Two rev D boards would not go into software DFU

[rusefillc]

Two out of two rev D boards I have would not jump into SW DFU

HW DFU works fine (both power on with PROG and reset under PROG) SW DFU just reboots the board into without not DFU mode - i.e. we loose USB connection, we change blue blinking frequency but we keep bilking blue

@BeerMoneyMotorsports has the other three rev D boards in the mail so he would confirm on his end hopefully @BeerMoneyMotorsports has a rev C to compare

[BeerMoneyMotorsports]

@rusefillc i have rev C to compare

[rusefillc]

i run experimenters with https://github.com/rusefi/rusefi/releases/tag/2021.06.28_release

i think once it has rebooted into DFU but only once.

so it seems sporadic? maybe some power supply drama?

[andreika-git]

Can you please make a close-up photo of the STM32 soldered on the board? We need to compare the chip labels with rev.C

[mck1117]

I've just tried on a Rev A hellen 121 and it rebooted back to the firmware, not DFU

[rusefillc]

@mck1117 welcome to the party!!! what about 121-vag-B?

@andreika-git any chance you can try Hellen-72 A or B you have?

[rusefillc]

@andreika-git @mck1117

H72-C is F429IIT6 H72-D is F407IGT6

[mck1117]

I have just flashed snapshot_20210628_214332_hellen121vag_rusefi on an old Proteus 0.4 of August 2020 vintage, and "reset to DFU" button WORKED

[rusefillc]

I wonder if there is a correlation or causation between exact CPU model and this issue

Which F4xx is Rev A hellen 121 vag? Which F4xx is Rev B hellen 121 vag and how does it vote? Which F4xx is old Proteus 0.4?

[andreika-git]

@andreika-git any chance you can try Hellen-72 A or B you have?

SW DFU works on H72b (STM32F427IGT6)

[rusefillc]

So one would suspect 407 right?

I've just tested with a small old DISCOVERY 407 and SW DFU works just fine? so 100 pin 407 is happy and 176 407 is not?

[rusefillc]

@mck1117 could this be about power supply? especially my sporadic one time good reboot

maybe I shall add AP2337 tomorrow?

[andreika-git]

Also you can try with an external +12V power

[rusefillc]

@andreika-git is in the lead - rev D enters SW DFU as long as powered by +12v

[mck1117]

It happens if connected to an actual USB host, but not if connected to USB power (usb power bank battery)

[mck1117]

I've just tested with 12v connected, and it DID NOT HELP

[rusefillc]

I've added one AP2337 to MCU module that made zero difference

I am not 100% but it looks like PROG holding + RESET does NOT always get me DFU only MOST of the time?! Need to made a video

[mck1117]

I see the same thing, button is not 100% reliable either

[rusefillc]

https://community.st.com/s/question/0D53W00000wyoGESAY/sporadic-reset-into-dfu-and-dfu-jump-any-links-between-power-and-dfu

[rusefillc]

Matthew Kennedy 10:29 PM I did some probing

The dfu software is running, then resetting after ~800us

Before the oscillator initializes

But it didn't even try the osc

[rusefillc]

https://github.com/rusefi/hellen121vag/issues/50 was done while investigating this issue

[mck1117]

Observation 1: powering from 12v (no USB connected), BOOT0 button is 100% reliable Observation 2: powering from 5v USB power bank (no 12v, but also no data lines), BOOT0 button is 100% reliable. Test performed with the same cable that is unreliable with laptop. ∴ the problem is not power, but is USB data lines.

[mck1117]

H72B is 100% reliable with F427IGT6

[mck1117]

Took the chip from the working H72B, put it on an H121B board. Works perfectly.

[mck1117]

DFU ROM matches, except for the odd duck that's a very old ST F4 discovery board:

[rusefillc]

I've shared the drama with @dron0gus

Dron says 407 has one VCAP while 42x have two VCAP or something? i like this line of inquiry

[rusefillc]

With VAG and NA6 reports this is now https://github.com/andreika-git/hellen-one/issues/79