[hpsdr] Boat Anchors..PreAmps

[Phil Harman]

Just note that the AGC needs to be before the ADC. In the case of Mercury 
you have two settings 0db and -20dB.

We would need a software programmable variable attenuator to use this 
technique or fit an external one perhaps connected to the I2C bus?

73 Phil...VK6APH




----- Original Message ----- 
From: "Lester Veenstra" <lester at veenstras.com>
To: <lstoskopf at cox.net>
Cc: <hpsdr at lists.openhpsdr.org>; <phil at pharman.org>
Sent: Wednesday, April 13, 2011 3:59 AM
Subject: RE: [hpsdr] Boat Anchors..PreAmps


The value of PTIC was arrived at experimentally rather than empirically. So 
the mathematicians out there can settle back down, unless one decides to 
prove the result is impossible.

The simplicity of this method is that you do not need to actually calculate 
the percentage. Just count for a long enough time to actually see about 1E-7 
percent, based on the sample rate. If before that time interval, you are 
counting values, you are overloading and should drop the AGC gain a step, 
and restart the count. If by the end of the period that should give the PTIC 
target, you have counted noting, increase the AGC gain a notch and try 
again.

The relationship between the PTIC value and the NPR value is a steep curve. 
That is, the PTIC will change a decade for a small (say 1/2 dB) change in 
the NPR, so a crude estimate of the PTIC is sufficient to get you in the 
right loading region. It is not necessary to measure until you have a 
statistically significant number of samples. In practice, the target is no 
counts in the expected period, step a half db and count some more. If you 
get counts, drop a half dB.  The period necessary to get an expected count 
will be long enough to ensure you have a long slow and easy AGC function. 
Yes you will bang up and down in succession to keep you in a known area, but 
the output will not have perceptible artifacts.

And to respond to an earlier question in this thread, optimum is optimum. 
What is optimum based on the total input to the A/D, all the power in all 
the nyquist or less bandwidth, is optimum for any small signal in the 
bandwidth. In other words, with a wide band A/D, you cannot get a better 
loading that will improve the S/N of a narrow band demod process after the 
A/D.



Lester B Veenstra  MØYCM K1YCM
lester at veenstras.com
m0ycm at veenstras.com
k1ycm at veenstras.com


US Postal Address:
PSC 45 Box 781
APO AE 09468 USA

Telephones:
Office:     +44-(0)1423-846-385
Home:     +44-(0)1943-880-963
Guam Cell: +1-671-788-5654
UK Cell:   +44-(0)7716-298-224
Jamaica:  +1-876-352-7504

This e-mail and any documents attached hereto contain confidential or
privileged information. The information is intended to be for use only by
the individual or entity to whom they are addressed. If you are not the
intended recipient or the person responsible for delivering the e-mail to
the intended recipient, be aware that any disclosure, copying, distribution
or use of the contents of this e-mail or any documents attached hereto is
prohibited.


-----Original Message-----
From: lstoskopf at cox.net [mailto:lstoskopf at cox.net]
Sent: Tuesday, April 12, 2011 8:39 PM
To: 'Tom Holmes'; phil at pharman.org; lester at veenstras.com
Cc: hpsdr-request at lists.openhpsdr.org; hpsdr at lists.openhpsdr.org
Subject: RE: [hpsdr] Boat Anchors..PreAmps

..........  Do you have a reference for those of us with 50 year old college 
math?

So we sample 10E6 samples...no 255, raise the gain, I'm guessing if you get 
255 before you are at .37E6 counts..drop the gain.  And in there somewhere 
must be a calculation of how much to raise. .......


N0UU

---- Lester Veenstra <lester at veenstras.com> wrote:
> As a point of interest, the optimum loading of an A/D is about 1E-6 
> percent
> time in clip (PTIC). The PTIC is percent of time the input voltage peak
> bangs on the rails, that is at minimum or maximum values. For an 8 bit
> example, the for some time interval, the sum of the number of times you 
> see
> a value of 0 or 255, divided by the total number of samples in the period
> counted.



-----
No virus found in this message.
Checked by AVG - www.avg.com
Version: 10.0.1321 / Virus Database: 1500/3565 - Release Date: 04/11/11



 1302654791.0