Hi All,
Whilst testing a regulator machine config I've run into a weird issue.
The preliminary data sheet I have suggests that there are two enable
registers
for ldo11 but neither rctl22 or rctl11 seems to have any effect (on
their own
or both cleared together). I can change the voltage just fine, but not
actually turn it off.
Anyone come across this problem or have any idea what is going on?
Thanks,
Jonathan
On Thu, 2009-06-25 at 15:39 +0000, Jonathan Cameron wrote:
> Hi All,
>
> Whilst testing a regulator machine config I've run into a weird issue.
>
> The preliminary data sheet I have suggests that there are two enable
> registers
> for ldo11 but neither rctl22 or rctl11 seems to have any effect (on
> their own
> or both cleared together). I can change the voltage just fine, but not
> actually turn it off.
>
> Anyone come across this problem or have any idea what is going on?
I've come across something similar in the past with two enable bits
causing I2C register cache coherency issues. The cache bits became out
of sync with reality and stopped some I2C writes happening.
Another option is that your hardware config doesn't allow ldo11 to be
disabled ?
HTH
Liam
Liam Girdwood wrote:
> On Thu, 2009-06-25 at 15:39 +0000, Jonathan Cameron wrote:
>> Hi All,
>>
>> Whilst testing a regulator machine config I've run into a weird issue.
>>
>> The preliminary data sheet I have suggests that there are two enable
>> registers
>> for ldo11 but neither rctl22 or rctl11 seems to have any effect (on
>> their own
>> or both cleared together). I can change the voltage just fine, but not
>> actually turn it off.
>>
>> Anyone come across this problem or have any idea what is going on?
>
> I've come across something similar in the past with two enable bits
> causing I2C register cache coherency issues. The cache bits became out
> of sync with reality and stopped some I2C writes happening.
Don't thing that could happen here.
>
> Another option is that your hardware config doesn't allow ldo11 to be
> disabled ?
Other than the obvious possibility that the hardware has a bug, I can't
seem to find a reason why this shouldn't work.
Thanks,
Jonathan
Jonathan Cameron wrote:
> Liam Girdwood wrote:
>> On Thu, 2009-06-25 at 15:39 +0000, Jonathan Cameron wrote:
>>> Hi All,
>>>
>>> Whilst testing a regulator machine config I've run into a weird issue.
>>>
>>> The preliminary data sheet I have suggests that there are two enable
>>> registers
>>> for ldo11 but neither rctl22 or rctl11 seems to have any effect (on
>>> their own
>>> or both cleared together). I can change the voltage just fine, but not
>>> actually turn it off.
>>>
>>> Anyone come across this problem or have any idea what is going on?
>> I've come across something similar in the past with two enable bits
>> causing I2C register cache coherency issues. The cache bits became out
>> of sync with reality and stopped some I2C writes happening.
> Don't thing that could happen here.
>> Another option is that your hardware config doesn't allow ldo11 to be
>> disabled ?
> Other than the obvious possibility that the hardware has a bug, I can't
> seem to find a reason why this shouldn't work.
>
I didn't meet this problem, Haojian & Paul, ever had this problem before?
Or try lookup the Errata for the Arava chip.
Eric Miao wrote:
> Jonathan Cameron wrote:
>> Liam Girdwood wrote:
>>> On Thu, 2009-06-25 at 15:39 +0000, Jonathan Cameron wrote:
>>>> Hi All,
>>>>
>>>> Whilst testing a regulator machine config I've run into a weird issue.
>>>>
>>>> The preliminary data sheet I have suggests that there are two enable
>>>> registers
>>>> for ldo11 but neither rctl22 or rctl11 seems to have any effect (on
>>>> their own
>>>> or both cleared together). I can change the voltage just fine, but not
>>>> actually turn it off.
>>>>
>>>> Anyone come across this problem or have any idea what is going on?
>>> I've come across something similar in the past with two enable bits
>>> causing I2C register cache coherency issues. The cache bits became out
>>> of sync with reality and stopped some I2C writes happening.
>> Don't thing that could happen here.
>>> Another option is that your hardware config doesn't allow ldo11 to be
>>> disabled ?
>> Other than the obvious possibility that the hardware has a bug, I can't
>> seem to find a reason why this shouldn't work.
>>
>
> I didn't meet this problem, Haojian & Paul, ever had this problem before?
> Or try lookup the Errata for the Arava chip.
Hi Jonathan,
Here's a hint, check register 0x1b - which looks suspicious to me to have
some side effects to LDO11.
Eric Miao wrote:
> Eric Miao wrote:
>> Jonathan Cameron wrote:
>>> Liam Girdwood wrote:
>>>> On Thu, 2009-06-25 at 15:39 +0000, Jonathan Cameron wrote:
>>>>> Hi All,
>>>>>
>>>>> Whilst testing a regulator machine config I've run into a weird issue.
>>>>>
>>>>> The preliminary data sheet I have suggests that there are two enable
>>>>> registers
>>>>> for ldo11 but neither rctl22 or rctl11 seems to have any effect (on
>>>>> their own
>>>>> or both cleared together). I can change the voltage just fine, but not
>>>>> actually turn it off.
>>>>>
>>>>> Anyone come across this problem or have any idea what is going on?
>>>> I've come across something similar in the past with two enable bits
>>>> causing I2C register cache coherency issues. The cache bits became out
>>>> of sync with reality and stopped some I2C writes happening.
>>> Don't thing that could happen here.
>>>> Another option is that your hardware config doesn't allow ldo11 to be
>>>> disabled ?
>>> Other than the obvious possibility that the hardware has a bug, I can't
>>> seem to find a reason why this shouldn't work.
>>>
>> I didn't meet this problem, Haojian & Paul, ever had this problem before?
>> Or try lookup the Errata for the Arava chip.
>
> Hi Jonathan,
>
> Here's a hint, check register 0x1b - which looks suspicious to me to have
> some side effects to LDO11.
>
Thanks Eric,
I'll have a play and see if this reg (start up control) effects the ability
to turn it off at other points.
Jonathan