"Stacked Amplifiers Lower Noise," Wurcer, Kitchin, EDN 1982.

A resource for technical documentation. Datasheets, application notes, instruction manuals, books and links to resources are found in the Document Library.
Post Reply
User avatar
mediatechnology
Posts: 5444
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

"Stacked Amplifiers Lower Noise," Wurcer, Kitchin, EDN 1982.

Post by mediatechnology »

"Stacked Amplifiers Lower Noise," Scott Wurcer, Charles Kitchin, EDN, October 13, 1982.
You can realize an improvement in the signal-to-noise ratio of an amplification system by stacking the system's amplifiers - connecting their inputs in parallel and their outputs in series. You can also modify figure 1's circuit to produce an active balanced transformer (figure 2). This circuit provides the same 3dB noise improvement as in figure 1, the signals add but the noise outputs don't correlate.
Note that figure 2 shows cross-coupled inputs...

Image
"Stacked Amplifiers Lower Noise," Scott Wurcer, Charles Kitchin, EDN, October 13 1982, Figure 2.

"Stacked Amplifiers Lower Noise," Scott Wurcer, Charles Kitchin, EDN, October 13, 1982.
Full pdf:
https://proaudiodesignforum.com/images/ ... 3_1982.pdf
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: "Stacked Amplifiers Lower Noise," Wurcer, Kitchin, EDN 1

Post by JR. »

Optimistic, hoping to sell 2x the ICs.

JR
Cancel the "cancel culture", do not support mob hatred.
User avatar
mediatechnology
Posts: 5444
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: "Stacked Amplifiers Lower Noise," Wurcer, Kitchin, EDN 1

Post by mediatechnology »

It's the same circuit fragment as Cohen's "Double-Balanced" output (figure 5, 1984) and a simple way to realize common mode rejection but keep the output balanced.
It also appears to have a distortion-cancelling mechanism since the individual op amp errors are similar but opposing.

I saw another similar article where the author asserts that the HF common mode rejection is improved by using a cross-coupled input followed by yet a third differential to single-ended stage.
I'll find that one to post.
And yes, it does sell twice as many parts. (Or three times as many.)

The noise argument was less influential but still valid.

Image
"Double Balanced Microphone Amplifier," Graeme Cohen, 1984, Figure 5.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: "Stacked Amplifiers Lower Noise," Wurcer, Kitchin, EDN 1

Post by JR. »

The noise argument ASSume a low Z source. Paralleling devices increases the input noise current term so depending on the source impedance the expected 3 dB improvement may not be realized.

I saw no discussion of NF in the brief, but mentioning an alternate super low noise part that was only 4nV/rt/Hz suggests this is archaic, and not all that quiet technology.

JR

{edit- paralleling low noise devices was well explored by Motchenbacher and Fitchen in their '70s low noise design book. [/edit]
Cancel the "cancel culture", do not support mob hatred.
User avatar
mediatechnology
Posts: 5444
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: "Stacked Amplifiers Lower Noise," Wurcer, Kitchin, EDN 1

Post by mediatechnology »

Sometimes these circuits are useful in ways not intended, or realized, by the author.
Wurcer mentions this article recently (this century) in a post, then shows two DACs piggybacked on top of one another to lower noise.
Since there was no web presence for the 1982 article it seemed worthwhile to post.

Douglas Self spends a lot of time on stacking in his book, "Small Signal Design."
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: "Stacked Amplifiers Lower Noise," Wurcer, Kitchin, EDN 1

Post by JR. »

I recently saw a discussion of stacking convertors on facebook of all places :lol: but it was to point out the fallacy of stacking convertors to add all their output bits together.

I kind of like the idea of using using two convertors on balanced opposite polarity feeds to buy a little more resolution, (ignoring the cost) but that then requires some digital post processing to get the full benefit.

The cheap micro platforms I use with built in A/D generally offer the option of of using the 4 internal convertors as 1x12b or 4x10b, so in principle we should be able to pick up as much as one full bit from stacking two convertors. So 6dB improvement and somewhat better than the analog stacking S/N benefit of 3 dB.

JR
Cancel the "cancel culture", do not support mob hatred.
Post Reply