Hi everybody, yesterday I got my Primer 2. It was perfectly working, I tried the applications.
Today I tried the supplied sample application "toggle", I suppose its scope was to make the two upper leds blink at different rates, and it was OK.
Then .... when trying again the same demo, I got an error from the Rite 7 stating that there was a communication error, that voltage was at 1.1V ... something like "error 103" ...
Now the unit is dead. - If I connect it to the PC using the debug connector it's correctly recognized as a WinUSB device - if I try to switch it on I get only a rapid blink of the display, then the upper red led blinks rapidly for a few seconds but with low energy (the light is very little with respect to normal) ... then nothing, everything stops - I tryed to charge again the battery but it's almost full, I measure it and it's OK. - I also tryed to reprogram the unit to factory default ... but the prolem is always the same "error 1an I do now?03: target power failure, measured voltage is 1.000000V". What can I do now?
Please download the schematic. There are two regulators L6928D. One is set at 2V8 (for the digital devices), and the second at 3V1 (to power the backlight and the loudspeaker). Please check the voltage for both of them. It could be that the regulator U9 (2V8) is dead. Please confirm by measuring the voltage at C20.
I'm having the same problem with my Primer2, I loaded the toggle program from the Examples/ARM/Primer2 folder and it worked for a little while, then I get this same error. I'm going to test my unit and take a look at the regulator in question... did you ever resolve this with your unit?
We received at Raisonance the following message from 'dzingoni'. It could hellp some other users:
<<Thank you for your reply. Yes, there are news ... after trying for hours to revive the Primer 2 I finally succeeded. I thought that the problem was during the startup phase because after a "false start" the Primer 2 actually did not start (low power red led blinkng fast ... and then nothing). It was impossible to have the unit ON when launching commands to re-program it or to download software. So I tried to start a "debug" of a demo application just after pressing the button on the unit (during the false startup phase in other words ....) and miraculously the Ride7 was able to download the software. Then I reprogrammed the unit to factory setting ... and now it works .... The problem is: am I going into the same troubles if I try to install some other app or debug ? As a further information please note that the "low voltage" error appeared the first time when I first tried the unit ... Initially I thought there was a problem with the Ride7, so I reinstalled it, then I thought that maybe I was using the wrong demo (maybe one for the ST7 ...) .... the error appeared for some time, then a "debug start" went OK .. and then the lock ... Looks like some problem is present in the driver for the auxiliary microcontroller. My suggestion, in any case, would be to modify the driver software in order to be able, in the extreme case, to bypass the voltage level check and to force the reprogramming of the unit.>>
The problem could be that the application shuts down the power supply by clearing PC13. If the application starts by shutting down the power supply, the embedded RLink cannot controls the Cortex M3. The only solution is to keep the button pressed until the debug session is launched and the FLASH erased.
I open an example project for Primer2 with RIDE, it compiles without errors, it flashes without errors. But the Primer2 freezes immediatly. Now I have to remove the battery jumper, remove usb.
When I insert the jumper and press the button I get either nothing (screen stays black) or it flashes one time and nothing works. I only get an error about voltage with varying values (1.100, 1.2000, 1.0000).
The only way to get the primer back to live is this procedure: remove jumper, unplug usb. plug usb. flash factory firmware. while flashing insert jumper. Now the unit is alive again.
What is odd about it I cannot flash the firmware with the battery jumper inserted. But I also cannot remove the jumper for adding an example program, because the Primer2 shuts down.
So I am somewhat diappointed by the device as I did not get anything working till now :-( (.... and god do I miss a JTAG connector)
GREAT!! Now I can reproduce what you report... I launch the batch file: \Program Files\Raisonance\Ride\Lib\ARM\CircleOS\Restore_Circle_Factory.bat and I see exactlele what you report... This batch reload the Primer1 hex files that are NOT compatible (only the source files are compatible since the pinning in not the same at all..).
To restore the orginal program, you have to launch another batch file: \Program Files\Raisonance\Ride\Lib\ARM\CircleOS\Restore_Primer2_Circle_Factory.bat
The problem is that the hex file for Primer1 switches OFF the power supply of the Primer. The solution consists in keeping the button pressed while executing this batch.
Conclusion: 1. The binary files of the two CircleOS primers are not compatible 2. but the CircleOS applications (maze, breakout, ...) are binary compatible.
I used this Restore_Primer2_Circle_Factory.bat to restore factory settings, that works. I did not try \Restore_Circle_Factory.bat.
But I came around an other strange thing: I Downlod the mems-demo frm this site. It says, it is compatible to primer1 and primer2. The project settings are on primer1 by default. With these settings - primer 1 - it compiles. Flasing it freezes the primer2 (s.a.).
When I set target to primer2, it compüiles but link stage failes:
crt0_STM32x.o: In function 'ResetHandler': crt0_STM32x.c:211: undefined reference to 'main'
The binary code and the source code work on both Primer1 and Primer2. However, the projects are set for Primer1. When you launch from RIDE a debug session, CircleOS is automatically loaded (the wrong one in this case). If you wish to debug such an application, you must create a new project, select a Primer2 application. It should work.
To load an application that work for both Primer1 and Primer2, you can use circle_mgr.exe instead of loading the RIDE project. It will work with the same binary code (and the executable reccognizes automatically the version of the Primer.
I see that there are some news about the issues of low battery and false startup, and that probably a "weak connection" exist for ground on the boards near the regulator ... I understand that at Raisonance you are experimenting a solution, based on adding a small wire to reinforce the current path .... Is it possible to have detailed instructions to do that at home? I think it would be preferable for most just to add this wire before the regulator gets eventually burned .... Thank you. And Happy New Year to everybody!
The following picture shows the wire to enforce the ground on U9. It has been applied on 90% of the first batch.. (the first 10% were already shipped) :
It is recommended to make this change. We have seen only a few Primers with a regulator died but we don't know how many are really concerned.
The additional wire is a 'T', as the original track that is too thin and creates a parasitic inductance.
I have a similar problem, which I cannot solve. I'm using STM32 Primer2 with CircleOS version upgraded to 3.5.
I cannot debug the example application toggle_with_CircleOS. When I try, Ride7 states that ther's a communication failure ans CircleOS dies. Then I must follow the recovery procedure below (execute Restore_Primer2_Circle_Factory.bat, while pressing joystick button to avoid error 103).
If, instead of reprogramming CircleOS I try to debug the application again, Ride7 complains about 'OPI Driver: Target power failure'.
When CircleOS is loaded and working, I cannot list applications using Circle_Mgr.exe L. This give me another 'error 103: Error reported by the DTC'.
So what can I do to load some test (and my own) applications? (I cannot download the picture linked about, and I don't know if it is related to my problem)
If you try to debug a 'Primer1' project, CircleOS for Primer1 will be reloaded and will switch off the power supply. You have to use a 'Primer2' project. Now, this problem has been fixed on the new CD-image (downloadable from this web site) . The new CircleOS for Primer1 checks the Primer version, and does not switch off when Primer2 is detected (it loops for ever in this case).
is for Primer1. That causes to switch off the power supply. ¿Is there an application to start working with CircleOS? I was following the STM32Primer-Manual. I will download and install the new CD-image.
But, even when Factory CircleOS is loaded, why CircleOS_mgr L fails? This is the error message:
... Reading FAT table... App0: Name=Maze, Addr=0x08006000, Size=8KB.... !!! Error 103: Error reported by the DTC
