The Popular Electronics article is December 1961. I think it is as
you are suggesting, two cascaded pots, both wired as voltage
dividers, so you get two ratios and hence mulitplication: Vout = Vin*
(R1p/R1t)*(R2p/R2t) where p & t are the position and total Rs, Vout
is then null balanced against the third pot. The 2nd pot R is much
larger than the 1st to minimise loading error on the 1st.
Apparently some versions used a meter and DC, and some used an audio
oscillator and AC, with headphones replacing the null meter, which
explains the caps and transistors I saw.
No logs and even simpler than I thought.
On 2013 Mar 10, at 9:58 PM, dwight elvey wrote:
It seems like I recall two types. There was the one
withthe linear
scale ( radio shack ) and another with log scales.Dwight
From: hilpert at cs.ubc.ca
>
> Well, I assumed it was doing a simple V R ratio / sum-of-logs of two
> of the controls with a null-point balance with the third, however one
> of the sites on the web showed a bag of caps and transistors involved
> in the construction so perhaps it's using some AC principle, or that
> may be a mislead.
>
> I'm going to try to find the PE article in the library at the radio
> museum this afternoon.
>
>
> On 2013 Mar 10, at 11:50 AM, dwight elvey wrote:
>
>> I did some searching on the web but I was surprisedthat no one has
>> posted a schematic for it.Dwight
>>
>>> Actually, after giving it a little thought, it would use
>>> linearpots and scales. It would be voltage dividers and a
>>> balancingmeter.Remeber, a pot forms a ratio across a
>>> resistor.Dwight
>>>
From: hilpert at cs.ubc.ca
>>>
>>>>
>>>> That's what I was expecting, but looking at the dial (as much
>>>> as one
>>>> can see in the photos), they appear to be linear, so I figured it
>>>> was
>>>> taking advantage of log pots (or is there a log/anti-log problem
>>>> there anyways?). Considering it more, a typical log pot may not
>>>> give
>>>> an accurate enough log function to be very good for the task, so
>>>> you're probably right. Maybe there's another principle involved.
>>>>
>>>> Another web site refers to a popular electronics article with a
>>>> similar design (I think I've seen the article in the past) but
>>>> doesn't say which issue.
>>>>
>>>>
>>>> On 2013 Mar 8, at 8:58 PM, dwight elvey wrote:
>>>>> Actually it makes more sense to use alinear one with log
>>>>> scales on
>>>>> the dial.Dwight
>>>>>
From: hilpert at cs.ubc.ca
>>>>>> ---snip---
>>>>>>
>>>>>> I like how the box says "As accurate as a slide-rule".
>>>>>>
>>>>>> Presumably it uses standard log pots, aka volume controls.