Dan Sheingold clarifies the origin of the "averaging filter ", which has occasionally been attributed to him, in the following email exchange.
Sent: Monday, March 01, 2010 7:33 AM
To: Sheingold, Dan
Subject: the "modified scheingold" filter
I just encountered a reference to some of your early work in a paper - a real, physical paper made of dead tree. The funny thing is that I can't locate it on the web (at least through Google).
I don't have a copy of the ADI "Nonlinear Circuits Handbook", I'm ashamed to say, I do have the BB equivalent, and of course your Transducer Interfacing Handbook - formative stuff, for sure.
Apparently you discuss averaging filters within, and specifically filters whose transfer functions approximate the ideal averager. My paper library at home as a copy of a paper by Voigtman and Winefordner, published in May '86. They tweak one of your poles, call it a "modified Sheingold" filter, and rate it as first equal in figure of merit (settling time versus noise
bandwidth) of all the 2nd and 3rd order functions evaluated. I'm writing a Filter Wizard piece on averaging filters, following a request from a reader, and I thought I'd better refresh myself.
So, you have a filter named after to you, and it's a best-in-class type!
But this paper's absence from the web has prevented you from receiving the fame this merits.
The full reference is
Low-pass filters for signal averaging
Edward Voigtman and James D Winefordner
(Chemistry Dept, U of Florida, Gainesville) Rev. Sci. Instrum. 57 (5), May 1986 0034-6748/86/050957-10$01.30 copyright 1986 American Institute of Physics
If you can't locate it, let me know and I'll scan it to a pdf for you. It's actually quite a useful paper, as are loads of the other old things I still have despite several culls over the last few years.
best regards - Kendall C-P
From: "Sheingold, Dan" <Dan.Sheingold@analog.com>
To: Kendall <email@example.com>
Date: Tue, 2 Mar 2010 13:45:27 -0500
Subject: RE: the "modified scheingold" filter
There are a couple of inaccuracies in the saga of the "Scheingold filter, not the least being the spelling of my family name (it's like the Sahara Desert, no "c"). More important is that this is a case of "blame the messenger": this version of a highly effective 3-pole averaging filter appeared in a couple of publications that I edited, but its real inventor was Dr. Peter D. Hansen, who was Chief Scientist (or some such) at Philbrick, and it was a modification of a filter idea introduced by the late MIT Prof. Henry M. ("Hank") Paynter.
The most accessible version of these filters is in the Philbrick Applications Manual (of which I was editor), a copy of which is available on the ADI web site:
The pages of especial interest are
47 (Figure 2.18)
76 (Figure 3.25)
The Hansen papers on which these filters are based can be found in the 1965 numbers of Philbrick's "The Lightning Empiricist" (of which I was editor). Joe Sousa has kindly posted them on the Philbrick Archive.
Here are the citations:
"New approaches to the design of active filters," THE LIGHTNING EMPIRICIST, Vol. 13, Nos. 1&2 (1965), pp. 3-16.
In the abstract, he says, "...we discuss the design of operational amplifier and analog computor circuits suitable for use as low-pass filters. We also discuss the commonly used mathematically designed filters, i.e., Butterworth, Chebyshev, and Bessel. In addition, we present two new types of theoretical filters, the Paynter and the Averaging filters. Design data necessary for realizing these theoretical filters with amplifier circuits is presented..."
"New approaches to the design of active low-pass filters," THE LIGHTNING EMPIRICIST, Vol. 13, Nos.3 3&4 (1965), pp. 2-12.
Many years ago I saw a reference to a filter with my name attached to it (I believe it was in RSI), and I wrote to either the author of the article, or the editor of the publication, explaining the problem. But it's almost impossible to rewrite history in a narrow bandwidth. So I will sail through future history with an unearned reputation as filter expert. Alas!
Yours in analog,