Link is unstable

[hossein]

The error "Programming @:..." just happned because of "^M" at the end of values, its becasue of editing file with vi editor.
After uploading bin/hex to stm board, the link is going to be unstable and as you can see in attachment gif

i can not work at all.
This is happening when i upload compiled file - that i got from site - too.

[AndreyB]

I've forked your message from the other thread since now this has nothing to do with "how to compile", let's stay on topic please.

I do not see anything attached? I am not sure what's going on.

[hossein]

I was trying to make a shorter gif or video to upload, and i failed.
This is the video http://79.127.54.245/stm32f.mp4

[AndreyB]

This site can’t be reached. 79.127.54.245 took too long to respond.

Would youtube be an option?

[kb1gtt]

I found the download was very slow. After 5 minutes it still claimed it was a 2hr download. So I aborted.

Something like Youtube would allow for multiple downloads with out bandwidth limiting the server. I have fast internet, and I'm very certain the slow was a server issue.

[hossein]

I'm so sorry for low bandwidth.
This is link:
[video]https://www.youtube.com/watch?v=oTKsmFdyY7M[/video]

[kb1gtt]

-- What is your hardware? Are you using just a OEM discovery board?
-- Are you using an OEM board or a lower cost alternative from eBay or other sources?
-- I'm assuming that's Windows emulation for ST link, which seems to function well. Can you install TS in that emulation and see if that is more stable?
-- Do we have reason to believe TS is stable in your distro of linux?

[hossein]

I'm using discovery board + frankenstein board
MS Windows is running on on a server physically connected to stm board and it was rdp. TS was/is stable before.
I can fire and run an engine of ford festiva with this board+ts. (last week)
[video]https://www.youtube.com/watch?v=EwkogYZi4BM[/video]
Reason of disconnecting - that just you'll see in clip - was a low quality of usb cable.
And not about past post.

[AndreyB]

Wow, you engine is running? Can you start a thread about it?

Is your link issue resolved now with a new cable?

[kb1gtt]

Cool to see your engine running. Also good to hear you had it working stable with a native windows PC. Having the hardware working stable with any system removes several concerns about hardware, as you mentioned a bad USB cable. I do not know how TS determines a loss in link, I also do not know if Linux is constantly connecting the USB and disconnecting it, or anything like that. Perhaps look at your Linux logs to see if you notice excessive connects and disconnects of the USB. If that is not the situation, then perhaps there is a latency issue with the Linux drivers which is causing TS to drop the link. Also it could simply be a buggy Linux driver. Perhaps try a different USB cable, it might have a different driver and might be less buggy. Some commands that might help track down driver issues include the below.
lsusb
dmesg
depmod -a
lspci
lsmod

[hossein]

Thank you for creating a wonderful tool
Yes i shared crankshaft data between my ecu and original ecu with two 220 n caps.
I make sparks and car injection.

Cool to see your engine running. Also good to hear you had it working stable with a native windows PC. Having the hardware working stable with any system removes several concerns about hardware, as you mentioned a bad USB cable. I do not know how TS determines a loss in link, I also do not know if Linux is constantly connecting the USB and disconnecting it, or anything like that. Perhaps look at your Linux logs to see if you notice excessive connects and disconnects of the USB. If that is not the situation, then perhaps there is a latency issue with the Linux drivers which is causing TS to drop the link. Also it could simply be a buggy Linux driver. Perhaps try a different USB cable, it might have a different driver and might be less buggy. Some commands that might help track down driver issues include the below.
lsusb
dmesg
depmod -a
lspci
lsmod

Actually I am *nix expert (GNU/Linx& FreeBSD). And yes it will be correct in case of no suitable driver or unloaded kernel; module and etc ...
And i ran it on FreeBSD and Debian GNU/Linux and windows successfully.
The usb cable issue: The connector in this video is damaged ...
and now ! about hanging out !

Code: Select all

Aug  6 22:52:50 virtual kernel: ugen0.5: <STMicroelectronics ChibiOSRT Virtual COM Port> at usbus0 (disconnected)
Aug  6 22:52:53 virtual kernel: ugen0.5: <STMicroelectronics ChibiOSRT Virtual COM Port> at usbus0
Aug  6 22:52:53 virtual kernel: umodem0 on uhub0
Aug  6 22:52:53 virtual kernel: umodem0: <STMicroelectronics ChibiOSRT Virtual COM Port, class 2/0, rev 1.10/2.00, addr 2> on usbus0
Aug  6 22:52:53 virtual kernel: umodem0: data interface 1, has no CM over data, has no break
Aug  6 22:54:07 virtual kernel: ugen0.5: <STMicroelectronics ChibiOSRT Virtual COM Port> at usbus0 (disconnected)
Aug  6 22:54:07 virtual kernel: umodem0: at uhub0, port 3, addr 2 (disconnected)
Aug  6 22:54:07 virtual kernel: umodem0: detached
Aug  6 22:54:10 virtual kernel: ugen0.5: <STMicroelectronics ChibiOSRT Virtual COM Port> at usbus0
Aug  6 22:54:10 virtual kernel: umodem0 on uhub0
Aug  6 22:54:10 virtual kernel: umodem0: <STMicroelectronics ChibiOSRT Virtual COM Port, class 2/0, rev 1.10/2.00, addr 30> on usbus0
Aug  6 22:54:10 virtual kernel: umodem0: data interface 1, has no CM over data, has no break
Aug  6 23:05:14 virtual kernel: ugen0.5: <STMicroelectronics ChibiOSRT Virtual COM Port> at usbus0 (disconnected)

These lines are my machine's message logs and was happened after upgrading to latest bin image that i downloaded from http://rusefi.com/build_server/separate_files/
and even my bin/hex file that i compiled it.
I think the link is stable now, but something in stm goes wrong.
This will happened in any Operating system (Ms windows | GNU/Linux | BSD | etc ) and the bug depends on whatever executed in STM board and i do not know what it is !!.

[hossein]

Hello guys.
At these links i will describe my issue. And sorry if my English is not good.
So The fist link is how to compiled the source code:
https://www.youtube.com/watch?v=H3ZQDkpKwQ
As you can see there is no problem and done perfectly !!!

in link two we have official image and issue :
https://www.youtube.com/watch?v=SYUnAqOBgLg

At least you can see the old version - i guess it is for 2016 - works fine ....
https://www.youtube.com/watch?v=BJxnYk_bD98

Notice: I did all tests on stm32f discovery board - just one board and yes its original and not fake -
Thanks

[kb1gtt]

I seem to recall a problem with the Discovery USB. Can you flash your discovery firmware? See attached picture.

Is there a wiki page for this firmware update thing? Should I create one?

[kb1gtt]

I just added some clarity to this wiki page. http://rusefi.com/wiki/index.php?title=Manual:Software:User/en#Load_firmware It's probably still garbage, but better than it was. The picture helps make it jump out that you need to do it.

[hossein]

Hi, The problem seems is not solved.
I can not connect with the TS.

[kb1gtt]

https://www.youtube.com/watch?v=H3ZQDkpKwQ

For this video, I get an error message that the video does not exist. Something about error occurred during validation.

Also can I confirm that you are using the discovery's built in serial port. The Frankenso board has a FTDI chip which is completely different than the discovery's built in hardware.

If using the discovery's VCP COM port driver instead of the Frankenso's FTDI COM port, I know the VCP driver needs to be installed separately. The windows auto-detected driver typically does not work properly. See link to VCP driver at the below wiki page.
http://rusefi.com/wiki/index.php?title=Manual:Software:User/en#While_you_wait_for_the_box

I do not know your basic setup. It would help if I better understood your general setup. Some other miscellaneous questions for a general understanding of what you are doing. Please answer these questions. Are you using WINE or is that a virtual windows? What version of windows is that? Is it 32 bit, or 64 bit?

Hmmm, you say an old version of TS worked but a new version does not. Have you updated your TS to the new version? I'm expecting an new TS will not work with an old rusEFI firmware. I see you are running TS 3.0.18, and the current TS is 3.0.22. Please update your TS to the current version.

Do you experience problems with the console or is this just a TS issue? If the console works, and TS does not then we know this is a TS issue. However if the console also does not work, then we know this is a driver issue. Also the console provides lots of good diagnostics information. A screen shot of the console would be helpful, as it indicates what version of rusEFI you are running, etcetera. See console at this link.
http://rusefi.com/wiki/index.php?title=Manual:Software:User/en#rusEfi_Console

[hossein]

That's my mistake ! Bad description ....
this what really my setup is :
1 stm board + Frankenstein
A server with MS Windows 8 R2 data center that connected to my stm board and i upload program in to the stm by that.
Some Linux & BSD machines that i have TS and Ruesfi console on theme including a laptop -debian - to test.

All of these machine are up to dated OS and TS and.
I use stm32link v4, upgraded firmware as you told.
And nothing is changed but triggering crankshaft is off and won't working anyway.
I mean after upgrading firmware from st-link menu and uploading the latest image i could not trigger ecu as i did before,
I changed trigger type to all of them and even custom with dummy 2 and 1 skip !
And nothing happened, and the ecu is dead .... -i hope not -
Rusefi console is working but TS ....
And suddenly i lost my ecu

[kb1gtt]

All of these machine are up to dated OS and TS

This does not appear to be accurate. Your video shows TS is at 3.0.18, and the current version is 3.0.22 Please update TS to 3.0.22.

Can you get this to work with discovery board only?

Can you get this to work with the console?

Can you post a picture of the console?

[kb1gtt]

Also please note, we expect that the current version of rusEFI firmware will NOT function with TS 3.0.21 and earlier. It has to be 3.0.22 or newer. TS 3.0.22 came out around May 5th of 2017.

[kb1gtt]

I just updated the below wiki page to note the 3.0.22, it had an older version noted, which in recent months has become wrong.
http://rusefi.com/wiki/index.php?title=Manual:Software:User/en#Tuner_Studio

[hossein]

This does not appear to be accurate. Your video shows TS is at 3.0.18, and the current version is 3.0.22 Please update TS to 3.0.22.

snapshot526.png (105.53 KiB) Viewed 18815 times

Can you get this to work with discovery board only?

I did it before exactly.

Can you get this to work with the console?

Yes ! as i wrote before yes it is working with rusefi console but the crankshaft data is not working .
I triggered it with my car's flywheel& and i have nothing.

Can you post a picture of the console?

snapshot527.png (349.97 KiB) Viewed 18815 times

Thanks for replies

[puff]

Hey!
The firmware on rusefi console screenshot seems to be outdated and incompatible with TS, doesn't it?
Try flash-erase, upload the latest firmware, download the latest rusefi console (check connection), create a new project in TS - this might help.

[kb1gtt]

I'm not sure I understand what the problem currently is. As puff noted you might have an old version of the rusEFI firmware. Please note that when you get the bundle, you need to use the console and such that come with that bundle. You cannot mix new and old pieces of a bundle.

I have just updated this wiki page which helps convey what information would be helpful for you to post. Much of this could have been much quicker if we had the screen shots of TS and console sooner.
http://rusefi.com/wiki/index.php?title=Manual:Support#Support_option_1_---_forum_request

Is your problem that the console connects and functions, but TS does not connect? Or has the problem changed? I suspect your problem with setting the crank and other items is that you are not using all the pieces from the same bundle. Try using the firmware, console, and TS files from the same bundle. It would be best if you use the most recent bundle.

It's not fully clear to me what the current problem is.

[kb1gtt]

Are you all set now? Do you still have a link stability problem?

[hossein]

Yes i did some upgrades .
I had to repair/fix some issues before starting again.
I fried max9926 Unintentionally by ringing the cell phone, the cellphone had too many missed calls and it was next to a gearbox sensor.
- this is what happened to my ecu, I'm trying to find out how is it possible and i will go to the next step, i think what way of blocking harmful noises -
This is the only reason for Lack of accountability to triggers .
I checked it out all parts are working fine but i can not do any trigger, i injected the signals to A5 port by an arduino and stm works fine.

Thank you, all of you

[kb1gtt]

Hmmm, I would not expect issues with the MAX9926. Did you have R111 installed? What version of Frankienstien do you have?

I seem to recall cell phones operate at very low power, and that the diodes on the MAX9926 are fairly beefy. I would expect the 5V would fail before the MAX9926 fails. I suspect you are fighting something different. Can your arduino simulate 0V to 5V to the MAX9926?

[hossein]

Yes , R11 5k is installed.
I'm using Frankienstien v 0.11.
I had checked the output on jumpers, and i have 5 volts stable with no change even with triggering max9926/7.
Am i right ? : the voltage might be down and up when triggering .
Yes but i didn't yet, i will try it with arduino.

[hossein]

There is no significant change even with arduino signals in max99, but the 3.3 volt on pA5

[kb1gtt]

We are getting a bit off topic. Do you have engine thread where we continue to chat about it a bit more on topic?

Do you have a scope? Can you analyze your analog signals? To me it sounds like you are not getting the signal to the MAX9926 chip.