paparazzi-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Paparazzi-devel] I2C startup on lisa/L


From: Christophe De Wagter
Subject: Re: [Paparazzi-devel] I2C startup on lisa/L
Date: Tue, 24 Jul 2012 23:10:09 +0200

I know a very VERY ugly fix... but it works for us:

https://github.com/tudelft/paparazzi/commits/tudelft3.9/sw/airborne/arch/stm32/mcu_periph/i2c_arch.c

It involves doing the init a few extra times with delays in between.

On Jul 24, 2012 5:30 PM, "Eric BAHIER" <address@hidden> wrote:
Hi,

I have received a new Aspirin2.0, with the lisa/L adaptator, it is realy better tha  my...... I will rework on it!!!

Bur I have some problems:

On power-up the mag (5883 on I2C) does not start one time on 5 try. 95% of the powerup test the mkk controlers does not start.

If I reset trought the jtag (reprog the cpu) no problem 100% good power up and the MAG and the MKK run correctly.

The other way is to power-up, the MKK led are on,  reset the STM32 (CC on the reset cap, C9) the MKK red leds are OFF. I have change C9  from 100nF to 1µF, the problem still present.

With logic analyzer, when the MKK fail, we can see 4 acces(to the 4 mkk adress) without ACK, then the STM32 make only READ acces to the MKK....

Probably we try to early to config/write to the MKKK, they are not "wake up", and the STM32 driver goes in an aproximative state

On lisa/M with aspirin 1.5, some time the mag hang to at power up.

Some one have seen this problems?

Eric






Le 20/07/2012 16:30, Eric BAHIER a écrit :
Hello,

I continue to work o my problemsss with Lisa/L and Aspirin 1.5like
I have solve some problems, wire connection length, and decoupling capa....

With the last GIT (V3.9) (with aspirin.c from july 19from Christophe). I have from 20sec to 5 minutes of stay in live I2C.
I have connect the logic analyzer on the I2C (SDA-SCL) and on Gyro-int and Mag-int.

The software read (@68) the  gyro when the interrupt occur, it "smell" perfect. each interrupt is followed by read cycle, then the interrupt request is clear. the interrupt remain every 1.8 ms.

Concerning the MAG (@1E) the software acces it every time, with or without interrupt.....
The MAG is able to give new measurement at 75hz (13ms), the interrupt request is present every 13ms, same as the specs.
 But the software read all the time.... evry about....400µsec, we have 100µs free time between the end a read, and the beginning of the follow on.

on lisa/M the stability is (probably) better, but some time the MAG does not run....


thank for your help,

regards

Eric



--
************************************
Lez Concept EURL

Eric BAHIER
15 chemin neuf
34 980 Montferrier sur LEZ
France

Tel : 33 4 67 59 92 53
Fax : 33 9 59 15 92 53
************************************


_______________________________________________
Paparazzi-devel mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel

reply via email to

[Prev in Thread] Current Thread [Next in Thread]