MeshTower v2 (30dBm version) clock reset problem w/ Meshcore

#1

So I’ve completed deployment of a Meshtower v2 (higher power, 30 dBm) version. I have flashed the device with the Heltec-provided Mechcore v1.16 firmware for the repeater configuration:

https://resource.heltec.cn/download/MeshTower-V2/firmware/meshcore

However, the node appears to be auto-rebooting every X minutes (about 5-10?), which resets the onboard clock.

For example, using the remote management meshcore CLI. I can retrieve the firmware version (which matches the heltec download), get the current clock value (note the 2024 year value). Set the click using the time command and the UNIX epoc value (seconds). Then immediately fetch the clock value again to retrieve the correct time:

> ver
v1.16.0-fe59589 (Build: 13-Jun-2026)
> clock
10:54 - 15/5/2024 UTC
> time 1783810804
OK - clock set: 23:00 - 11/7/2026 UTC
> clock
23:00 - 11/7/2026 UTC

However, if I wait several minutes and again use the command line to retrieve the current time, you see that it has reverted to 2024 - the gab between these two commands was about 5 minutes.

> time 1783810804
OK - clock set: 23:00 - 11/7/2026 UTC
> clock
23:00 - 11/7/2026 UTC
> clock
10:53 - 15/5/2024 UTC

I have been periodically resetting the clock, only to see it revert back to the 2024 value throughout the day. When retrieving overall device status, I will typically see uptime values below 10 minutes:

Battery
100% / 4.21v
Clock • At Login
15/May/2024 3:59AM
Uptime
0 days Oh 6m 45s
Total Airtime
TX: 0 days Oh Om 5s
RX: 0 days Oh Om 5s
Last RSSI
-48
Last SNR
10.8
Noise Floor
-119dB
Packets Sent
Total: 6, Flood: 5, Direct: 1
Packets Received
Total: 8, Flood: 5, Direct: 3
Duplicate Packets Seen
Flood: 1, Direct: 0

Has anyone else experienced similar problems using the meshcore firmware on this hardware configuration?

1 Like
#2

I am having the exact same issue with ALL 4 of my Heltec Meshtowers. I have bumped into some threads that have mentioned this is a bug with the watchdog which causes the reboot and there is a git commit for the updated pull. Unfortunately being that the version we are running is built by Heltec it may not include this yet. :frowning: Regardless, this is very frustrating and I hope Heltec fixes it sooner then later as I am sitting on 4 very useless repeaters now.

#3

So I’m a very impatient person… :slight_smile: I went ahead and did a new build from the dev branch of meshcore which included the fix I previously mentioned. Also looks like the dev branch includes the meshtower v2 definitions now. I have flashed this unofficial build to one of my meshtowers and have had an uptime of over 20 minutes now. Much better then the 9 it was constantly resetting at. I would share the file here, but not sure that’s allowed?

#4

Ugh, Now I am wondering what special sauce Heltec added to their firmware. Unfortunately their commit of fe59589 seems to be from their own fork and not located on meshcore’s repo. Problem is, using the stock meshcore builds battery does not get read, I receive SOME messages, but very few, no one receives anything from me (only very local companions). When I switched to the Heltec build of fe59689 suddenly more messages were received and people actually could see my messages, and reads the battery BUT, the meshtower reboots every 9 minutes. So as I previously mentioned, built a new file from the dev branch of meshcore, which did include the new meshtower v2 definitions, and now, I am back to receiving some messages, no one can see mine (only very local companions), but I can at least read the battery, and there is no reboot after 9 minutes. So please Heltec, either add the meshcore commits to your publicly accessible github repo, or let us know what special sauce we need to add to this custom build to make it work like yours.

*Update: Okay, I was wrong, seems it may just be the conditions, my repeater was picked up by the next repeater 10km away, but it is intermittent. So seems my build is working. Giving it a few days. Apologies for the back and forth on this thread. Heltec staff/peeps, if you have any further info to what you changed in your meshcore release that would be helpful (Maybe).

#5

I did receive an email reply from Heltec Support (which had been forwarded to their Product Manager, Richard Zhang) with a test firmware file attached. I have yet to test this, and am unsure if it includes the open source patch (from here - https://github.com/meshcore-dev/MeshCore/pull/2936).

FWIW, that pull request does included a prebuilt binary attachment for a repeater node if you don’t/can’t build it yourself easily.

I plan to deploy & test later this week. But I agree, it doesn’t look like anyone from Heltec actually tested their own firmware release :frowning:

#6

My uptime is now at 3h 44m 1s. After about the first hour of it being back online I started connecting with the rest of the mesh. Seems all my problems are solved using my build.

#7

Using the firmware build supplied by Heltec:

v1.16.0-cd70894 (Build: 14-Jul-2026)

My node’s uptime is now 18h and climbing.

2 Likes
#8

That’s great, hopefully they publish it soon for other users. As for me I have now flashed my build onto all my towers and all are working great again with no reboots.