[OLPC-devel] making DC measurements with "mic in", progress

[Miller Puckette]

Hi Jaya,

On Sun, Jun 18, 2006 at 08:02:30AM -0400, Jaya Kumar wrote:
> On 6/17/06, Miller Puckette <mpuckett at imusic1.ucsd.edu> wrote:
> >Hi OLPC developers (particularly Mark perhaps)
> >
> >Simon and I have got as far as to verify that, with VREFD asserted on the
> >AD1888 codec, whatever voltage we put on the MIC input actually gets to
> >pin 21 of the AD1888 chip.  I think this implies among other things that
> >we are indeed using the "modded" board (in fact Jim found the solder blob).
> > We only made two measurements, 0V and 2V, but that seemed to make the case
> >that we've made it to the codec.
> 
> Cool.
> 
> >
> >Unfortunately, we still can't seem to get DC results out of the CODEC; 
> >some-
> >body further downstream is high-pass filtering the signal.  I doubt ALSA
> 
> This might be a silly question. I interpret what you said about the
> step function earlier to mean you are driving _down_ from 2V to 0V and
> you are seeing a dying exponential. On the scope, when probing pin 21
> as you drive your step signal, do you see the step waveform itself?
> Just to double check, when you measure resistance through the short
> you've made, it's ~0 right. I'm just wondering if the cap that you've
> shorted with the blob is discharging resulting in the dying
> exponential. Even so, the schematic says it's 1uF so it'd be hard to
> get an RC of 0.5s. If you stick a 1Hz sine wave in, does that come out
> looking okay?

It's the classic: when you drop the input voltage (step down) , you see a 
negative-valued dying exponential; when you raise it (step up) you get a
positive-valued one.  Always asymptotes back to about zero.

Since I'm trying DC signals, I don't think anti-aliasing caps, or anyhting
short of a cap right in the signal path (not shunting off from it) would
have this kind of effect.  It's hard to imagine any such thing in the CODEC
itself, of course.

I don't have a sine wave generator handy but will improvise something and
try that later today.

> 
> >would do such a thing and am only hoping there's another setting we need
> >to change in the codec.  I'm going to start looking through the codec doc
> >to see if I can find anything, but if anyone else can think of something,
> >I'd be eager to hear it.
> 
> 
> I'm wondering if the ADC antialiasing filter caps with the 270pF on
> pin 29 and 30 on page 13 have any effect.
> 
> Other than that, I can't think of anything concrete. I assume you're
> setting the sampling rate from ALSA and not writing those regs
> directly. Maybe try varying the rate to see if it has any effect if
> you haven't already.
> 
I haven't tried that and will, also will just repeat everything and hope
to notice something useful.  
> jaya