The Cyber-Spy.Com Usenet Archive Feeds Directly
From The Open And Publicly Available Newsgroup
This Group And Thousands Of Others Are Available
On Most IS NNTP News Servers On Port 119.
Cyber-Spy.Com Is NOT Responsible For Any Topic,
Opinions Or Content Posted To This Or Any Other
Newsgroup. This Web Archive Of The Newsgroup And
Posts Are For Informational Purposes Only.
From: "Bill Sloman"
Subject: Re: Niosy A/D converter - an advantage ?
Date: Fri, 15 Nov 2002 00:07:54 +0100
Organization: Planet Internet
NNTP-Posting-Date: 14 Nov 2002 23:07:51 GMT
X-Newsreader: Microsoft Outlook Express 6.00.2800.1106
"Roger Lascelles" wrote in message
> We have 12 bit ADC in a product. Due to grounding problems, the ADC
> readings fluctuate - I don't know exactly by how much, but lets say at
> the last two bits dither around. The noise is digital hash from the
> microprocessor which has external RAM and ROM.
> Because the AD for is only used to measure periodic or DC quantities, the
> programmer has added a lot of averaging code which produces smooth
> to an equivalent of 13 bits resolution. He claims that the noise is an
> advantage, because it dithers the AD to permit averaging.
> Has anybody got experience of ADC dithering / noisy ADCs ? What are the
> dangers or limitiations with this sort of thing ? Are there gotchas with
> some types of noise ? Do you know of any articles or web links which
> help ?
"Dithered Quantizers" by Robert M.Gray and Thomas G.Stockman, published in
the IEEE Transactions on Information Theory, volume 39 pages 805-812 (1993)
is a useful paper with 18 references to earlier work.
John Watkinson's "The Art of Digital Audio" ISBN 0-240-51270-7 contains
useful discussion of dither which lists several of these references.
James Potzick didn't manage to find any of these references in his short
paper on the subject in Review Scientific Instruments 40 pages 2038-40
(1999) and I sent them a rude comment, which was published ...
The good papers tend to make the point that you have to be careful to get
more or less the right probability distribution in your dithering signal to
get the best result. The Gray and Stockman paper has a great example of how
things can go wrong, drawn from professional audio recording.
Digital hash from from a nearby microprocessor probably isn't ideal.
Bill Sloman, Nijmegen
Go Back To The Cyber-Spy.Com
Usenet Web Archive Index Of
The sci.electronics.design Newsgroup