[hpsdr] QSD gain and noise.

[Johan Maas]

Hi,

A quick thought.... when we are setting the lsb of the divisor always to 0 
will that solve the problem?

By the way iam not using the msb of the accu but yousing the msb of the 
output of the cordic algorithm (see svn trunk of Phil Harman for the 
cordic.v)

73 Johan PA3GSB


----- Original Message ----- 
From: "Philip Covington" <p.covington at gmail.com>
To: "Robert McGwier" <rwmcgwier at gmail.com>
Cc: "Phil Harman" <phil at pharman.org>; "Johan Maas" <johan.maas at hetnet.nl>; 
<hpsdr at hpsdr.org>
Sent: Tuesday, September 25, 2007 5:13 AM
Subject: Re: [hpsdr] QSD gain and noise.


> On 9/24/07, Robert McGwier <rwmcgwier at gmail.com> wrote:
>> I can't believe how many times this mistake is made.  I made it myself
>> about four times.  I keep "rediscovering" this guaranteed source of
>> jitter and phase noise.  The frequency will have to be an even divisor
>> of the clock on you will have more spurs at sin(x)/x amplitudes than you
>> can even believe.
>>
>> Bob
>>
> Yes, an interesting exercise is to connect the MSB of your accumulator
> to one of the output pins on the FPGA - then look at what you get on a
> spectrum analyzer when the divisor is not even - not very pretty :-)
>
> 73 Phil N8VB 


 1190783505.0