how many USB-Cs are there out there?
I’d still also like to understand the interaction that caused it be ok with the USB plugged in vs not.
how many USB-Cs are there out there?
I’d still also like to understand the interaction that caused it be ok with the USB plugged in vs not.
I’ll send you one if you wanna play with it? I have a oscilloscope if you have any ideas I should try. I think I am going to apply the capacitor fix to a USBC version, send a regular microusb version, and send a “fixed” microusb version as well. We can see if any or all of them work.
Good point 4 maybe, but they do work for me. I will be testing this more on my end as well. Just about to head out to the shop to do that.
Was the capacitor the ultimate fix, or was it the 10k resistor from gpio0 to 3v3 that was the ultimate fix? Just scanning back through the Discord thread…
The USBc needs the cap to make boot work better and the micro need the resistor to make reset work better. I think I have the summary here, Genuine vs Knock off esp32's - #4 by vicious1
Seems a bit obvious at this point but I was actually pretty sure I sent the microusb as the replacement. I didn’t know we were still on the USBC version.
When Keith ran fuildTerm, it reset the esp32. But the TMC drivers had already booted, and were quiet. It is only when the TMC drivers are booting at the same time as the esp that it gets stuck in boot loader mode.
That’s my theory anyway. Pushing the reset on the esp after the power is on should also have fixed it. But I think it didn’t (I can’t find it in the evidence above).
Well thanks again, and I’m happy to serve as a problem environment for any future revisions if it is ever helpful. If you need me to test something related to this issue, you now have a location.
It might sound ridiculous, but you put yourselves out to create things, so I’m happy to give a little tiny bit back by serving as a lab rat once in a while.
It didn’t fix it unless a USB battery or computer was connected to the USB C port. Note that I also tried this with a common ground from the PSU to the metal housing on the USB C (I just pressed it there) and it didn’t appear to work either (I was literally sitting there with my finger on it, so I may not have been as patient as I was for the other variations.)
I don’t understand how all these uarts can run on one wire… I am sure it is logical, but it looks weird.
The 0(boot) pin should be pulled up on the board. The BOOT button pulls it back down. So maybe the pullup resistor on the USBC version is a little larger, causing it to get closer to ground when there are enough TMC drivers installed.
You’re being very patient with our random test requests. I really appreciate a good “patient”.
I think that is very plausible.
Okay. I can make it happen now!!!
And…
The added capacitor does not fix it. Boots into flash mode just fine, but still does not broadcast an SSID on power with 5 drivers in the socket.
Not sure why it tested fine before. I have had three different batches of USBC version come in so maybe some were different. The first ones I did flash and test with the USB connected, so that is obvious but that last one I sent worked fine, maybe I only had one driver in it to test…I did that for a while. Crap who could have known.
I will send out a microUSB version and take the USBC version off as an option.
If anyone needs some USBC ESP32’s I have a bunch…
I have learned alot since being here!
There must be a solution! Do you have any old 8825 s or (the 4000 series, crap i hate getting old!) To see if they work?
Maybe you can sell with those!
Um, I still need to do some homework, but didn’t I read somewhere you can use the esp32 with wled, to do fancy things with leds?
Sounds like the shop has one of the parts for those wanting bling lights on our machines. ![]()
Or am I the only one that Johnathan has influenced? Ha
Actually just came across that earlier today
I’ll order at least one or two just to have around to set up a test bench with.
Indeed.
Nope, that’s going to break UART communication once booted, as noted elsewhere in the thread.
If I were to work on an enhancement, I would indeed submit a pull request.
You’ll note from these very threads that I have advocated for funding FluidNC, and Ryan can confirm that each order I’ve ever placed has given a tip specifically to support the open source nature of his products and his business model. I DO speak with my wallet.
Can you confirm it works several times in a row after only changing to the micro version?
Can you put a scope or at least a volt meter on that boot pin while plugging it in via 24V? I am curious how close it is.
Yeah, I tried a bunch on several devices. I tried two boards as well with 6 drivers on there.
Good idea, see if that cap does anything at all. Let me see if I can set it up again right now.
You’ll need the scope to see if the cap does anything. Need scope shots with/without the cap.