Ahh. Okay, I know there’s been issues with using the web flasher first so I was curious. Do you have a picture of your device?
Hear a picture during the boot process of SoftRF (v1.3) all sensors seem to be properly recognized and initialized. I received the device 2 day ago.
Hmm. Okay, definitely using the right hardware and correct firmware version. Have you tried flashing it directly to the core without it inserted into the body? And then testing to see whether it worked? If that doesn’t work then I would definitely open an issue here: Issues · meshtastic/firmware · GitHub
@Crichton
I tried and flashed the firmware on the pure core board. Unfortunately the same result. When I boot without the body the 2 minute pause after „Scanning for i2c devices on port 1“, which occurs when connected to the body, is ‘missing’. For me it looks like that scanning for i2c devices on port 1 does not work properly and get a ‘time out’. I will open an issue.
Thanks for your support
It would be good to open a GitHub issue and I can tag a developer from lilygo.
@garth
thanks for your support, please see [Bug]: T-Beam Supreme - Display stays black · Issue #2519 · meshtastic/firmware · GitHub.
I’m not to familiar with how to create issues. If more details are required please let me know
A fix has been submitted Fix t-beam-s3-core display not working by lewisxhe · Pull Request #2520 · meshtastic/firmware · GitHub
1 Like
Has anybody spotted this new release yet?
1 Like
Awesome, thanks Lewis!
Yes, but also that’s unrelated to this post. Please create a new post if you wish to discuss. **Edit: Looks like a post for this product has been created, you can reply there.
Fix is working for me 
. Thank you very much for your quick reaction.
The BMP-280 sensor is now also detected, but no environmental data available.
I opened a new issue for this: [Bug]: T-Beam Supreme - No environmental telemetry data (BMP-280) · Issue #2524 · meshtastic/firmware · GitHub
Hi, Apologies for the delay in replying
flashing done via CLI, now using fixed build and screen is now working, bootup time significantly improved now too as its not getting stuck probing.
1 Like
I just got a TBeam Supreme off AliExpress tonight, and tried to use and found multiple issues. First, the 18650 battery holder seems just a little too small (and/or the spring tabs just a little too tight) to support a battery. I feel like it needs to be 1-2mm longer.
Second, and more importantly, the uBlox M10 performance is awful. It took 4.5 minutes to get an initial fix while out in an open area outside with a clear view of the sky. When I walked inside, with a solid fix, I lost the fix almost immediately. The GPS performance is unacceptable and unusable.
Anyone have any ideas?
I can believe this is a protected cell and may not be compatible, although frankly, it sucks that protected cells aren’t compatible. Is there a requirement to have a cell in there to begin with? I know on NanoVNA’s the noise goes up exponentially if powered from a USB connector vs a battery, so I’m wondering if running it off a battery may improve performance.
I want to be clear on this. The performance of the curret setup is poor. It’s not a little poor, it’s not just marginal inside, it’s marginal outside. Ublox quotes a COLD startup time of approximately 30 seconds or less, but this took me 4.5 minutes to start up OUTSIDE with a clear skyview. My other GPS units like a VK-162 dramatically outperform this thing, as does… I think, older versions of the TBeam. I spent extra to get the ublox M10 because I wanted top notch performance and frankly, this is below bargain basement
I’m not sure if this will work for the T-Beam Supreme board, but lilygo released a GPS reset firmware that can be flashed that clears any lingering GPS data. This solved the GPS acquisition issue for older v1.1 T-Beams. Then you flash the firmware again and it should acquire GPS lock pretty quickly.
lilygo only recommended this as a last resort, and I have no idea if it will brick your Supreme board so I don’t recommend you give it a try without asking an expert. However I do know it worked perfectly well on older boards.
I’m not sure what to say at this point. I kept power cycling in my house, eventually, I was getting a fix in about 4 minutes or so. Over time the fix kept getting better. After a half hour, I was seeing it fluctuate between 9 and 12 satellites, whereas when I get the initial fix it’s always just 4.
It kind of feels like there may be bad configuration data in the GPS, as if it’s trying to get a fix on China and only slowly realizing it should be looking for satellitea over the US… and it’s only looking for the GPS constellation. I really don’t understand why it takes another 20 minutes to pick up the additional satellites… that seems like a really long time horizon.
On Friday, I’m going to go outside and plug my laptop into the VK162, uBlox M8 GPS in my aircraft and see how long it takes to lock and then do the same thing with the TBeam side by side. That works be a bit more objective. Although perhaps the VK162 has some internal power that keeps it in a better state to start…
Also this.
If the GPS hadn’t been turned on since it was mfg in China then the first time will/could take a long time. Every single t-beam i have, i have switched out for a better GPS antenna. The stock GPS antennas suck, the stock Lora antennas suck.
I’ll be posting a T-Beam Supreme Case shortly (been testing various iterations on my printer, Ender 3 Max Neo) so hopefully will have a useable case for these.
I want to again thank everyone for the feedback. I’m a bit confused about all of this.
- I don’t understand this whole almanac thing. The cold start spec is about 30 seconds. I can imagine that a lack of an almanac might prevent you from getting a fix, but it shouldn’t prevent you from acquiring a SV. It shouldn’t even slow it down much. I presume there’s
It certainly appears to me as if there is no data storage between power cycles, no battery (my 18650 isn’t installed since I don’t have one that fits), and no keep-alive cap. So the question is… is there some sort of non-volatile memory in the uBlox or information stored in the ESP32S3 (which certainly does have non-volatile) that is loaded into the uBlox? If so can it be corrupt or incompatible with my geographic location (USA)?
If there is a GPS firmware reset, can someone point me to the firmware? Does it work for the TBeamSupreme with the ESP32S3 and uBlox M10?
- So far the only way I know to interface to it is through SoftRF which it comes loaded with (and is my intended application). It simply says how many satellites it has and whether there is a fix. I am somewhat surprised at the behavior where it starts off with a few satellites and slowly adds more until it stops at 12. It’s conceivable to me that the more satellites it has a fix on, the easier it is to obtain new satellites because the “time error” (the disrepancy between the internal clock and those of the atomic clocks) is reduced as the fix gets better, which might make it easier to correlate.
Is there a way to hook it up to the uBlox Center-2 software so that I can get better data (perhaps other firmware that will give more useful results). I found a “GNSS_Access” firmware on SoftRF’s Github repository, but I don’t know what it is. When I tried simply plugging in the TBeam (with the SoftRF 1.3 installed) device repeatedly connected and disconnected (switching COM ports) – it didn’t reboot though, and the issue was only present WHILE running uBlox Center2.
- I do have to wonder why they are shipping this with an inadequate GPS antenna. It’s clear the antenna is needed for most uses – so why is it acceptable to have a non-functioning antenna. Again, I have a USB GPS mouse that cost $10 that cold starts in 50 seconds and then warm starts within seconds, and as far as I know it doesn’t have a battery or “keep-alive” memory.
I can order the antenna mentioned on Amazon, it’s only $11, not a big deal. I’m not sure what LoRA antenna (915MHz) to get though.
- Or is it possible it’s actually defective?