Design Article
Comment
Guru of Grounding
It likely was me on the AES tape. I've been trying to get the audio industry to ...
kendallcp
Strange coincidence - I turned out some old cassette tapes (remember them?) ...
Designing precision analog front ends (Part 1 of 2)
Jason Seitz, Staff Applications Engineer, National Semiconductor Corp
8/16/2010 3:53 PM EDT
A common analog design challenge is to convert a small, noisy, or perhaps near-rail signal into an accurate digital representation. Any analog design engineer worth their salt would know to employ an analog front end (AFE) to achieve this task. However, what would be the best topology: non-inverting, difference amplifier, or other, to obtain accuracy and precision at minimal complexity and cost?
Before deciding on a topology, it is important to understand the purpose of an AFE and its principles of operation, such as amplification, filtering and setting common-mode voltage. Then a deeper dive can be taken into the advantages and disadvantages of each of the popular AFE topologies to determine the best fit for a specific application. Completing the AFE design with the proper ADC is critical to guarantee the correct digital signal.
This article looks at three basic AFE topologies--the non-inverting amplifier, the difference amplifier, and the instrumentation amplifier--and their key attributes and tradeoffs, along with sources of error and non-linearity. It is presented in two parts, as pdf files (no registration required):
Part 1: Introduction, the non-inverting amplifier, and the difference amplifier, click here.
Part 2: The instrumentation amplifier, IA input/output range, amplifier specifications, tools, and conclusion, will be posted on August 19, 2010.
About the Author
Jason Seitz is staff applications engineer for National Semiconductor’s Precision Systems group working on National’s precision, low-power and low-voltage operational amplifiers. Previously, Seitz worked as a product quality engineer in National’s Worldwide Quality Network. Seitz received his bachelor’s of science degree in electrical engineering from the University of California at Davis and his master’s of science degree in electrical engineering at Santa Clara University.|
|
|
|
|
Clyde_rel
8/18/2010 11:35 AM EDT
No link to the article?
Sign in to Reply
Guru of Grounding
8/18/2010 1:00 PM EDT
If there's a link on the page, I sure can't find it! Is this a puzzle? Do we get a prize?
Sign in to Reply
rpell2
8/18/2010 1:07 PM EDT
The link is hard to see. The word "here" is hyperlinked where it says "click here".
Sign in to Reply
Guru of Grounding
8/18/2010 1:19 PM EDT
Finally found the link "here". Anyway, I'd like to point out that, in 99% of the pro audio world, the difference amplifier (with its low common-mode input impedances) is nearly universally used for balanced audio inputs. The designers forget that the signal source impedances are in series with those carefully-trimmed input resistors and, more importantly, the signal source impedances are seldom matched to any precision. As a result, an advertised CMRR of 90 dB often becomes 50 dB when a real source is connected rather than a laboratory generator. A new test procedure was issued in 2000 by the IEC (see IEC 60268-3) which intentionally imbalances the source by 10 ohms, first in one leg of the input and then the other. The CMRR readings thus obtained more accurately predict real-world noise rejection. An IC that overcomes this problem is the 1200-series InGenius(R) input stage ICs by THAT Corp. (see www.thatcorp.com). It desensitizes to the effects of source impedances by using a patented bootstrap of the input bias resistors (required for op-amp operation). Transformers enjoy excellent CMRR largely because their common-mode input impedances approach infinity.
Sign in to Reply
kendallcp
8/22/2010 12:41 PM EDT
Strange coincidence - I turned out some old cassette tapes (remember them?) last week, and amongst them were some recordings from a 1997 AES (Audio Engineering Society) Convention. I listened to them in the car (yes, old enough that it still plays tapes) and I believe that one of the presenters was a certain Guru of Grounding (was it you?). Amazing how timeless the advice and analysis is in such AES technical sessions - just yesterday I was at a concert where a slight amount of PA buzz, probably from an unmanaged "pin 1 problem", was clearly audible when the musicians stopped playing.
Sign in to Reply
Guru of Grounding
8/25/2010 1:54 PM EDT
It likely was me on the AES tape. I've been trying to get the audio industry to understand the true nature of balanced interfaces since my first paper on the subject was published in the AES Journal back in 1995. I write and lecture (to anyone who'll listen!) at every trade show and university opportunity I get. My real name is Bill Whitlock and you can check my speaking schedule at www.jensen-transformers.com under "seminars". I'll be presenting an AES paper this fall about the true (and surprising to some) source of most system ground voltage differences.
Sign in to Reply