[PSIG]

This is a repetitive issue. I wonder if we could easily make calibrations part of the tune, at least for reference?

[Mr.Grimes]

Yep, it was calibration issue, thank you. I was thinking its tied to the tune.
It seems that the pressure calibration stays the same with the tune.

BTW: for Spectre and similar boards should be stated to not use USB on STMboard while powering the ecu from 12v. Cumunication works, it just slowly cooks your 5v regulator.

[moussie369]

Hi, I finally found the culprit for my problems it was simply a bad solder joint inside my rs232 adapter that decided to let go, I kept struggling since I rebuilded an other board and tried multiple things thinking I had fried something.

I'm coming back because I think I've found a bug with version 202210-dev from https://github.com/Tjeerdie/speeduino

I have a spare F407ZG that has no problem running it so it took me a while to understand that my f407VE wasn't fucked. Updating from version 202204 seems to have broken RS232/TTL, when using the detect function it finds firmware version "RQRQRQ", sometimes it's able to connect but isn't stable and will crash after a few seconds

[porschenstein]

I'm trying to get running on a chinese "DevEBox" stm32f407vgt6. I'm working in Visual Studio Code with PlatformIO. The code builds but I can't get it to upload via DFU. I've been fighting with it for a while - initially there were no drivers but I downloaded some "old" ones from STM and can see the DFU driver in Device Manager when it put the board in "boot mode".

I tried the "black_f407ve" and edited it to "black_f407vg" - both build but neither will load. The error I get is:

Cannot open DFU device 0483:df11 found on devnum 19 (LIBUSB_ERROR_NOT_SUPPORTED)
No DFU capable USB device available
*** [upload] Error 74
==================================================== [FAILED] Took 161.92 seconds ====================================================

Has anyone ever been successful in getting this board to work? Any idea how to fix it? I'm stumped & frustrated...

[JHolland]

Is that VID/PID combination listed in the inf file for the drivers that you installed?

[pazi88]

Does the board show up as STM32 Bootloader in boot0 mode on your PC? If not, that's your problem.

Untitled.png (35.54 KiB) Viewed 2247 times

Some cheap STM32 boards have bit wrong clock capacitor values which prevent those from going to DFU mode. In that case you should see "Unknown USB Device (Device Descriptor Request Failed)" -error on the devices.

[porschenstein]

Thanks for the responses. I had previously downloaded some drivers from STM that were pre-Cube. Those just plain don't work (at least on my computer). I finally got it working by downloading the STM32CubeProgrammer application. It contains the drivers that are necessary for Win10 to properly recognize and interface with the HW. At first, I changed the driver to "DFU" as I thought that is what would be required to upload from Visual Studio Code. But, that's not the case - it started working when I switched it back to STM Bootloader as in your responses.

[porschenstein]

I now have the base "black board" code running on my processor board. Can anyone point me to a concise list of the changes I need to make to move from 4x4 to 8x8 sequential?

I've been all-through the code and it's a little confusing with all the options. I could just start changing things, but I'd rather do it the "proper" way so I don't accidentally break anything. At first I thought that selecting 8 cylinder sequential in TS would do it, but that doesn't work. Do I need to create a new "board" definition? It seems the DropBear is the only 8x8 in the list and it's using a Teensy. It would be great if the manual were updated with a section on making this type of change.

As always, I respect the work that everyone here's put into this code base. It's pretty awesome. Once I get up the learning curve I'd like to contribute.

[porschenstein]

Hi all,

I have my STM32F407 processor board loaded with a PlatformIO build of "black_F407VE". I'm directly connecting the crank and cam inputs (60-2 - dual wheel) through a resistor-divided (scaled to 3V) V8 test rig from everythingfuelinjection.com.

In TS, I've selected the BMW PnP board and configured the ignition and ve settings. I've connected the processor board pins according to the " #elif defined(STM32F407xx)" definition in init.ino.

The RPMs are stable and I can change the RPM pot on the test rig and see the RPMs move around on the TS gauge.

However, looking at the coil outputs via logic analyzer, I can't get a reasonable or consistent firing order or dwell time. I set the dwell for 3ms but get something more like 10ms. The frequency of each coil output is around 3.4Hz and doesn't change much when I increase the RPM from 1k to 4k. Even though I have a fixed 3ms dwell in the table, the dwell changes substantially with RPM.

Is there something I'm overlooking to get this to work?

[pazi88]

Do you have battery reference voltage present? The dwell wont be correct without it.