[Trac #42] Audio driver for AD1888 - sensor input mode

[Zarro Boogs per Child]

#42: Audio driver for AD1888 - sensor input mode
---------------------+------------------------------------------------------
 Reporter:  jg       |        Owner:  dilinger
     Type:  task     |       Status:  new     
 Priority:  blocker  |    Milestone:  BTest-2 
Component:  kernel   |   Resolution:          
 Keywords:           |  
---------------------+------------------------------------------------------
Comment (by wmb at firmworks.com):

 Here are some measurements that I took on a pre-B2 board.

 Test conditions:

 MIC input.  Record gain register (index 0x1c) set to 0 (== 0dB gain), mic
 boost off (no +20dB extra gain).  Codec hipass filter off (0x1000 bit set
 in register 0x5c). DC coupled input (5536 GPIO01 high).  Vrefout (MIC
 bias) on (2.3 V open circuit). I control the input voltage with a 26K POT
 from MICin to ground, measuring the voltage with 3-digit VOM and adjusting
 POT for various voltages.  (The POT forms a voltage divider with the 3K
 MIC BIAS resistor).  Using the OFW audio driver to capture 1/2 second of
 samples at 8000 samples/sec, calculating the average value of the left
 channel samples.

 The digital output value is 0 at 1.18 V, -32768 at 0.41 V, +32768 at 1.96
 V. Over most of that range, the analog-to-digital gain is 43000
 counts/volt, and appears to be fairly linear within my measurement
 precision.

 If I change the record gain register value to 0x808 (+12 dB, a voltage
 gain of 4), the 0 point remains at 1.18 V, and the range shrinks to 0.99 V
 .. 1.37 V, which is 1/4 the previous range.

 A similar result hold for the maximum record gain setting - the range
 shrinks by the amount corresponding to the voltage gain equivalent of 22.5
 dB.

 Note that this input range (1.55 Vpp) is rather less than the AD1888 data
 sheet implies (2.83 Vpp), but that might be related to the fact that we
 are running with
 Avdd = 3.3 V instead of the 5 V that the data sheet calls out.

 Mitch

-- 
Ticket URL: <http://dev.laptop.org/ticket/42#comment:15>
One Laptop Per Child <http://laptop.org/>