Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > sci.electronics.design > #741707 > unrolled thread

one potentiometer dividing multiple signals

Back to article view | Back to sci.electronics.design

Contents

  one potentiometer dividing multiple signals Christopher Howard <christopher@librehacker.com> - 2026-03-16 07:21 -0800
    Re: one potentiometer dividing multiple signals Martin Brown <'''newspam'''@nonad.co.uk> - 2026-03-16 15:45 +0000
    Re: one potentiometer dividing multiple signals "Edward Rawde" <invalid@invalid.invalid> - 2026-03-16 11:50 -0400
    Re: one potentiometer dividing multiple signals john larkin <jl@glen--canyon.com> - 2026-03-16 09:42 -0700
    Re: one potentiometer dividing multiple signals Don Y <blockedofcourse@foo.invalid> - 2026-03-16 09:54 -0700
    Re: one potentiometer dividing multiple signals liz@poppyrecords.invalid.invalid (Liz Tuddenham) - 2026-03-16 20:03 +0000
      Re: one potentiometer dividing multiple signals liz@poppyrecords.invalid.invalid (Liz Tuddenham) - 2026-03-19 10:01 +0000
    Re: one potentiometer dividing multiple signals antispam@fricas.org (Waldek Hebisch) - 2026-03-17 21:28 +0000
      Re: one potentiometer dividing multiple signals Christopher Howard <christopher@librehacker.com> - 2026-03-18 08:25 -0800
        Re: one potentiometer dividing multiple signals John R Walliker <jrwalliker@gmail.com> - 2026-03-18 16:34 +0000
        Re: one potentiometer dividing multiple signals Martin Brown <'''newspam'''@nonad.co.uk> - 2026-03-19 12:42 +0000
          Re: one potentiometer dividing multiple signals Christopher Howard <christopher@librehacker.com> - 2026-03-19 11:51 -0800
            Re: one potentiometer dividing multiple signals john larkin <jl@glen--canyon.com> - 2026-03-19 13:16 -0700
            Re: one potentiometer dividing multiple signals liz@poppyrecords.invalid.invalid (Liz Tuddenham) - 2026-03-19 22:18 +0000
            Re: one potentiometer dividing multiple signals Jasen Betts <usenet@revmaps.no-ip.org> - 2026-03-20 07:29 +0000
            Re: one potentiometer dividing multiple signals liz@poppyrecords.invalid.invalid (Liz Tuddenham) - 2026-03-20 09:05 +0000
        Re: one potentiometer dividing multiple signals Jasen Betts <usenet@revmaps.no-ip.org> - 2026-03-20 07:25 +0000
          Re: one potentiometer dividing multiple signals liz@poppyrecords.invalid.invalid (Liz Tuddenham) - 2026-03-20 09:05 +0000
    Re: one potentiometer dividing multiple signals someone <cffbf4deb9142bce48974efc0e64dede@example.com> - 2026-03-17 22:30 +0000
    Re: one potentiometer dividing multiple signals liz@poppyrecords.invalid.invalid (Liz Tuddenham) - 2026-03-19 18:44 +0000
    Re: one potentiometer dividing multiple signals Jasen Betts <usenet@revmaps.no-ip.org> - 2026-03-20 07:12 +0000
    Re: one potentiometer dividing multiple signals john larkin <jl@glen--canyon.com> - 2026-03-20 09:08 -0700
      Re: one potentiometer dividing multiple signals Ross Finlayson <ross.a.finlayson@gmail.com> - 2026-03-20 11:52 -0700
      Re: one potentiometer dividing multiple signals Christopher Howard <christopher@librehacker.com> - 2026-03-23 09:09 -0800
        Re: one potentiometer dividing multiple signals piglet <erichpwagner@hotmail.com> - 2026-03-24 08:48 +0000

Page 1 of 2  [1] 2  Next page →

#741707 — one potentiometer dividing multiple signals

In my analog computer setup, a very common thing is to use
potentiometers as coefficients, i.e., as voltage dividers that attenuate
an input signal. But something wasteful about this is that sometimes
multiple different signals have to be divided by the same coefficient —
for example, dividing various forces by the mass of the object. So, I
have to use multiple potentiometers for that, and I have to keep them
set to the same value.

So, I was wondering what would be the simplest analog approach to get
around this problem, without having to use expensive
multiplier/dividers. Maybe just have one potentiometer connected to a
bunch of cheap transistors? 

-- 
Christopher Howard

[toc] | [next] | [standalone]

#741711

On 16/03/2026 15:21, Christopher Howard wrote:
> In my analog computer setup, a very common thing is to use
> potentiometers as coefficients, i.e., as voltage dividers that attenuate
> an input signal. But something wasteful about this is that sometimes
> multiple different signals have to be divided by the same coefficient —
> for example, dividing various forces by the mass of the object. So, I
> have to use multiple potentiometers for that, and I have to keep them
> set to the same value.
> 
> So, I was wondering what would be the simplest analog approach to get
> around this problem, without having to use expensive
> multiplier/dividers. Maybe just have one potentiometer connected to a
> bunch of cheap transistors?

If the signals are being combined later then you can sum them with a 
unity gain buffer and then divide by one potentiometer.

The other easy way out is use multiple linear potentiometers (like on a 
mixing desk) and a piece of wood to move them all at the same time.

-- 
Martin Brown

[toc] | [prev] | [next] | [standalone]

#741713

"Christopher Howard" <christopher@librehacker.com> wrote in message news:87ldfr6bqo.fsf@librehacker.com...
> In my analog computer setup, a very common thing is to use
> potentiometers as coefficients, i.e., as voltage dividers that attenuate
> an input signal. But something wasteful about this is that sometimes
> multiple different signals have to be divided by the same coefficient -
> for example, dividing various forces by the mass of the object. So, I
> have to use multiple potentiometers for that, and I have to keep them
> set to the same value.
>
> So, I was wondering what would be the simplest analog approach to get
> around this problem, without having to use expensive
> multiplier/dividers. Maybe just have one potentiometer connected to a
> bunch of cheap transistors?

Sounds like you need one potentiometer connected to a unity gain buffer.
Find a spare op amp and connect the wiper to the + input.
The voltage at the wiper should stay between the op amp supply rails.
Connect the - input to the output.
The output should now follow the wiper voltage.
The difference is that the output provides a low impedance drive which
can now be connected to many things without changing the output voltage.
Sometimes you might want more than one buffer connected to the same wiper.

>
> -- 
> Christopher Howard 

[toc] | [prev] | [next] | [standalone]

#741716

On Mon, 16 Mar 2026 07:21:51 -0800, Christopher Howard
<christopher@librehacker.com> wrote:

>In my analog computer setup, a very common thing is to use
>potentiometers as coefficients, i.e., as voltage dividers that attenuate
>an input signal. But something wasteful about this is that sometimes
>multiple different signals have to be divided by the same coefficient —
>for example, dividing various forces by the mass of the object. So, I
>have to use multiple potentiometers for that, and I have to keep them
>set to the same value.
>
>So, I was wondering what would be the simplest analog approach to get
>around this problem, without having to use expensive
>multiplier/dividers. Maybe just have one potentiometer connected to a
>bunch of cheap transistors? 

How about DPOTS? You'd have to program them somehow, but all with the
same data.

Cheaper than MDACs.


John Larkin
Highland Tech Glen Canyon Design Center
Lunatic Fringe Electronics

[toc] | [prev] | [next] | [standalone]

#741717

On 3/16/2026 8:21 AM, Christopher Howard wrote:
> In my analog computer setup, a very common thing is to use
> potentiometers as coefficients, i.e., as voltage dividers that attenuate
> an input signal. But something wasteful about this is that sometimes
> multiple different signals have to be divided by the same coefficient —
> for example, dividing various forces by the mass of the object. So, I
> have to use multiple potentiometers for that, and I have to keep them
> set to the same value.

You are using the pots as resistive dividers (?)  The ratio of the two
"sides" of the wiper effectively determines the divisor.  I.e., if
60% of the total resistance is on one side of the tap, then you are
dividing by 100%/60% (or, 100%/40%, depending on your point of
reference).

[There are other ways of configuring the pot, too]

To have multiple pots act in unison, you need a multiple-gang pot:
<https://nolelc.com/wp-content/uploads/2026/01/multi-gang-rotary-potentiometer.webp>
Here, both pots are operated by the same rotary shaft.

You can typically see such devices in use in a legacy "stereo"
where one pot controls the volume of the left channel while
the other (on the same shaft) controls the volume of the
right channel.

The problem with this approach is that the two don't always track
perfectly.  E.g., one may yield 60% while the other reflects 59%.
Also, the tapers may have slight differences.  So, the "60" may
become "61" as the pot is turned while the "59" may become "59.5"
despite the shaft having turned the same amount in each.

It's also more expensive and there are limits to just how many
pots you can stack on a shaft.

> So, I was wondering what would be the simplest analog approach to get
> around this problem, without having to use expensive
> multiplier/dividers. Maybe just have one potentiometer connected to a
> bunch of cheap transistors?

In that approach, you are trying to replicate a *voltage* for distribution
to multiple consumers.  You are only "dividing" the ONE signal that is
presented to the pot ("divider"); your further attempts ("transistors")
are just *replicating* it (and buffering each).

Look at the "math" you are attempting to perform.  See if you can
exploit associative and commutative properties to rearrange the
"equation" to reduce the number of "divisions":

x/K + y/K would require two divisions by K.  But, (x + y)/K
requires only one.

[toc] | [prev] | [next] | [standalone]

#741726

Christopher Howard <christopher@librehacker.com> wrote:

> In my analog computer setup, a very common thing is to use
> potentiometers as coefficients, i.e., as voltage dividers that attenuate
> an input signal. But something wasteful about this is that sometimes
> multiple different signals have to be divided by the same coefficient —
> for example, dividing various forces by the mass of the object. So, I
> have to use multiple potentiometers for that, and I have to keep them
> set to the same value.
> 
> So, I was wondering what would be the simplest analog approach to get
> around this problem, without having to use expensive
> multiplier/dividers. Maybe just have one potentiometer connected to a
> bunch of cheap transistors? 

I had this problem with a multi-stage audio filter and solved it with
variable on/off ratio CMOS switches.  

www.poppyrecords.co.uk/other/500CylinderAudioAmplifier.gif

One switch (Ic536) is the 'master' and is in one arm of a bridge
circuit, with the 'pot' (actually switched resistors in this case)
(R533) in another arm.  The off-balance voltage of the bridge is
amplified and alters the switching ratio until the bridge comes into
balance.  In the balanced condition, the effective resistance of R534 in
series with the chopping switch, Ic536 is the same as the selected value
of R533.

The resistors handling the signals are all switched by CMOS switches
(Ic528 - Ic626) in synchronism with the 'master', so they track it 


-- 
~ Liz Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

[toc] | [prev] | [next] | [standalone]

#741886

Liz Tuddenham <liz@poppyrecords.invalid.invalid> wrote:

Apologies for following-up my own post but I discovered the link below
wasn't working - I have now put it right.


> www.poppyrecords.co.uk/other/500CylinderAudioAmplifier.gif
> 
> One switch (Ic536) is the 'master' and is in one arm of a bridge
> circuit, with the 'pot' (actually switched resistors in this case)
> (R533) in another arm.  The off-balance voltage of the bridge is
> amplified and alters the switching ratio until the bridge comes into
> balance.  In the balanced condition, the effective resistance of R534 in
> series with the chopping switch, Ic536 is the same as the selected value
> of R533.
> 
> The resistors handling the signals are all switched by CMOS switches
> (Ic528 - Ic626) in synchronism with the 'master', so they track it 


-- 
~ Liz Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

[toc] | [prev] | [next] | [standalone]

#741790

Christopher Howard <christopher@librehacker.com> wrote:
> In my analog computer setup, a very common thing is to use
> potentiometers as coefficients, i.e., as voltage dividers that attenuate
> an input signal. But something wasteful about this is that sometimes
> multiple different signals have to be divided by the same coefficient —
> for example, dividing various forces by the mass of the object. So, I
> have to use multiple potentiometers for that, and I have to keep them
> set to the same value.
> 
> So, I was wondering what would be the simplest analog approach to get
> around this problem, without having to use expensive
> multiplier/dividers. Maybe just have one potentiometer connected to a
> bunch of cheap transistors?

Two potential ideas:
- build a bunch of dividers based on termistors.  A single potentimeter
  can provide control signal to a heater heating the whole assembly.
- similar idea but using fotoresistors and a light bulb.

Or you can go half way to digital and use pulse with modulation.
Namely, one potemeter can control duty cycle of a generator.
Such generator can control several analog mutiplexers.  Filtering
signal after the multiplexer you get input signal time duty cycle
of the generator.

-- 
                              Waldek Hebisch

[toc] | [prev] | [next] | [standalone]

#741848

antispam@fricas.org (Waldek Hebisch) writes:

> Two potential ideas:
> - build a bunch of dividers based on termistors.  A single potentimeter
>   can provide control signal to a heater heating the whole assembly.
> - similar idea but using fotoresistors and a light bulb.
>

I'm nervous about burning the bench down with the thermistor approach.
But the photoresistor idea sounds appealing, as being a rather simple
approach, that could be done with a cool LED. What approach would you
suggest as far as having output voltage correspond to change in
photoresistor resistance? Like, have the photoresistor pull input signal
to ground? Or have an op amp with negative feedback resistor on the
output side?

> Or you can go half way to digital and use pulse with modulation.
> Namely, one potemeter can control duty cycle of a generator.
> Such generator can control several analog mutiplexers.  Filtering
> signal after the multiplexer you get input signal time duty cycle
> of the generator.

I few folks have mentioned variants of this idea. It sounds a little
more complicated than I would prefer, but I'm keeping it in the back of
my mind.

-- 
Christopher Howard

[toc] | [prev] | [next] | [standalone]

#741849

On 18/03/2026 16:25, Christopher Howard wrote:
> antispam@fricas.org (Waldek Hebisch) writes:
> 
>> Two potential ideas:
>> - build a bunch of dividers based on termistors.  A single potentimeter
>>    can provide control signal to a heater heating the whole assembly.
>> - similar idea but using fotoresistors and a light bulb.
>>
> 
> I'm nervous about burning the bench down with the thermistor approach.
> But the photoresistor idea sounds appealing, as being a rather simple
> approach, that could be done with a cool LED. What approach would you
> suggest as far as having output voltage correspond to change in
> photoresistor resistance? Like, have the photoresistor pull input signal
> to ground? Or have an op amp with negative feedback resistor on the
> output side?
> 
>> Or you can go half way to digital and use pulse with modulation.
>> Namely, one potemeter can control duty cycle of a generator.
>> Such generator can control several analog mutiplexers.  Filtering
>> signal after the multiplexer you get input signal time duty cycle
>> of the generator.
> 
> I few folks have mentioned variants of this idea. It sounds a little
> more complicated than I would prefer, but I'm keeping it in the back of
> my mind.
> 
It will probably give you better linearity and repeatability
across channels than any of the other approaches.

However, if you want to keep things simple, you could connect
the leds of a string of series connected H11F1 FET optocouplers
to a current source controlled by the potentiometer and use the
light sensitive FETs themselves in your various circuits.
John

[toc] | [prev] | [next] | [standalone]

#741898

On 18/03/2026 16:25, Christopher Howard wrote:
> antispam@fricas.org (Waldek Hebisch) writes:
> 
>> Two potential ideas:
>> - build a bunch of dividers based on termistors.  A single potentimeter
>>    can provide control signal to a heater heating the whole assembly.
>> - similar idea but using fotoresistors and a light bulb.
>>
> 
> I'm nervous about burning the bench down with the thermistor approach.
> But the photoresistor idea sounds appealing, as being a rather simple
> approach, that could be done with a cool LED. What approach would you
> suggest as far as having output voltage correspond to change in
> photoresistor resistance? Like, have the photoresistor pull input signal
> to ground? Or have an op amp with negative feedback resistor on the
> output side?

A summing amplifier is probably what you want to use and combine all the 
terms with the same coefficient using that *before* the potentiometer.

If you provide a netlist or specimen differential equation then we could 
sketch out how to do this in ASCII art or another netlist.
> 
>> Or you can go half way to digital and use pulse with modulation.
>> Namely, one potemeter can control duty cycle of a generator.
>> Such generator can control several analog mutiplexers.  Filtering
>> signal after the multiplexer you get input signal time duty cycle
>> of the generator.
> 
> I few folks have mentioned variants of this idea. It sounds a little
> more complicated than I would prefer, but I'm keeping it in the back of
> my mind.

Thermistors and LDRs tend to be somewhat non-linear and getting some 
that will track well is nigh on impossible. I have used them once in a 
theramin where isolating the RF side from the audio was critical.

Dual gang pots are fairly common for stereo audio anything beyond that 
is specialist stuff.


-- 
Martin Brown

[toc] | [prev] | [next] | [standalone]

#741923

> If you provide a netlist or specimen differential equation then we
> could sketch out how to do this in ASCII art or another netlist.

I was working on a simplified flight model using angle of attack to
calculate lift and drag, and thus vertical and horizontal acceleration.

* vertical acceleration is lift force minus gravity force, divided by
  mass
* gravity force is mass times a gravity variable
* horizontal acceleration is thrust minus drag force, divided by mass
* the horizontal acceleration is integrated to give horizontal velocity
  which is squared and then fed into the circuit that calculates lift
  and drag force based on angle of attack (coefficients of lift/drag).

Ignoring for the present the detail of different air behavior at
different altitudes, and banking.

With analog computing, multiplication or division by a constant can be
done with a potentiometer, though it is physically speaking a voltage division.

It is not obvious to me how combine those three mass related
calculations into something done with a single summing amplifier and a
single pot, but I am open minded.

-- 
Christopher Howard

[toc] | [prev] | [next] | [standalone]

#741924

On Thu, 19 Mar 2026 11:51:16 -0800, Christopher Howard
<christopher@librehacker.com> wrote:

>> If you provide a netlist or specimen differential equation then we
>> could sketch out how to do this in ASCII art or another netlist.
>
>I was working on a simplified flight model using angle of attack to
>calculate lift and drag, and thus vertical and horizontal acceleration.
>
>* vertical acceleration is lift force minus gravity force, divided by
>  mass
>* gravity force is mass times a gravity variable
>* horizontal acceleration is thrust minus drag force, divided by mass
>* the horizontal acceleration is integrated to give horizontal velocity
>  which is squared and then fed into the circuit that calculates lift
>  and drag force based on angle of attack (coefficients of lift/drag).
>
>Ignoring for the present the detail of different air behavior at
>different altitudes, and banking.
>
>With analog computing, multiplication or division by a constant can be
>done with a potentiometer, though it is physically speaking a voltage division.
>
>It is not obvious to me how combine those three mass related
>calculations into something done with a single summing amplifier and a
>single pot, but I am open minded.

Why not Spice it?


John Larkin
Highland Tech Glen Canyon Design Center
Lunatic Fringe Electronics

[toc] | [prev] | [next] | [standalone]

#741929

Christopher Howard <christopher@librehacker.com> wrote:

> > If you provide a netlist or specimen differential equation then we
> > could sketch out how to do this in ASCII art or another netlist.
> 
> I was working on a simplified flight model using angle of attack to
> calculate lift and drag, and thus vertical and horizontal acceleration.
> 
> [1]* vertical acceleration is lift force minus gravity force, divided by
>   mass
> [2]* gravity force is mass times a gravity variable
> [3]* horizontal acceleration is thrust minus drag force, divided by mass
> * the horizontal acceleration is integrated to give horizontal velocity
>   which is squared and then fed into the circuit that calculates lift
>   and drag force based on angle of attack (coefficients of lift/drag).
> 
> Ignoring for the present the detail of different air behavior at
> different altitudes, and banking.
> 
> With analog computing, multiplication or division by a constant can be
> done with a potentiometer, though it is physically speaking a voltage
> division.
> 
> It is not obvious to me how combine those three mass related
> calculations into something done with a single summing amplifier and a
> single pot, but I am open minded.

In the first and third equations you are dividing by mass but in the
second equation you are multiplying by mass, so even ganged
potentiometers won't directly do what you want.  If you make one of them
divide the feedback around an op-amp, this will give multiplication
instead of division.



-- 
~ Liz Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

[toc] | [prev] | [next] | [standalone]

#741939

On 2026-03-19, Christopher Howard <christopher@librehacker.com> wrote:
>> If you provide a netlist or specimen differential equation then we
>> could sketch out how to do this in ASCII art or another netlist.
>
> I was working on a simplified flight model using angle of attack to
> calculate lift and drag, and thus vertical and horizontal acceleration.
>
> * vertical acceleration is lift force minus gravity force, divided by
>   mass
> * gravity force is mass times a gravity variable
> * horizontal acceleration is thrust minus drag force, divided by mass
> * the horizontal acceleration is integrated to give horizontal velocity
>   which is squared and then fed into the circuit that calculates lift
>   and drag force based on angle of attack (coefficients of lift/drag).
>
> Ignoring for the present the detail of different air behavior at
> different altitudes, and banking.
>
> With analog computing, multiplication or division by a constant can be
> done with a potentiometer, though it is physically speaking a voltage division.
>
> It is not obvious to me how combine those three mass related
> calculations into something done with a single summing amplifier and a
> single pot, but I am open minded.

perhaps time division, or frequency division multiplexing ... it seems 
kind of tricky to me too.

-- 
 Jasen.
 🇺🇦 Слава Україні

[toc] | [prev] | [next] | [standalone]

#741946

Christopher Howard <christopher@librehacker.com> wrote:

> [1]* vertical acceleration is lift force minus gravity force, divided by
>   mass
> [2]* gravity force is mass times a gravity variable
> [3]* horizontal acceleration is thrust minus drag force, divided by mass
> * the horizontal acceleration is integrated to give horizontal velocity
>   which is squared and then fed into the circuit that calculates lift
>   and drag force based on angle of attack (coefficients of lift/drag).

[1]   Va = (L-G)/M
[2]   G = M*k
[3]   Ha = (T-D)/M


Equation [1] can be expressed as:

[4]   Va = (L-(M*k))/M

...but this does not help.  The simplest way would be to have two
separate inputs, one for M and the other for G (regardless of the fact
that G is related to M).  Then the M potentiometer only had to cope with
two inputs and both of them are divisions, which are easy to do with a
pot.

Then you could either use off-the-shelf twin pots (if the accuracy isn't
critical) or a chopper or time-division system running at much higher
frequency than the input signals.


-- 
~ Liz Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

[toc] | [prev] | [next] | [standalone]

#741938

On 2026-03-18, Christopher Howard <christopher@librehacker.com> wrote:
> antispam@fricas.org (Waldek Hebisch) writes:
>
>> Two potential ideas:
>> - build a bunch of dividers based on termistors.  A single potentimeter
>>   can provide control signal to a heater heating the whole assembly.
>> - similar idea but using fotoresistors and a light bulb.
>>
>
> I'm nervous about burning the bench down with the thermistor approach.
> But the photoresistor idea sounds appealing, as being a rather simple
> approach, that could be done with a cool LED. What approach would you
> suggest as far as having output voltage correspond to change in
> photoresistor resistance?

use n+1 all the LEDs in series and the zeroth photoconductor in the 
feedback path of the amplifier that drives the LEDs

> Like, have the photoresistor pull input signal
> to ground? Or have an op amp with negative feedback resistor on the
> output side?

use them as one arm of a voltage divider - however works best.
getting all to track equally may be tricky

-- 
 Jasen.
 🇺🇦 Слава Україні

[toc] | [prev] | [next] | [standalone]

#741945

Jasen Betts <usenet@revmaps.no-ip.org> wrote:

> On 2026-03-18, Christopher Howard <christopher@librehacker.com> wrote:
> > antispam@fricas.org (Waldek Hebisch) writes:
> >
> >> Two potential ideas:
> >> - build a bunch of dividers based on termistors.  A single potentimeter
> >>   can provide control signal to a heater heating the whole assembly.
> >> - similar idea but using fotoresistors and a light bulb.
> >>
> >
> > I'm nervous about burning the bench down with the thermistor approach.
> > But the photoresistor idea sounds appealing, as being a rather simple
> > approach, that could be done with a cool LED. What approach would you
> > suggest as far as having output voltage correspond to change in
> > photoresistor resistance?
> 
> use n+1 all the LEDs in series and the zeroth photoconductor in the 
> feedback path of the amplifier that drives the LEDs
> 
> > Like, have the photoresistor pull input signal
> > to ground? Or have an op amp with negative feedback resistor on the
> > output side?
> 
> use them as one arm of a voltage divider - however works best.
> getting all to track equally may be tricky

The problem with analogue computing is the accuracy - especially as we
now learn (in another reply) that one of the resultants has to be
squared. and another is the result of subtracting two quantities.


-- 
~ Liz Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

[toc] | [prev] | [next] | [standalone]

#741796

If it's analog computation then those inputs are independent and are required to be  electronically isolated from one another.

There are electronically controlled pots that come in much cheaper than a multiturn mechanical, but there's a lot of design overhead involved.

-- 
For full context, visit https://www.electrondepot.com/electrodesign/one-potentiometer-dividing-multiple-signals-4399004-.htm

[toc] | [prev] | [next] | [standalone]

#741921

Christopher Howard <christopher@librehacker.com> wrote:

> In my analog computer setup, a very common thing is to use
> potentiometers as coefficients, i.e., as voltage dividers that attenuate
> an input signal. But something wasteful about this is that sometimes
> multiple different signals have to be divided by the same coefficient —
> for example, dividing various forces by the mass of the object. So, I
> have to use multiple potentiometers for that, and I have to keep them
> set to the same value.
> 
> So, I was wondering what would be the simplest analog approach to get
> around this problem, without having to use expensive
> multiplier/dividers. Maybe just have one potentiometer connected to a
> bunch of cheap transistors? 

One approach that nobody has mentioend is time-division multiplex based
on a Baudot Distributor.  If the signals are slow enough, they could be
each switched in turn to the potentiometer and the output switched in
turn to a storage capacitor or filter.  

Something like a ring-of-8 counter could drive two CMOS one-of-8
selector switches for the input and one for the output of the pot.  As
long as the clock frequency was well above the analogue frequencies
being switched, it should work reasonably well.


-- 
~ Liz Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

[toc] | [prev] | [next] | [standalone]

Page 1 of 2  [1] 2  Next page →

Back to top | Article view | sci.electronics.design

csiph-web