rusefi.com

@JRD McLAREN
Can you assist?
have you updated any of your boards recently?

JRD McLAREN wrote: ↑

Wed Aug 24, 2022 4:08 pm

I can give it to you ....

Status update: it does not turn on at all. I might have soldered wrong USB connector part number

stayros32001 wrote: ↑

Sun Nov 27, 2022 3:17 pm

It is the custom version of @JRD McLAREN MRE+sd 5.55rc

Best I can do is ship my unit of custom version to @mck1117

AndreyB wrote: ↑

Sun Nov 27, 2022 3:31 pm

stayros32001 wrote: ↑

Sun Nov 27, 2022 3:17 pm

It is the custom version of @JRD McLAREN MRE+sd 5.55rc

Best I can do is ship my unit of custom version to @mck1117

to try to start the board and reproduce the issue?
Is any way that I can help also?

stayros32001 wrote: ↑

Sun Nov 27, 2022 3:40 pm

to try to start the board and reproduce the issue?

yes

stayros32001 wrote: ↑

Sun Nov 27, 2022 3:40 pm

Is any way that I can help also?

How many of affected boards do you have? Worst case you would be shipping to Matt if we cannot get this one already in USA to start.

..OK .. I'm here ...

Now I have two boards ..and I can test FW update ...

..wait ...

20220422 works
20220703 works
20220812 not - tested twice
20220914 not - tested twice


programed via STLINK ...

last release ended with RED FATAL, no serial port...

AndreyB you have "my UA" version ...
and stayros32001 you have "purple" ..??

it have TLE8888QK from Digikey
STM32F407 VGT6 from Mouser - desoldered MCU from Mikroe EasyMx Pro v7 board

stayros32001 would you can make photo of "both" MCU ..??
if you still have "old burned" board ... and actual one ..

(old one may have "pre-covid" MCU, bought from TME)
(try to flash it ... )

Both ( usefull and burned where purple)
these images are the best i can do

(i will try to locate the burned)

So any ideas?
is it something at newer firmware's that needs to be changed for "STM32F407 VGT6" ?

AndreyB wrote: ↑

Sun Nov 27, 2022 3:43 pm

Worst case you would be shipping to Matt if we cannot get this one already in USA to start.

Matt is at US?
This means customs/taxes and maybe months without my daily-driver

I have the same series on my boards ...
AndreyB probably too ...

stayros32001 both boards have the same MCU series .. ?? (from Malaysia) ..

EDITED:
hmm AndreyB's old one has the same MCU rev.
https://rusefi.com/forum/viewtopic.php?p=41521#p41521

TLE8888-QK vs. TLE8888-2QK ... maybe ..

software perspective:

This might explain why every next version after july fails!
https://github.com/rusefi/rusefi/releases/tag/20220812_release

...YES ...

as we discuss here: https://rusefi.com/forum/viewtopic.php?p=46765#p46765

some MCUs / MREs have problem to start after fw upgrade, due to OpenBLT.
last working version is 20220703 - without OpenBLT

I'm try to flash "all my boards" with STM32F407VGTx as I can..
I have boards with MCU from Malaysia, Philipines, none from China

all STM32F407VGT6 from MYS has problem (my last used one, purple and "UA" boards) (aprox 8 boards, AndreyB has one)
STM32F407VGT7 from PHY has problem (first black edition) (only 2 boards, one in my car)
STM32F407VGT6 from PHY is OK (first black edition)
STM32-DISCOVERY sometimes work, sometimes not ... (strange) full erase & re-flash can fix it (tested from 20220703 to 20221117)

and the questions are ...

Can rusefi developers do something with this or not .. ??
It is question for OpenBLT / Feaser developers ..??
Can I make / build "custom firmware" without OpenBLT from source .. ?? ... and test it ..??
(..with STM BLT)

We can't hold the origin source of MCU for the future, or make a choice before buying...

(Affected user are: "me" - not important, stayros32001 and Martin Matousek)
I'm not sure, if "last" MCUs from MYS (desoldered from Mikroe) are counterfeit or not.
Works well in 3 cars with 20220422

JRD McLAREN wrote: ↑

Mon Nov 28, 2022 9:13 pm

Can I make / build "custom firmware" without OpenBLT from source .. ?? ... and test it ..??

I have to be blunt and harsh: you have my permission to do anything with the firmware within license.

At the moment I see clear correlation between custom microRusEFI hardware and the firmware issue. At the moment I see very weak evidence that OpenBLT is to blame here.

The issue was reported enough to get noticed. There is no need for additional reports in the next few weeks.

Your next step is to reproduce the problem with any official rusEFI hardware and report what exact version of what exact official board is affected.

AndreyB wrote: ↑

Mon Nov 28, 2022 9:40 pm

I have to be blunt and harsh: you have my permission to do anything with the firmware within license.

maybe another question ...
Is stm bootloader in source code .. or not ..??
I'm never compiled custom firmware before....

At the moment I see clear correlation between custom microRusEFI hardware and the firmware issue. At the moment I see very weak evidence that OpenBLT is to blame here.

The issue was reported enough to get noticed. There is no need for additional reports in the next few weeks.

Your next step is to reproduce the problem with any official rusEFI hardware and report what exact version of what exact official board is affected.

Black MRE are build from official repo. It is not custom PCB ...
Would you can send me your official MRE or STM only.. ??
.. I can solder your MCU to my board and test it ...

AndreyB wrote: ↑

Mon Nov 28, 2022 9:40 pm

Your next step is to reproduce the problem with any official rusEFI hardware and report what exact version of what exact official board is affected.

this is very ineffective and that upsets me a lot.

report what exact version of what exact official board is affected

Would you can flash this "your old official board" please ... ??
if you have one .. (or somebody else - mck1117 .. ?? )

https://rusefi.com/forum/viewtopic.php?p=41521#p41521

it has "MYS 99" mcu - the same as my. one dot on left-down corner

"your new official version" has "PHL" mcu - two dots - left-down & right-up corner

JRD McLAREN wrote: ↑

Mon Nov 28, 2022 10:00 pm

Would you can send me your official MRE or STM only.. ??

You can..??

Board revision is on the back but at least I see microUSB and that says old board

I have already disassembled all microUSB for spare parts. Here is a random 0.5.2 which works fine.

Until https://github.com/rusefi/rusefi/issues/4847 I was sure that all chips are the same. Well, https://github.com/rusefi/rusefi/issues/4847 tells us that there are minor nasty differences

I can put this board on the plane to EU on Dec 23rd.

SAME SAME .... , BUT DIFFERENT ....

I suspect that the problem is in different MCU revisions - country of origin ...

AndreyB wrote: ↑

Mon Nov 28, 2022 10:49 pm

I can put this board on the plane to EU on Dec 23rd.

OK, you need my address ..??

BTW:
Original STM32F407-DISC1 with F407VGT6 "MYS" has problem
aprox. 4y old, bought from TME.eu

I have 4 proteus, 3 with F429ZGT6 "KOR" - Korea
and 1with F429ZIT6 "PHL"
I can test it and send results.

AndreyB wrote: ↑

Mon Nov 28, 2022 10:49 pm

Board revision is on the back but at least I see microUSB and that says old board
I have already disassembled all microUSB for spare parts. Here is a random 0.5.2 which works fine.
Until https://github.com/rusefi/rusefi/issues/4847 I was sure that all chips are the same. Well, https://github.com/rusefi/rusefi/issues/4847 tells us that there are minor nasty differences

I can put this board on the plane to EU on Dec 23rd.

This your "old random board" have STM32F427VGT6..
not F407VGT6 ...

@AndreyB

Do you think that is possible to locate users with the MCU revision and board that @JRD McLAREN mention here?

https://rusefi.com/forum/viewtopic.php?p=46791#p46791

Do you keep for example an xls-db with MRE versions and STM versions that the users bought?

@stayros32001 I've shipped a custom board to @mck1117 yesterday and that's all I can do here. I am very grateful that you are playing with my hobby project but I have to make calls on priorities. At the moment this strange issue is weirdly isolated and it's just not on top of my priority list considering the weird localization of problem.

Time, price, quality. Pick any two.

Your quickest way forward is getting a unit from Simon@FutureProof https://rusefi.com/forum/memberlist.php?mode=viewprofile&u=848

Only 3 boards with "suspected" MCU are out of me... in the world...

stayros32001 and M.Matousek have these boards.

But it looks, that all F407VGT6 MCUs from MYS (Malaysia) have problem with Openblt.

@stayros32001 - would you check mcu on your's boards .??
old - burned one ... and working one ..??

it is very likely, that different series of processors were sold in Europe and America or China.

stayros32001 MRE can be fixed with right MCU - PHL one ... at this moment ...
..but, what we do in the future ... when .. . . ...

updated gitub content ...

https://github.com/JRDMcLAREN/hw_microRusEfi

two boards waiting for F427VGT6 mcu ...
To mck1117:
Can I ask for any test results .?

"My" Proteus with "KORean" mcu has the same problem as MRE with "MYS" one

...it's there ..

@JRD McLAREN made for us an firmware with the "tranditional" bootloader and it works fine.
so, bootloader seams to be the one that we should blame.

@mck1117 did you worked on JRD McLAREN unit that that @AndreyB send you?