Page 1 of 1

The HP 8903A "Pontis" Floating Cross-Coupled Output

Posted: Tue Apr 11, 2017 3:03 pm
by mediatechnology
Around 1980 HP introduced the 8903A Audio Analyzer featuring an active "floating" cross-coupled output.

The Pontis output was the pre-cursor of the SSM2142 and DRV134.

Image
The HP8903 Floating "Pontis" Output

Complete output stage of the HP8903A with attenuator: https://www.proaudiodesignforum.com/ima ... _Tiled.jpg
HP8903A Service Manual: https://www.proaudiodesignforum.com/ima ... ervice.pdf (21 mb)

The "Pontis" output was published in the August 1980 HP Journal

(I think U2's inputs are mis-drawn....)

Image
G.D. Pontis, "Floating A Source Output," HP Journal, August 1980.

August 1980 HP Journal: https://www.ka-electronics.com/images/p ... df#page=12 (pdf 7.7 MB)

Related reading:

"Balanced Outputs" https://www.proaudiodesignforum.com/for ... ?f=6&t=635

Re: The HP 8903A "Pontis" Floating Cross-Coupled Output

Posted: Thu Apr 13, 2017 11:03 am
by JR.
I vaguely recall adding some small caps for stability to work in all conditions, but indeed a nice topology that eliminates the service calls when outputs are one legged, and level drops with simple differential outputs. While that topology does preserve level you still lose headroom when one leg is shorted.

JR

Re: The HP 8903A "Pontis" Floating Cross-Coupled Output

Posted: Thu Apr 13, 2017 11:29 am
by mediatechnology
I vaguely recall adding some small caps for stability to work in all conditions
Walt Jung provides the only cross-coupled floating output schematic I've seen published with added stability caps.
The THAT1646 is a different topology but it also has a HF rolloff (and reduction in Zocm) courtesy of an internal 10 pF capacitor.

The "big" HP8903A schematic (linked above) shows a 270 pF on the "-" leg.

Image
SSM2142 Cross-Coupled Output With Added Stability Caps

Re: The HP 8903A "Pontis" Floating Cross-Coupled Output

Posted: Thu Apr 13, 2017 12:43 pm
by JR.
Yup, when you have competing NF paths and interface with the outside world where load impedances and RF can be unpredictable, we have to cover all bases.

JR