[Lowfer] WOLF Sampling Rate Cal

[John Andrews]

All,

Something to do while on vacation stuck in the woods with wet weather...<g>

I have posted a new article on an easy way to find your sound card's 
actual sampling rate at:
http://www.w1tag.com/WOLFSamp.htm

Here's the idea. Several of us have suggested that an AM receiver tuned 
to Loran-C on 100 kHz provides an excellent spectrum of accurate audio 
frequencies which can be used to calibrate sampling rates. The problem 
is that your rate may be SO far off that you wind up using the wrong 
Loran harmonic. My approach is now to do the cal in two steps: an 
approximation using a very low audio frequency, and then a fine-tuning 
using one of the harmonics. Any comments on the article would be 
appreciated.

For those who may be unconvinced about the need to nail the sampling 
rate for WOLF, here's an example:

My laptop has an actual sampling rate of 11098.786 Hz instead of the 
nominal 11025 Hz. Here's part of a WOLF run on 185.800 kHz with a weak 
signal, using the correct rate:
2008-08-05 19:16:08 >WOLF  -r 11098.79 -f 700.170 -t 1 -d 16 -w 0.0000
t:  24 f: 0.003 a: 0.0 dp: 60.0 ci:12 cj:120 TAG RAYMOND ME  -
t:  48 f: 0.001 a: 0.2 dp: 60.8 ci:12 cj:120 TAG RAYMOND ME  -
t:  96 f: 0.001 a: 0.2 dp: 61.2 ci:12 cj:120 TAG RAYMOND ME  -
t: 192 f: 0.000 pm:18.06 jm:720 q:  3.5  2.6 TAG RAYMOND ME  -

Now, if I change the rate to the nominal 11025 Hz, and lower the 
receiver tuning to 185.795 kHz to get the signal back in the passband, 
the following results:
2008-08-05 18:46:26 >WOLF  -r 11025 -f 700.170 -t 1 -d 16 -w 0.0000
t:  24 f: 0.312 a:-0.9 dp: 63.9 ci: 8 cj:  0 IUGISRYOZPVHPRF ?
t:  48 f: 0.312 a:-0.9 dp: 63.3 ci: 4 cj:324 B33I5SPSXY45J4J ?
t:  96 f: 0.312 a:-0.9 dp: 62.3 ci: 7 cj:325 IYAHR4IM/2KEHH5 -
t: 192 f: 0.312 pm:1.974 jm:314 q:-12.3 -8.5 XMNE*OAEHZMCPQO ?
t: 288 f: 0.312 pm:2.533 jm:319 q:-11.5 -8.4 D*CVQ07UFR4FYH* ?
t: 384 f: 0.312 pm:2.590 jm:319 q:-11.2 -6.1 TWU8DTIFJNF/H9H -
t: 480 f: 0.312 pm:2.617 jm:319 q:-10.3 -7.8 D*E1YG1QUBSAVZU ?
t: 576 f: 0.312 pm:3.030 jm:319 q: -9.7 -8.0 H YKU7UJGKQYTSB ?
t: 672 f: 0.312 pm:3.054 jm:319 q: -9.0 -7.8 .KRK0J *KTMVL4H ?
t: 768 f: 0.312 pm:3.091 jm:319 q: -8.6 -7.6 U D6FLDZ6X15E/8 ?
t: 864 f: 0.312 pm:3.221 jm:361 q: -8.2 -7.6 CLCAUBP48BXX1HG ?
t: 960 f: 0.312 pm:3.478 jm:341 q: -7.5 -7.8 FUCV9734SEY7UP4 ?
t:1056 f: 0.312 pm:3.574 jm:341 q: -7.1 -8.3 34H*M7GW. /K57X ?
t:1152 f: 0.312 pm:3.576 jm:341 q: -6.7 -8.1 /X0358UEJ9GCK30 ?
t:1248 f: 0.312 pm:4.479 jm:386 q: -5.6 -6.0 91XJY0USP7M3UT* -
t:1344 f: 0.312 pm:4.518 jm:386 q: -5.2 -9.0 UGENBXXY3SQ2OSM ?
t:1440 f: 0.312 pm:4.554 jm:386 q: -4.9 -8.1 LCFT*NBAM LAKU3 ?
t:1536 f: 0.312 pm:4.585 jm:386 q: -4.9 -6.1 QJU9MQ.RQNNBX6R -
t:1632 f: 0.312 pm:4.679 jm:386 q: -4.7 -7.5 EFYE2WHVR/UU5FS ?

I.E., same signal, and it never decodes! The rapidly increasing "jm" 
values show the shifting of the framing, and the "q" values never get 
above the magic -3 or so that it takes to decode a signal.

John Andrews, W1TAG