Bug 12754 - Knob stops working after a while when running on battery (was: UI sluggish and music stutters / stops when running from battery)
: Knob stops working after a while when running on battery (was: UI sluggish an...
Status: RESOLVED WONTFIX
Product: SB Radio
Classification: Unclassified
Component: Product Quality
: Include FW version in comment
: PC Other
: P3 enhancement (vote)
: 7.4.0
Assigned To: Caleb Crome
:
Depends on:
Blocks:
  Show dependency treegraph
 
Reported: 2009-07-09 05:48 UTC by Felix Mueller
Modified: 2009-09-08 09:29 UTC (History)
0 users

See Also:
Category: ---


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Felix Mueller 2009-07-09 05:48:58 UTC
Running Baby from battery (the one Rico gave me during last on-site) makes the UI unresponsive and music breaks up or stops completely.

As soon as external power is reapplied everything goes back to normal.

This happens on a PB1 and PB2 with r6430.

Caleb guessed: "It must be that the MCU is continually interrupting the CPU.

BTW: My first PB1 still had r5974 installed and it did not happen with that firmware revision.

Top output while on battery: ([events/0] uses a lot of CPU)

Mem: 36772K used, 24916K free, 0K shrd, 6628K buff, 11544K cached
CPU:   7% usr  39% sys   0% nic  33% idle   0% io  18% irq   0% sirq
Load average: 1.03 0.35 0.12 1/64 989
  PID  PPID USER     STAT   VSZ %MEM %CPU COMMAND
   15     2 root     SW<      0   0%  22% [events/0]
  914     1 root     S    22956  37%  21% /usr/bin/jive
   83     2 root     SW<      0   0%  11% [IRQ-4]
  302     2 root     SW<      0   0%   6% [IRQ-182]
  976   915 root     R     2696   4%   2% top
    9     2 root     SW<      0   0%   2% [sirq-tasklet/0]
    5     2 root     SW<      0   0%   1% [sirq-timer/0]
  903     2 root     SW<      0   0%   1% [ksdiorqd]
  876     2 root     SW<      0   0%   0% [AR6K Async]
  310     2 root     SW<      0   0%   0% [IRQ-9]
   41     2 root     SW<      0   0%   0% [IRQ-34]
  908     1 root     S     1892   3%   0% /usr/sbin/wpa_supplicant -B -Dwext -ie
  340     2 root     SW<      0   0%   0% [IRQ-32]
  866     1 root     S     2772   4%   0% /usr/sbin/inetd
  915     1 root     S     2696   4%   0% /bin/sh
  916     1 root     S     2696   4%   0% /sbin/getty tty3 9600 VC vt100
  954     1 root     S     2696   4%   0% udhcpc -R -p /var/run/udhcpc.eth1.pid
    1     0 root     S     2692   4%   0% init
  784     1 root     S     2692   4%   0% /sbin/klogd
  781     1 root     S     2692   4%   0% /sbin/syslogd -S
Comment 1 Caleb Crome 2009-07-09 10:03:06 UTC
Can't reproduce with MSP fw 185 on a PA5 board.  

Will move this to an MP bug.
Comment 2 Felix Mueller 2009-07-10 08:34:43 UTC
Re-testing with r6491 / MSP fw 185 on PB1 and PB2 doesn't show the sluggishness or music break up anymore, but after running from battery for a couple of minutes I now see the knobs (and sometimes all buttons and IR) stop working. When this happens I sometimes see the following serial console output:

i2c-adapter i2c-1: ACK not received
i2c-adapter i2c-1: ACK not received
i2c-adapter i2c-1: ACK not received
i2c-adapter i2c-1: ACK not received
msp430 i2c bus failed

Reapplying external power doesn't recover. Only a reboot helps. Battery voltage is about 13.2 volts.


more serial console output:

GOING DOWN NOW!
i2c-adapter i2c-1: ACK not received
i2c-adapter i2c-1: ACK not received
i2c-adapter i2c-1: ACK not received

i2c-adapter i2c-1: Could not grab Bus ownership
i2c-adapter i2c-1: Bus Arbitration lost
i2c-adapter i2c-1: Could not grab Bus ownership
i2c-adapter i2c-1: Bus Arbitration lost
i2c-adapter i2c-1: Could not grab Bus ownership
i2c-adapter i2c-1: Bus Arbitration lost
i2c-adapter i2c-1: Could not grab Bus ownership
msp430 i2c bus failed
i2c-adapter i2c-1: Bus Arbitration lost
Comment 3 Richard Titmuss 2009-07-27 01:09:44 UTC
Reset priority before triage.
Comment 4 Caleb Crome 2009-07-30 09:12:31 UTC
This is due to a hardware bug with PB1 and PB2 units (msp supply voltage too low).  The PA5/MPQ should not show this problem at all.

The "i2c-adapter i2c-1: ACK not received" is the smoking gun.

Once I get the other MSP stuff out of the way, I can rework the MSP code to be more graceful on the PB1 and PB2 units, but I'm not sure the effort is worth it.

-Caleb
Comment 5 Caleb Crome 2009-07-30 09:15:36 UTC
FYI, you should be able to test this by lifting pin 3 (Vin) of U39, and connecting that pin to the + side of C255 or C254.

-C
Comment 6 Felix Mueller 2009-07-30 14:24:13 UTC
Hi Caleb

I just wanted to let you know that your proposed fix in comment #5 indeed fixes the issue. Tested on a PB1 Baby.

And I agree, it's possibly not worth fixing, except if we need people testing with batteries on PB1 and PB2.

Felix
Comment 7 Caleb Crome 2009-08-12 09:15:05 UTC
This is a hardware bug that's fixed in MPQ/MP boards.

We can't fix it.