Motion Sensor reg dump

[Andy Green]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi folks -

Here is a register dump, looks okay AFAICT, there is some interesting
value in Z (ads 0x2D) when it sits on my desk suggesting it feels the
gravity of the situation ;-)

[    2.925000] LIS 0x00: 00
[    2.930000] LIS 0x01: 00
[    2.930000] LIS 0x02: 00
[    2.935000] LIS 0x03: 00
[    2.935000] LIS 0x04: 00
[    2.940000] LIS 0x05: 00
[    2.940000] LIS 0x06: 00
[    2.945000] LIS 0x07: 00
[    2.945000] LIS 0x08: 00
[    2.950000] LIS 0x09: 00
[    2.950000] LIS 0x0A: 00
[    2.955000] LIS 0x0B: 00
[    2.955000] LIS 0x0C: 00
[    2.960000] LIS 0x0D: 00
[    2.960000] LIS 0x0E: 00
[    2.965000] LIS 0x0F: 3B
[    2.965000] LIS 0x10: 92
[    2.970000] LIS 0x11: 93
[    2.970000] LIS 0x12: B2
[    2.975000] LIS 0x13: 13
[    2.975000] LIS 0x14: 13
[    2.980000] LIS 0x15: 02
[    2.980000] LIS 0x16: 06
[    2.985000] LIS 0x17: 84
[    2.985000] LIS 0x18: 13
[    2.990000] LIS 0x19: 12
[    2.990000] LIS 0x1A: 0A
[    2.995000] LIS 0x1B: 51
[    2.995000] LIS 0x1C: 00
[    3.000000] LIS 0x1D: 00
[    3.000000] LIS 0x1E: 00
[    3.005000] LIS 0x1F: 00
[    3.005000] LIS 0x20: 47
[    3.010000] LIS 0x21: 00
[    3.010000] LIS 0x22: C9  <-- CTRL_REG_3
[    3.015000] LIS 0x23: 00
[    3.015000] LIS 0x24: 00
[    3.020000] LIS 0x25: 00
[    3.020000] LIS 0x26: 00
[    3.025000] LIS 0x27: FF  <-- X
[    3.025000] LIS 0x28: 00
[    3.030000] LIS 0x29: FF  <-- Y
[    3.030000] LIS 0x2A: 00
[    3.035000] LIS 0x2B: 00
[    3.035000] LIS 0x2C: 00
[    3.040000] LIS 0x2D: 35  <-- Z
[    3.040000] LIS 0x2E: 00
[    3.045000] LIS 0x2F: 00
[    3.045000] LIS 0x30: 00
[    3.050000] LIS 0x31: 00  <-- FF_WU_SRC_1 (b6 = int asserted)
[    3.050000] LIS 0x32: 00
[    3.055000] LIS 0x33: 00
[    3.055000] LIS 0x34: 00
[    3.060000] LIS 0x35: 00  <-- FF_WU_SRC_1 (b6 = int asserted)
[    3.060000] LIS 0x36: 00
[    3.065000] LIS 0x37: 00
[    3.065000] LIS 0x38: 00
[    3.070000] LIS 0x39: 00
[    3.070000] LIS 0x3A: 00
[    3.075000] LIS 0x3B: 00
[    3.075000] LIS 0x3C: 00
[    3.080000] LIS 0x3D: 00
[    3.080000] LIS 0x3E: 00
[    3.085000] LIS 0x3F: 00

Unfortunately the INT1 and INT2 pins which are tied together and pulled
up on this design will NOT budge from 0V, one or both of them, or the
CPU input is being driven low by ~13R to 0V, which is right for a normal
N-Channel output transistor.  (Determined by pulling it up to 3.3V via a
known resistance and figuring out what the hidden resistance to 0V must
be by the change in voltage)

I changed the CPU GPF0 and GPG8 (hooked to the INT lines for the two
devices) explicitly to INPUT config temporarily in case they were
somehow configured as Output, but the net remains pulled low the same.
The only explanation I could think of that makes any sense is that the
motion sensors, in defiance of the datasheet, are pushing it low.

I also ran a test where every interrupt it cycles the value in
LIS302DL_REG_CTRL3 which controls the inversion and output type of the
interrupt pins.  If there was any correct value that took it high, it
would stick there (since no more interrupts would come to move it off).
 But the pin STILL does not go high.

Therefore I guess there is some variance of that device from the
description in the datasheet, for example the address of
LIS302DL_REG_CTRL3 given in there is wrong.

- -Andy
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.7 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org

iD8DBQFHl3goOjLpvpq7dMoRAs/KAJ4vwTNnu+ORmoQWA6ltT+8tgfiEKQCfZwV/
IL/uFsfObXf/lpKve+9GXzg=
=XCfm
-----END PGP SIGNATURE-----