Hi Miklos,

I had assumed that the code for stopping the ADC also took care of 
disabling the pull-up resistor, but I guess that's not the case. Anyway, 
thanks for the suggestions.

Sergio


On 29/08/2012 7:51 PM, Miklos Maroti wrote:
> Hi Sergio,
>
> I think what you wrote shows that the consumption might come from the
> light sensor. Of course it is not enough to turn off the ADC, you also
> have to power down the light sensor with some power pins and possibly
> you have to pull up or down the ADC pins to prevent leakage.
>
> Miklos
>
> On Thu, Aug 30, 2012 at 3:56 AM, Sergio Gonzalez <[email protected]> wrote:
>> Hi Miklos,
>>
>> I physically removed the light sensor that I was monitoring through an ADC
>> line, and the consumption dropped to 37 uA (I've also got an accelerometer
>> draining ~ 25 uA). So, for some reason the MCU won't go into low power mode
>> even if I stop the ADC. I'll dig deeper to try and see why this is
>> happening.
>>
>> Thanks,
>>
>> Sergio
>>
>>
>>
>> On 29/08/2012 10:12 AM, Miklos Maroti wrote:
>>> Hi Sergio,
>>>
>>> The timer periodically wakes the mote up (to run tasks and to maintain
>>> time), this can add up to some thing, since it wakes up 128 times per
>>> second (if I remember correctly), maybe only 4 times, depending on the
>>> prescaler. I would try to augment McuSleepC to see how deeppower state
>>> it goes into.
>>>
>>> Miklos
>>>
>>> On Wed, Aug 29, 2012 at 5:59 PM, Sergio Gonzalez <[email protected]>
>>> wrote:
>>>> Hi Janos,
>>>>
>>>> Yes, I already did that according comments of another (older) thread.
>>>> The consumption is above 10 mA when the radio is on, but when I turn off
>>>> the radio and the UART, then the consumption drops to 0.4 mA.
>>>> ActiveMessageC is included, but not wired. I've got my radio code wired
>>>> to RF230ActiveMessageC instead.
>>>>
>>>> I wonder whether there's a similar situation with the ADC and whether I
>>>> need to include SomeADCComponentC to achieve the same result. Also, I
>>>> know that at least one of the UART ports is off for sure, but I can't
>>>> tell whether the USART port that I use to program the device via ISP is
>>>> off as well.
>>>>
>>>> Thanks,
>>>>
>>>> Sergio
>>>>
>>>>
>>>> On 12-08-29 08:30 AM, Janos Sallai wrote:
>>>>> Sergio:
>>>>>
>>>>> Just a wild guess: when the RF230 is turned on, it's not in its lowest
>>>>> power state. You have to explicitly include ActiveMessageC in your
>>>>> application to allow for the RF230 init code to run, which will put
>>>>> the radio chip into its lowest power state.
>>>>>
>>>>> Janos
>>>>>
>>>>> On Wed, Aug 29, 2012 at 12:17 AM, Sergio Gonzalez <[email protected]>
>>>>> wrote:
>>>>>> Hello,
>>>>>>
>>>>>> I've been able to somewhat reduce the power consumed by an
>>>>>> ATM1281-based
>>>>>> Zigbit device by switching off the radio and the UART port (via
>>>>>> SerialActiveMessageC) during periods of inactivity, but I've been
>>>>>> unable
>>>>>> to get below 0.4 mA. I tried to explicitly disable the I2C bus as well
>>>>>> as, but there was no change. The corresponding fuses for the watchdog
>>>>>> timer, the brownout detector and the on-chip debugger are already
>>>>>> (un)programmed. I also tried lowering the MCU's clock frequency, but
>>>>>> anything other than 8 Mhz won't compile. Stopping the ADC converter
>>>>>> does
>>>>>> not seem to make any difference either. Has anybody had better luck
>>>>>> with
>>>>>> an Iris, Meshbean, or any other device with the ATM1281/RF230 combo?
>>>>>>
>>>>>> Thanks,
>>>>>>
>>>>>> Sergio
>>>>>>
>>>>>> _______________________________________________
>>>>>> Tinyos-help mailing list
>>>>>> [email protected]
>>>>>>
>>>>>> https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
>>>>>>
>>>> _______________________________________________
>>>> Tinyos-help mailing list
>>>> [email protected]
>>>> https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
>>

_______________________________________________
Tinyos-help mailing list
[email protected]
https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help

Reply via email to