Page 2 of 3

Re: OTA update problem

Posted: Tue Oct 27, 2015 8:05 am
by danicampora
It can be changed, but it needs to be changed in the bootloader, so there's no hope for the boards that are already shipped (~2500). GP28 was chosen because is the only pin that doesn't have alternate functions, so you are likely not going to miss it much.

Cheers,
Daniel

Re: OTA update problem

Posted: Tue Oct 27, 2015 8:34 am
by Sebastian
How is the bootlloader flashed onto the board? I'm just asking because the board is still "working" but will become more and more useless with newer python versions. If there is a way to "repair" it and only loose a dfferent pin it would be OK.

Re: OTA update problem

Posted: Tue Oct 27, 2015 8:56 am
by danicampora
With the help of the expansion board (or any USB-serial adapter, but the expansion board is already using the right pins), and a bit of soldering (to connect SOP2 on the test points below to 3V3), you can use Uniflash (Texas Instruments flasher tool), to reprogram the bootloader. Before all that you would have to build a bootloader version with a different SAFE BOOT pin, but that's easy to change, just one #define...

Re: OTA update problem

Posted: Tue Oct 27, 2015 9:39 am
by Sebastian
Thank you...sounds good. I will try to get it working...

Re: OTA update problem

Posted: Tue Oct 27, 2015 9:54 am
by danicampora
Hi,

Detailed instructions for flashing can be found here: https://github.com/micropython/micropyt ... /README.md

Bare in mind that you have to do it all, reformat, flash bootloader and application (mcuimg.bin), etc.

Cheers,
Daniel

Re: OTA update problem

Posted: Tue Oct 27, 2015 11:44 am
by torwag
Hey,

I checked the schematics and it looks like you protected the wipy against wrong polarisation via T2.
Since Vin seems to only used at that point and for the solder header, one method to protect against wrong polarisation would have been to swap safeboot with GND. In case of wrong orientation, one would also generated wrong polarisation and hence T2 would shut down the power to all subsystems. The wipy would remain completely dead which is a good indicator for "something is going wrong here". Is the V_in ground isolated from the common ground of the wipy board. I couldn't get this from the schematics. Maybe something more sophisticated as you did on the extension board (looks like T2P and T2N act as polarisation and battery protector) would be even better to avoid positive potentials on all ground lines.

Nevertheless, I also would suggest to wire V_in behind T2. At the current state the wipy board is protected but all external electronic would blow up for wrong polarisation since it is directly wired to Vin. Is there a design decision for adding it that way? T2 should scope with about 6.6A that should be enough to allow current going through to power external peripherals as well. I know all this is not an easy decision as you might face people with inductive loads which suddenly blow up T2. Just wanted to add my part to the confusion around a V2 or a V1.4 version. ;)

Cheers

Torsten

CC. A new order of wipys should be on the way....

Re: OTA update problem

Posted: Tue Oct 27, 2015 8:07 pm
by killwas
Hi
I'm sure I did not seat the WiPy wrong on the expansion board, nor did I connect GP28 with 5 V. But I got all the symptoms. To cut a long story short. My WiPy recovered after I disconnected the expansion board and the power.
It even started with version 1.5 although I could not upload it before (or so it appeared), though I tried many times and also rebuild the filesystem. Seems hard reset does not equal disconnecting from the power.
Regards,
Killwas.

Re: OTA update problem

Posted: Tue Oct 27, 2015 9:49 pm
by Sebastian
Can anyone tell me which test point is SOP2 on the bottom?
I did not find any documentation about the test points.
Thanks.

Re: OTA update problem

Posted: Tue Oct 27, 2015 10:15 pm
by danicampora

Re: OTA update problem

Posted: Wed Oct 28, 2015 1:12 am
by Sebastian
Thank you for the link.
The solution worked for me. I compiled both files and uploaded them to the wipy. After that I did a normal OTA update with v1.1.0.
The only strange thing is, that the bootloader ist not working as expected now. I thought i set it to GPIO 22 but the factory firmware is not booting when connecting the pin to 3.3Volt. But if I use machine.reset() the wipy is booting the factory firmware. After pressing the reset button the newest firmware is booting again. So the bootloader seems to work.
In the WIPY folder I changed the define in the mpconfigboard.h from pin 18 to pin 15 to use GP22 instead of GP28.
Did I made a mistake here?