OK, Thanks to all for replies on the noise generator calculations. I'm now
quite convinced its a simple SQRT(2*BW / Fclock) power division from the total
calculated from peak waveform voltage. The lack of any PI seems obvious now,
provided both clock and bandwidth are specified in the same units.
Now need to just make up a 32 bit shift register clocked at 4MHz for a PRN
sequence that generates spectral lines 0.93mHz apart (17.9 minutes repeat
interval) - enough for most LF modes used - so far.
BTW, I found a web page listing ALL the shift register taps for maximal length
sequences up to 32 stages. If anyone is interested it is
http://homepage.mac.com/afj/taplist.html
Andy G4JNT
-----Original Message-----
From: James Moritz [SMTP:[email protected]]
Sent: 2003/12/12 11:17
To: [email protected]
Subject: RE: LF: Related Technical Query - Soundcard calibrator
Dear Andy, LF Group
I hate to throw a spanner in the works, but... A noise source of exactly
this type is discussed in Horowitz & Hill "The Art of Electronics" 2nd
ed., section 9.34. This gives the formula (in the form Andy has used)
Vrms^2 = Vpk^2 * 2 * BW /Fclock, (giving 224mV rms) which is different
by a factor of sqrt(pi) from the value Andy gets. I spent some time last
night trying to figure out which one is correct, and came out with a
reasonably convincing case for Horowitz & Hill's formula - if anyone is
interested, I will put it into an e-mail compatible form
Cheers, Jim Moritz
73 de M0BMU
|
