Sure will now
Sent several from both my nodes. No response.
I asked about the check mark thing in the help section. I mean who would think a check mark would mean the message was delivered? 🤷♂️
Ok so this is what I learned: (for Android)
- only DMs can be fully acknowledged by end user
- when fully acknowledged by end user you'll see a person with a checkmark
- the cloud with checkmark only means the message was received by another use in the mesh
I thought the person check was automatic by the final destination node tho
so therefore if you get cloud check it just means your message went out but maybe not the final desired recipient
I think I just got a "test" to you
I saw your test and responded
I dunno but that's what they told me in the help section
@abstract iron I'm getting your test
nice
now can someone get a node on governors island
and tie us all together over the water
my node's been back outside for 24hr now
it's heard 88 nodes in just a day
nice, that's more than 4x what i could see about a year ago
Gotta make a trash node. Where it just floats around. lol
the only issue is that it doesn't help much given the event horizon
Is that when you get sucked into a black hole? 😐
You don't love my humor? 😁
I'm not seeing the starfleet mesh on any of the hailing frequencies up here in astoria queens.
you might have their old names and haven't heard nodeinfo since the update?
Possible.
yeah i still have the old name on the window node and haven't seen > 1hr, same for RSR1 / RSG2 ☹️
that one is on 2.6.0alpha but in theory that should only affect DM routing, and in 2 more days it'll be 2/2.6 #alphanauts-testing message
I rebooted both and changed them back to client. Maybe that helps a bit
We gotta work on our S BK to N BK links!
on foot in S Slope
you want the window as CLIENT and the other as CLIENT_MUTE otherwise they'll both retransmit everything they hear and will end up creating collisions/interference with each other before you get to the rest of the mesh
https://meshtastic.org/blog/choosing-the-right-device-role/ explains the roles really well
I was just doing this temporarily to see if it helped other nodes update for my new naming scheme. I'm just throwing poo at the wall at this point. I understand your point. I usually keep them in this config.
ahh you can cheese the normal NODEINFO update interval a bit... double-tap of the user button will send an ad-hoc ping to the mesh which should include your node's current name/short but kinda rude to do more than once or twice. other way is just reboot it, which should also send out the current NODEINFO shortly after it boots
but its set at a 3hr interval because it's somewhat needless to repeatedly send the NODEINFO other than aesthetics (e.g. seeing pork d446 /prk1 instead of !865ad446) and should naturally resolve itself over time... which is kinda frustrating but also the way Meshtastic works to not just be constantly packets over the shared airwaves
what is the user button?
The main button on most Meshtastic nodes, excluding the reset button.
lovely. Thanks math
My T-Echo has the user switch. I will need to add one to the WisBlock RAK19007 base board.
yawp, that's right for 19007 - just be sure to also configure the button for pin 31 in Meshtastic app (pretty sure that's mentioned on the same RAK page for adding the button that image is from... or at least it sure looks like the one I remember in their docs)
USER button is also useful if your node has a GPS, you can triple-click to toggle the GPS on/off to save some battery
ohhhhh interesting
Some models work better than others to full cut off power vs more a “sleep” lole behavior where it just doesn’t (re)poll for sats/position but both help save batt
Anyone have any consensus for the heltec 114m
v1 had some show-stopper issues w/Meshtastic but v2 is fine far as I'm aware (and looks like a nice replacement for the Heltec v3 as a go-to recommendable node)
I'm a little torn at the moment as I have 6 1276 enabled devices that now can't talk to my 4 sensecaps
cheapest option would probably be the Seeed Xiao ESP32 + Wio SX1262 for $10 as that can talk to all of them but not exactly a "fun" device since it's just basically a bluetooth device with no screen or extra sensors
Too bad there isn't a way to slap on a different chip onto the t beams
Save the rest of the hardware
ahhh i hadn't seen there was some newer discussion in there... honestly my t-beam basically fried its radio over time which is why I went back to RAK in the window, was seeing ~50% airtime when reality was more like 15-20%
Ya. I mean. But fark. 200 bucks of hardware that can't be used kinda sucks
Unless I just go full privaye
So... It's a sync word thing
yeah i didn't realize that was the crux of the issue, seems like something that might get fixed in 3.x but also wouldn't want to bet on it
If they don't fix it .. is there anything useful for these devices?
Like can we pull the radio off? Swap over the 1262?
if the pinout and package are the same in theory, yeah - in reality... I'm not 100% but def have a t-beam w/bad radio to try on
is there an alternative the Meshtastic GUI? I want to keep one stationary node connected but not on a phone. I tried the web GUI but it doesn't save messages with the Wisblock RAK4631. Bummer.
I tried to run it on a iphone but it's quite frustrating as Android is miles ahead
the node will store the last... 20? or 30? messages (pretty sure its 30) until a "client" app connects, can be Meshtastic.app or Meshsense or a cli one like contact ~ https://github.com/pdxlocations/contact
N1UGK car I RX you over in Williamsburg but can't reply here
it stores zero unless its a ESP32 device
I
I've been trying to get the above to work for 2 hours
I'm not sure where you picked that up but it's 100% also applicable to nRF52840 devices as I've done it repeatedly. maybe you're confusing the diff between ESP32 and nRF52840 around NodeDBs where the former will hold 100 entries in the device's NodeDB while the latter is only 80?
haven't seen contact get stuck on that screen, but only use it on the femtofox where meshtasticd is running as local service - does seem like something with python meshtastic talking to the device, confirmed that's working?
i gave up on the Contact. Not worth 4 hours just to get that working. Bluetooth passthrough on LXC containers are a pain
I used the Mesastic web GUI all day and the messaes kept dissapearing. Not really more I can say but that. They dissapeat when refreshing the browser or if the bluetooth disconnects
this is where i read about only ESP32 devices storing the messages:
This module allows you to resend text messages after a device has been temporarily not in LoRa range of the mesh.
Before browser refresh:
After browser refresh:
😦
Ahhhh! store and forward is a module, it def has the limitation you’re thinking of but now how normal client : app messages are exchanged… and it’s not just LXC, passing native host hardware to a container or VM almost always starts getting funky real quick
Far as the web client goes .. I haven’t touched it since 2.4.x but in general found to be functional and somewhat awkward vs the iOS app
Solution: get an android tablet. Problem solved
haha, whatever works!
Wow, nice. Was driving around with the 1 watt amplifier connected. Thx for the report
Portable is just an extender , client_mute, it’s usually off. UGK2 is my new node, replaces the QTH node
If you can hit NERO then it should work
See you here and there from
Chatham but not able to contact you usually
Ridges on all sides limit my transmit I think
I have the rest of the hill to my west, I can’t get anywhere in that direction
Sometimes I see you from the linden Rahway area as well but again not in any manner which I can reach you back with my Techo and whip antenna or T1000
I didn’t get anywhere without the increased power, other than with my own nodes. I heard many more nodes that could not hear mine.
still the amp on your site?
Yes, moved that same amp to the car
nice. what antenna you using on the RP-SMA port? or did you adapt it back
The antenna is a cheap mag mount (on the roof of the car), that runs down into the output of the amp with an sma male connector. The amp has an sma female output connection. On the input side of the amp, same, sma female, so I have a very short jumper male to male to connect the T-beam to the input of the amp.
ok the more I look at this t -beam
the pin outs are the same
as far as I can tell- comparing pictures of the tbeam 1.1 and the tbeam sx1262... they appear identical in every aspect- except the GPS receiver.(and of course the lora chip itself).
mmm not sure, the footprint looks right but it doesn't match the pinouts of any modules I'm familiar with (waveshare core, wio1262, ra01sh/htra62... maybe something else I'm forgetting)
hrmm, yeah i'd.. just prefer to actually pin out what's on the board cause it could be something else under the ttgo sticker
it def seems likely
the GNDs line up
(with my v1.1)
i'm yawning but leaving this on the bench for tomorrow to see if i can match anything else up
I just picked up some new ground coffee- might have "fnished" up my ethopian blend...
and now... a little wired.
haha yeah i knew if i didn't just put the tbeam down and go to sleep i'd have been up another hour or two poking at it... its much easier to see PCB traces when I'm awake but have a bit more work before i can poke at it
Sure looking like the LoRa32 will drop in but need to pop the OLED off and get at the esp32 to verify the SPI pins since they’re not broken out
Received “antenna B” test in Rahway/Linden but couldn’t not reply with a TEcho
yeah saw same - though 1 hop to UGK2 / 2 hops to UGKP
Did anyone get my airplane message?
If your NodeID (not long nor short name) is !15436b83, then yes. The node is no longer in my NodeDB.
Oakland garden solar is all by its lonesome. Its also my closest link
It is one thing to throw away 40 dollars of equipment.. its another thing to spend 10 dollars to upgrade (+ time/energy/skil etc)
hah I'm already way over $40 worth of my time/energy but if I'm doing a thing... I wanna do it right - and with the confidence of the pin outs matching I'd move onto slapping it on the hot plate and see if I can get it off without messing up the rest of the board at which point it's just a matter of getting a LoRa32 chip to solder in place - and if that works, it'll be an option for more than just me and you to keep a t-beam useable
and if it's not... well at least know "yeah the pinout doesn't match so not an option w/o jumper wire spaghetti"
Nope I'm 3270779875
I think I would need to update a soldering station
But part of me also wonders why nobody has thought of this
Or worse case.. they did think of it .. and we are going down the same path
#1300525496945348648 has a good soldering station.
I know someone figured out how to modify the Waveshare Pi Hat to be TCXO instead of XTAL and that's an even narrower niche.
I'm more of a hammer and drill guy 🤣
Not sure if it’s weather conditions or what but I see more of the mesh tonight than my normal “almost nothing”
I gotta say it's pretty wild how much my node could "hear" inside of a bag under my chair in an airplane
I was picking up stuff all over the flight path to Cancun
HAH! Well my dinky hotplate maxes out at 260C which wasn't enough to free the radio... but it was enough to lift the sticker revealing the ACTUAL part - HPDTEK HPD13a v1.1 which matches the LoRa-C1 v1.0 pins far as I can tell
so... we have determined a viable SX1262 replacement... my T-Beam however... lost a pad for the battery terminal and I had to munch the OLED to get it out of the way (I have spare 0.98" OLEDs though I think I'm out of VCC/GND/SCL/SDA ones and just have GND/VCC/SCL/SDA at hand)
HAHA $8 off ali... which is basically what a 1W E22-900M30S costs
hrmm, where did you find the pinout/image for LoRa32-C1 v1.0 ? i'm only finding another SX127x chip by that name
Let me looksie... Was very caffeinated when I went that
amazon. lol... I'll find anotherone for you. Finding stuff online is about my only skill... thats what led up to keithmon making those LTO BMS boards.. lol
Yes, but for the price of the new chip, you may have an actual functional device- which makes it less ewaste
on a very very odd thing, the lilygo lora32's are... somehow working with the Seedstudio
not.. perfectly..
they are the t-lora-v2-1-1p6
but still something incredibly wrong. the t-echo gets more (sitting on the same desk location)
like I haven't picked up a SINGLE other node for the last 24 hours
Ahh but would you replace with the same make/model 10? yr old problematic SX1276 … or an SX1262? Def with you on preventing ewaste but I thought we were trying to update the radio chip 😃
I think a tiny PCB with a bare Wio 1262 to break the pins out to match the tbeam PCB would be the move, with enough kapton tape to cover everything else a heat gun could probably get the existing one off
Far as the nodes T1000E seems to be pretty amazing RX from what I’ve heard and the little time I’ve played with myne. If the T-Echo seems to have the best RX try that as the “base” CLIENT and everything else CLIENTMUTE since that’s SX1268 and should work with both new LR1110 and the old SX1278
OH I saw this earlier... getting closer and closer to Great Neck everyday
Me personally? Eh... I dunno. Sometimes you chalk it up to hobby money. Kinda like this instead of going out drinking. So from that standpoint, why spend the effort. The flip side of my thought is, software should fix this issue, because it was a software design issue. And then also a hardware design issue. Replacing the chip works for the few of us willing to do it. It does nothing for the thousands and thousands of nodes in the wild, that we now have communication issues with
Like if the philosophy is to be able to speak to strangers nearby or bounce off their device...
Granted.. if there was no valid software that makes no sense either
Yeah, one of many reasons I like Meshtastic is the low bar to entry - cheap radios and no licenses + process of pair BT, set region and you’re meshing. Neither hardware or software is perfect, let alone the anarchy of mesh routing with random hw and sw versions, but it does work, mostly and gets better all the time because of the awesome community. Is it worth coming up with a potential ~$10 upgrade for tbeams that only a handful of people, if that, might do? Idk, but I’m still curious enough and have a basically useless tbeam to sacrifice.
Ideally 3.x might go back to the standard sync-word, and then T-Beams in theory will magically just work again, which is a software fix and way more useful to people, but also if we can figure out a hw solution - well that's even more of a solution (for the few people that do it)
btw found this for guerilla nodes while shopping for clear double sided for my bird feeder
"attaches to rough surfaces" BQE? heh
haha bqe would probably shake it off - maybe magnets?
https://meshtastic.org/docs/hardware/devices/seeed-studio/sensecap/card-tracker/ has magnets built-in. Someone'll just have to swap them for a charged one every other day. They could charge multiple at a time using this: https://www.seeedstudio.com/Card-Tracker-T1000-E-Charger-Accessory-p-6263.html
This accessory is specifically designed for charging T1000-E devices and can charge up to six T1000-E units simultaneously. It features a PCBA board and includes six magnetic cables. Users can customize the casing based on the provided 3D shell file, allowing flexibility to meet various usage scenarios.
i just checked, the T1000E* magnets aren't strong enough to stay attached to metal workbench... it holds for like .5s and then drops
krazy glue a >40lbs magnet to a holster style case and it'd work though
I agree with the low bar to entry but they didn't tell me there is no exit in sight
Facebook Marketplace is a potential exit.
If the paint holds up.. lol. Actually if you mount it to the bqe and just stick use a yellow paint marker and write "TBTA capital project #834521"
Then nobody will bother it
mmmm sexy
In a fit of even more crazy, the lora32 with the older firmware seems to transmit and see the sensecaps just fine... It doesn't give hop info, but it could a version problem. But they are both 1672 chips
which version of 2.6.0 was that? apparently one of the releases had to be hot-fixed as it was swapping the public/private keys 😬
that is def weird, but with 2.6 dropping today it's a good excuse to upgrade and get them all on same version (if you got the time for it, and with a cfg backup before you start)
Just regular 2.6.0 I believe
2.6.03f51297 was one of the bad ones (I d/l'd but hadn't flashed anything with it) 2.6.0.f5db94e2.6.0.0b106d4 is the most-current version and what i've been upgrading to
oh man, the colored heart emoji gang are taking over williamsburg (i like it!)
Honest question here: has anyone here actually successfully added anyone via direct message and been able to regularly text back and forth? Other than to your own second node?
not reliably but I hop neighborhoods so much and need to deploy my bird feeder in Williamsburg
who is this??
nice location
Based on their reply to https://www.reddit.com/r/meshtastic/comments/1iyycvw/comment/meygbjn/, and their Reddit username, I'd guess @balmy osprey…
According to this Reddit thread, it's NickMNYC on Reddit: https://www.reddit.com/r/meshtastic/comments/1iyycvw/comment/meypqow/
It's me, yes 😉 Oakland Gardens... Building a 2nd one as we speak. BTW this is the best software to use to determine coverage. I'm trying to talk to my parents who are in Flushing
See posts, photos and more on Facebook.
Welcome.
Many nodes were recently put in place by someone with access to NYC Mesh's ISP infrastructure. I'm curious if your node(s) can reach any of them.
Thanks... What's your node? And a question.. how do I see a node in... Iowa? (I have MQTT off)
I just added another 2 foot extension today, but no way it's through air only
Either there's an airplane up really high, or you forgot to enable Ignore MQTT.
mqtt it was.. thanks
My nodes contain Supermath101 in their long names.
I've DM'ed probably 60 people and only got one actually response. I'm kind of bummed that the "connectability" of the mesh is what it is. It does work great for your one or two who are within direct contact but I'm shocked how many people are on the local mesh who don't get messages.
I'm with you... My problem is, I send a good morning, I get the check box. Some people respond, I respond back, and I get the line in the cloud (which means they didn't get it) .. why did they get it the first time and not the 2nd?
I mean the amount of effort people are putting into this you'd think it actually meshed people together effectively. It makes me wonder if I'm missing something
Maybe a packet collision, or maybe interference from other 906.875 MHz users in the area. What's your channel utilization like?
I can message people here on discord who are also in my nodeDB and they don't get any of the messages
I found out that height is your friend. I can't talk to my t-echo more than a few blocks.... So I said f*ck it, I put it on my roof, so now it works a few miles reliably.. but line of sight it is
Its hard to believe it's just my radios. I haven't talked to anyone who's been able to make consistent contact.
My theory is that many misplaced ROUTER(_CLIENT)s, REPEATERs, etc. are likely causing packet collisions, and causing more ideally placed CLIENTs to inhibit rebroadcasting messages themselves.
Yeah I can message myself from my T-echo to my WisBlock from a decent distance. But I'm not confident that the mesh works at all for messaging.
I also strongly believe it's line of sight... I was able to get a text to my house from a few miles away... then I moved closer and I couldn't.. but the elevation was lower... So I think we just need more nodes in high places
I hear great thing about Austin but are they actually able to send messages through hopping? I just don't believe it.
One way to mitigate that issue is to switch to a non-default frequency slot and/or radio preset. The other is to try ROUTER_LATE, at the cost of higher chances of packet collisions, due to higher channel utilization.
I haven't seen it but one time since 2-17 when I started
How many nodes are reachable in exactly zero hops?
in my whole list only 3.. I get a LOT of ? and maybe five 3s
there used to be a guy directly across the east river from me. i could reliably message him and have a whole conversation. i even asked him to change his node from router to client and he did it.
but my main use case is communicating close range with my friends at music festivals and it works flawlessly. This is probably closer to what meshtastic was intended for rather than texting with strangers across nyc or large distances.
i can also put a node in my shop (ground floor) about 1.5 blocks away with no line of sight and reliably message it from my apt window
yeah for small distances it works.. if I take my car and drive and test it, it doesn't... My main reason I went down the rabbit hole 🙂 is to communicate with my parents house (just in case)... I'm thinking of putting the same mast and node on top of their roof and see what I get
But would love to have a NYC working mesh and contribute
I even emailed a couple of Engineering teachers, if we can put a node on their roof...
BTW.. if you want to go east of Queens forget it.. the elevation is HIGH... there are only those 2 towers that can help us, but good luck getting the board to approve a node on their roof
I grew up east of Queens.
*north shore towers
I’m working on putting an outdoor node on my friends 35th floor balcony near Marsha P Johnson park. it faces north and can see to the horizon so might help with queens connectivity.
(Great Neck, NY)
I was in Astoria Park.. I got a message to my node (but only one)
Seven seas diner
yeah I'm talking getting to Nassau County and more east
That no longer exists (the building is a different restaurant).
yep.. I know very well
not sure i can see that far but i see arthur ashe stadium in flushing meadows from her balcony
if you can see my OGS node, we're getting somewhere lol
ok once i get that node up, ill report back here. just need to link up with the guy donating the antenna for this. i have the node already.
Haha and Supermath sending him our way - nice! Now… say hi Nick, that’s a good looking solar setup!
thanks man
Oh durp phone was stuck in scrollback - OGS reachs Williamsburg (me at least) consistantly
There’s been some talk about a node for Elechester but it’s dependent on Gardener’s schedule as he’s the one with access to NYCMesh (the wifi group) locations and about to disappear for work until … it’s cold again
Yes this is a good example of how it's useful. And I've been successful at simplex messaging between my nides.
So why use all the overhead that is required to operate as a mesh system? Currently you just can't really message your friend of loved one etc through the mesh system. It's just doesn't work. At all. At least here in NYC and with all my attempts.
Maybe it works somewhere? I'm just not understanding the hype I've seen after 10 days of zero contacts through the mesh. (Well one with Zlla ☺️)
There's a guy in Staten Island that has his Role set to ROUTER
I think initially, people have unrealistic expectations of how far you can communicate with Meshtastic (myself included).
I was humbled really quickly. Most of us live in apartments that the windows block a bunch of RF and then there’s lots of obstacles between us and who we are trying to talk to.
for e.g. I just got messages acknowledging nodes 2AB9 and 33c0, but I never got their original ack requests
i got the initial messages from 33c0
but i’m assuming you don’t see my ack from f25a
yeah.. even a car blocks the signal.. I was like how can't I see my node a few block away.. got out of my car, held my t-echo up in the air, and was able to communicate
I got f25a
I just send f25a a direct message
has a checkbox
I’ve gone on my roof and had my friend go on her roof… and we could reliably exchange messages.
Line of sight is key. No obstacles is key (including thick UV coated apartment windows).
which means you got it?
awesome! where are you located
i read you
Domino Park facing manhattan
41st floor apartment window
T-Echo with a 16.9cm Gizont antenna
on 2.6
My node is up about 100' in my window. I was able consistently message back and forth to my other node about 7 blocks away in the opposite direction that my window faces.
No line of site
Must have been bouncing
I can chat with HAQER.. awesome
where are you located?
yeah we are chatting!
oakland gardens.. exactly 11 miles line of sight
nevermind, i figured out from the name lol
You are f25a?
that’s wild cuz my node is facing west, not towards queens
I sent you a DM
yes that’s one of my T-Echoes
I just did a traceroute however, nothing.. another thing that is weird with meshtastic
don’t see it yet
it said not response.. did you get my "let's see how reliably" message?
ok.. so we can chat just fine... finally.. it's cool to confirm with this discord.. thanks @daring moth
this discord has been very helpful for testing
are you 9149?
My hops away says ? for you HAQER
but we can reliably talk, I"ll take it
@toxic warren where are you located?
Yes I sent you 2 from 9149 and 2 from 84ce
North end of Gowanus
i received Test 1 from 9149 only. also my time is off on my T-Echo so that may affect things idk
me and HAQER.. not bad.. but for sure good line of sight
that’s pretty impressive considering my node is almost facing the opposite direction
unless your antenna is directional, it doesn't matter
starfleet so you're in an apartment building? Do you have other buildings around you?
I should have line of sight to you also @toxic warren
My T-Echo time is fixed after turning off gps and rebooting. I don’t get a gps signal from my window.
i’m showing you one hop away in my node list
got your last test @balmy osprey
as I understand if we get everyone running older firmware to update it should help TRs quite a bit, when 2.6.1 is out (of alpha) we should start pushing people to update since all the LFS issues with 2.5.0-2.5.19 are avoided
ok.. I'm happy I can reach more than 11 miles away to HAQER very reliably... I have more hope now! : )
(and especially get the ROUTER_CLIENTs to regular CLIENT )
if you're both on 2.6.0 it should help a lot since they'll have a know "good" route
@normal osprey Got a tracerout to go through
yeah I'm on 2.6 on all 3 of my nodes.. but only the solar one is client.. the rest are client_mute
my window t-echo is on 2.6 so far
That was 9149
don’t have you in my node list yet
nodeinfo is sent every 3 hours by default
Yeah I’ve had that issue too. I asked on the discord they say you have to wait 🤷♂️
yeah my t-echo and T1000-E have an option to send ad-hoc nodeinfo rather than wait the 3 hours
i use that when trying to discover my own devices nearby
the ad-hoc ping is standard for anything with user button far as I know... and definitely have not inadvertantly double-tapped the cheap tact switch on pork mobile / 🙊
try double clicking the user button
if device has an *OLED screen it should show "Sent Ad-Hoc Ping" on the second to last screen in the carousel
not sure how that works on eink display though
yup my t-echo showed this (before the InkHUD in 2.6 at least)
The unit that reached you is a WisBlock. It doesn't have that ad-hoc button.
many more nodes discovered since the last photo
inkhud really got me thinking about which node to slap an eink display on
who is the Gobo solar node.. I see him too
Gardener seems to know them, was able to get Woodbine updated to 2.5.x but Gobo is still pre-2.5.x/ROUTER_CLIENT - they seem to be a good/reliable ROUTER_CLIENT but would be good to have them try ROUTER_LATE until we can figure out who the hidden REPEATER is also causing collisions over here (not to mention Brooklyn Solar Node + OOPOO30 which both popped up within a day of each other last month running ROUTER_CLIENTs on older firmware)
the real mystery is Harlem Relay - they showed up on default channel few weeks ago and Zumble tried to get them in here but not sure if they heard
designed a little mount to go outside my double hung window ledge. Will print it in the morning
hah, I usually print over night so it's ready in the AM
same but might keep an eye on the first few layers
too loud for my little apartment. The bambu x1c is loud. My Ender and CR-10s were whisper quiet but way slower
makes sense. i have an a1 mini and its in the back of my retail store so noise isn’t a concern.
that's the place to keep it. Not in a little apartment.
i like the sounds of the my printers going... lulls me asleep - even the loud af VzBot
I have 2 X1Cs now. One here and one in Raleigh. Much better to not have that droning noise all night. Wife hates it too 🙂
heh gotta keep the wife happy!
so... how about you sneak on your rooftop and put a solar node : )
i dont mind it. But the wife is VERY forgiving of my nonsense happenings at home. But after nudging me about the printers running at night I try to keep it during the day
Yeah that’s totally fair, I also acknowledge the noises aren’t for everyone, my ex HATED it but she would insist of falling asleep with the TV on which drove me crazy, one of many reasons she’s an ex
i spent a lot of effort making the Creality printers quiet. Almost couldnt hear them running. But the bambu printers dont work as well going slow. But the time savings is nice. Still love my Crealitys though !
Ooooo those are nice
My Ender got converted to a Voron Switchwire, makes more noise but it’s a fire off a print and forget it champ at PLA. Voron 2.4r2 is basically idle since the VzBot, Voron v 0.1 needs a rebuild at this point, the v0.2 I’ll use for smaller things but the Vz is the beast and just does everything but PLA/PETG amazing (never calibrated for it). Also have RatRig VMinion that’s good for PETG and not much else
always takes me forever to find that pic
I have a buddy who got heavily involved with his company's maker lab team and they were sending me kits to build / give feedback on in exchange for getting to keep em to see what they were gonna stock for their employees
I may have gotten ridiculous with the v0.2 https://www.youtube.com/shorts/0ev64qfkbQc
@balmy osprey just came across your reddit post 😂
hello gentlemen
welcome!!
watsup watsup
My node is "MESHEDUP" in the manhasset area if you guys have seen mine
i very recently got bridged over to the city the other day, either the mesh is growing rapidly or my devices only started picking things up well recently
nice! ideally it's a little bit of both
i just got through nuking + updating all my nodes so I think only one has a nodeDB >few hours old, lemme see if you're still in there
OH! I bet this was you #1202833898376138752 message
iit likely was
or well, maybe not bet but that's the furthest East I've seen from Williamsburg and we have someone in Great Neck
did you see Oakland Solar Garden / OGS at all?
f840
leme see
nope not yet
are you north or south shore? and for some reason i havent picked up anyone east of me yet
but ive even connected to a couple of people in CT across the water and 80 miles upstate
OGS would likely be the one to relay you over to the city, and he's joined us today in discord
Ahhh he definitely helped add more, but for some reason i didnt see him on the map until yesterday or the day before, but since then i've had so many nodes be detected so he's definitely helped a lot
ahh yeah there's a group in North CT that have 7 hops and have occassionally seen them here (same for a small group in N PA). I'm in Williamsburg, pork d446 / prk1
his node is pretty high up, but the area im in in manhasset is very high as well
i think i got you
nice! I'm at 4 or 5 hops currently to try and reach Zumble in Astoria more consistently but probably drop back to 3 once 2.6.0 leaves alpha and worth pushing people to upgrade
yes sir!
I offset my position slightly as everyone ends up stacked up ontop each other at that corner in williamsburg
aahhh okay hell yeah, for some reason it just shows a different name for me
yeah you've heard me send a POSITION (gps position) or TELEMETRY (temp + humid sensor, and like air time %) which are on a shorter schedule than a NODEINFO packet which has my node's name / short
if you're hearing OGS consistently should just be a matter of hours though is a narrow 1s window of my node sending it out to the mesh and then enough relaying that over to you
Ahh ok I see so if i receive that name first, it wont automatically change in my node list
this stuff is all very new to me lol, i've only very recently become interested in tech like this and electronics in general
having a blast tho idk why lol
OGS?
haha, awesome - yeah it's a weird hobby, and the running joke is we use discord to talk about how we're not able to talk over RF but we're friendly and just want things to work better so ask any questions
OGS = Oakland Gardens Solar node, the one in the picture I posted earlier
lol yes i feel like most of the time i send direct messages to a node that isnt very close to me then it doesnt deliver, but somehow every so often i get a message from a passing plane
Ohhh got you
hehe yeah, "height is might"
makes sense
lmao
so as far as the primary channel, is that sort of like a group chat for your local mesh?
yeah, it's basically "public" so anyone with meshtastic that also using LongFast modem preset will also hear you, vs a "private" channel using a preshared key to encrypt it - so most everyone will still relay both channels but the "private" channel can only be decrypted if you have the private key
Got ya thank you i appreciate you
so you can setup a private channel between friends/family/whatever that's totally private and also participate in the public chat when it occassionally happens (or just ignore it entirely)
like setting up another channel besides the primary one
for your friends/family if you wanted to
by sharing the keys
and my one last question that i cant seem to find online is that I notice some nodes have the unlocked vs the locked icon next to it, i can see this means that they have a shared key, how do i make my node have a shared key as well?
ive been trying to figure this out for like a week lmao
oh HAH - that is actually very silly, it merely indicates the other node isn't using PKI (public + private keypair) encryption, and means they are either a> running on firmware before 2.5.x where PKI was introduced and/or b> they're running in "licensed" mode which doesn't allow for encryption
for nodes with the lock you've got their public key already on your node so if you DM that's also encrypted using the other party's private key
LOL ok that makes a whole lot of sense as to why i couldnt figure it out😂
thank u dude ahaha
noooo prob! I gotta head to bed but stick around, there's def some folks out your way in here that'd be great to be able to talk to over RF
and def ask silly questions - its how you learn silly answers!
of course man if i was scared to ask i wouldnt have been asking you 😂 appreciate you dog Have a good night God bless
There's a few as tall as those. Creedmorenbeing one
I always wanted to eat there but was too poor.
has anyone noticed that ch utilization is really high? idk if this has to do with 2.6 but im not sending or receiving any messages
Isn't that the usage of what channel you are using? Not your usage. And I assume most people here are using the same in NYC. I thought there was a way to manually set this but I don't see how
Yes, this would be the utilization of the default LongFast frequency slot 20. I know it’s not just my usage.
Usually it’s not this high unless I’m sending messages (my usage) or I’m receiving messages from others (other people’s usage).
It’s possible a bunch of people are DM’ing but unlikely this late at night.
I just watched a video explaining that you can change your frequency to something less congested. But I don't see that setting in Android
It’s the frequency slot setting but I don’t want to change my frequency slot. I’m just trying to determine if channel util is just legit high right now or if it could be related to 2.6 as that’s the only change I’ve made recently.
If you change your frequency slot, you won’t see any of our nodes anymore because we’re all on the default, but yes there would be less congestion.
I'm really hoping this isn't a crypto mining scheme that we are all unknowingly participating in 😐 There's so much promotion for Meshtastic and all the devices but being that the mesh messaging is completely non functional I can't see where all the hype and funding is coming from.
I got sucked into something like that before...
There is supposed to be a setting where you can manually change the frequency but I can't find it
That’s such a wild theory that I don’t even know how to respond to it. 🤦🏻♂️
Wild maybe but something is off.
Meshtastic is an open source project meaning that all of the code is visible to everyone. Everyone would immediately know if there was some sinister crypto mining code baked in lol
This would also be the absolute worst platform to crypto mine with because it’s such low bandwidth.
Is the app itself open source? I didn't think it was
I think you don’t understand Meshtastic very well which is ok but i’m gonna excuse myself from this conversation.
Aren't most of the devices, antennas etc cross compatible or entered the market for use in Helium crypto mining?
And the venders who sell them also sell most of the Helium crypto mining devices as well
Or is the bandwidth different for Helium?
So accasionally some nodes come online in a position to connect me to the rest of the NY/NJ mesh. Exciting! Longest successful traceroute I've seen yet:
That’s the key with higher frequencies, they reflect off different types of surfaces somewhat, which is why cell phones operate on “microwave” frequencies, such as LoRa. It’s basically line of sight. But keep in mind collisions , even spread spectrum, packets can be missed due to collisions, retries can help but often make things worse. Lastly, LoRa is in the ISM band, which is used for commercial microwave heating and medical device heating, so there’s interference sources there too. It affects S/N ratio or the “background noise” - that could explain why sometimes it works, sometimes it doesn’t given the nodes are fixed- unless buildings move or some other object obstructs the view temporarily. Not likely. Amateurs have band privileges on 33cm, same area as LoRa but as secondary users, so there’s that activity there as well (very low usage around here). And, lastly, there’s not just LoRa in this band, there’s helium as some mentioned, Amazon sidewalk and dozens of other similar spread spectrum activity
Forgot to mention the power level here, 0.1 watts out of the box without a power amplifier. Things change significantly if you bump it to 10 watts.
I'm on a 1W MeshAdv Pi Hat. I was looking into switching to the licensed mode and bumping my power with an amplifier but I've read that there's not much activity in the licensed mode.
So I did some testing. Can someone explain? I was about 2 miles away from my house, in a car. I direct messaged my node, 2 were successful (got the check box in the cloud and then the little man), and 2 just had the line across the cloud, which means unsuccessful. I then shut down my portable t-echo. But when I went home, it showed that I got all 4 messages.. So could it be that our messages go through but people just don't respond when direct messaging others?
I thought licensed mode was just more power not a different channel
@radiant cairn 's blog had a comment "When I first set up my nodes, I disabled encryption and enabled the licensed operator mode. I saw no nodes other than mine after a week of testing." so I assumed that even Long Fast with its known encryption key was disabled.
@normal osprey sent you a message this morning.. It did not go through.. 😦
I just got a traceroute through. Try hitting Grand St as that’s who we are hopping through.