At 02:24 PM 4/12/2011, Sheingold, Dan wrote:

Hi Barrie:

 This exchange has aroused my curiosity, but the link to the item in question is missing (MULTIPL.PDF, P_SUPPLY.PDF).

 However I did find a February, 1955, Electronics article about a less Gargantuan quarter-square multiplier design that used the QK-329 (and several K2-Ws). (I didn’t get to read it at the time because I had been drafted into the U.S. Army and was doing Basic Training in South Carolina.)

Philbrick had designed early quarter-square multipliers around “red tube” versions of the 12AU7, which appeared to have approximate square-law diode relationships. But 1%-2% was the best that could be done (on a particular day!)

Dan

----------------------------------------

Sent: Wednesday, April 13, 2011 10:55 PM

To: Sheingold, Dan; Gilbert, Barrie

Subject: RE: Fwd:Re: QK-329

Dan:

I attached the relevant articles about tube multipliers. 

Your remark about early Philbrick use of the 12AU7 for quarter square multipliers is of interest to the Philbrick archive. May I post it with you credited as the source?

[snip]

-Joe

----------------------------------------

Date: Thu, 14 Apr 2011 11:05:13 -0400

Thanks, Joe. That is one formidable computer! You’re welcome, as always, to quote me. My only caveat is that the memories that many of these comments are based on are filtered through the mists of time.

Interesting that Fifer discusses the relationship between grid voltage and plate current (plate voltage constant). My recollection was that the square-law relationship we used was between the plate voltage and plate current, at some “optimal” value of grid bias. Problem with most tube types was discontinuities at low plate voltage.

Dan