Author Topic: Any way to do DC voltage "cross-fading"?  (Read 760 times)

0 Members and 1 Guest are viewing this topic.

Offline K3mHtHTopic starter

  • Regular Contributor
  • *
  • Posts: 91
  • Country: ca
Any way to do DC voltage "cross-fading"?
« on: October 27, 2023, 02:54:10 am »
I want a circuit that takes an input control voltage (0-5V DC) and as that voltage goes from 0 to 5V the output voltage goes from A to B.

A and B are arbitrary between 0 and 5V, set with potentiometers with either one being bigger than the other.

Seems pretty easy, but nothing I've come across in any of my books.. any ideas?







 

Offline ArdWar

  • Frequent Contributor
  • **
  • Posts: 373
  • Country: sc
Re: Any way to do DC voltage "cross-fading"?
« Reply #1 on: October 27, 2023, 03:22:21 am »
Seems pretty easy
Heh, you bet

Are you looking for an analog multiplier function? (i.e Vout = Vin * Vctrl, or in your case Vout = Vin * Vctrl/5)

That's indeed not exactly trivial circuit to do purely in analog. Depends on your bandwidth, dynamic range, linearity, and smoothness requirement it may require fairly exotic parts like AD633 or finicky log amplifier circuits.
It might be easier to do it with back to back ADC-DAC, or even opamp with digipot controlled gain if you're okay with some quantisation. You may also want to look at Variable Gain Amplifiers/Programmable Gain Amplifiers, but their gain response are often logarithmic (in dB)
« Last Edit: October 27, 2023, 03:30:28 am by ArdWar »
 

Offline Whales

  • Super Contributor
  • ***
  • Posts: 1899
  • Country: au
    • Halestrom
Re: Any way to do DC voltage "cross-fading"?
« Reply #2 on: October 27, 2023, 03:29:42 am »
Simplest to design: software on a microcontroller that has a DAC output.  Pots could be read with ADCs.

Staying analog, inaccurate: mix two channels, attenuate them individually by driving a transistor in the linear region.  Will require effort to combat nonlinearity.

Staying analog, accurate: motor-driven potentiometers.

Staying analog, accurate: analog multipliers (such K3mHtH suggests) followed by summing.

Offline K3mHtHTopic starter

  • Regular Contributor
  • *
  • Posts: 91
  • Country: ca
Re: Any way to do DC voltage "cross-fading"?
« Reply #3 on: October 27, 2023, 05:34:44 am »
Heh, you bet

Are you looking for an analog multiplier function? (i.e Vout = Vin * Vctrl, or in your case Vout = Vin * Vctrl/5)

That's indeed not exactly trivial circuit to do purely in analog. Depends on your bandwidth, dynamic range, linearity, and smoothness requirement it may require fairly exotic parts like AD633 or finicky log amplifier circuits.
It might be easier to do it with back to back ADC-DAC, or even opamp with digipot controlled gain if you're okay with some quantisation. You may also want to look at Variable Gain Amplifiers/Programmable Gain Amplifiers, but their gain response are often logarithmic (in dB)

There's actually 3 inputs, Vctrl, Vin-x and Vin-y. The input Vctrl going from 0-5V gets "mapped" to going from Vin-x to Vin-y.

 

Offline EPAIII

  • Super Contributor
  • ***
  • Posts: 1069
  • Country: us
Re: Any way to do DC voltage "cross-fading"?
« Reply #4 on: October 27, 2023, 06:32:53 am »
Your 0 to 5 Volts goes to a pair of OP amp circuits with the same gain: one inverting and the other non-inverting.

The common gain factor is selected to power a pair of LEDs. One LED is bright when the other is dim and vice-versa. They cross fade.

Each LED shines on a photoresistor with external light excluded (heat shrink, tube and some chalk, etc.). Of course, the light from the other LED should also be excluded.

https://www.mouser.com/c/?q=photoresistor

The photoresistors are used as two input resistors of a third OP amp with an adder circuit configuration.

A is input through one photoresistor.

B is input through the other photoresistor.

The output of that third OP amp is a signal that fades back and forth from A to B and B to A.


I am sure this will work beyond DC. But I have no idea of what bandwidth can be possible. Audio? Probably. Base band, analog video? Perhaps, perhaps not. Linearity? Perhaps a challenge.

Have fun experimenting.
« Last Edit: October 27, 2023, 06:36:11 am by EPAIII »
Paul A.  -   SE Texas
And if you look REAL close at an analog signal,
You will find that it has discrete steps.
 

Offline fourfathom

  • Super Contributor
  • ***
  • Posts: 1884
  • Country: us
Re: Any way to do DC voltage "cross-fading"?
« Reply #5 on: October 27, 2023, 07:14:35 am »
What's the response rate needed?  It might be easiest to do this with digital duty-cycle control of a 2:1 analog mux, with the common output fed to a low-pass filter.  The 0-5V input could be read by an uController ADC, and the controller would then generate the duty-cycle output.  Or you could also skip the controller, and use a couple of comparators to generate the duty-cycle signal (one to generate a triangle wave, the other to compare the triangle to the input control voltage.)

We'll search out every place a sick, twisted, solitary misfit might run to! -- I'll start with Radio Shack.
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 19533
  • Country: gb
  • 0999
Re: Any way to do DC voltage "cross-fading"?
« Reply #6 on: October 27, 2023, 08:21:16 am »
The  AD633, or if you need something faster and more accurate, the MPY634.
https://docs.rs-online.com/6f7d/0900766b8077fe4b.pdf
https://www.ti.com/lit/ds/symlink/mpy634.pdf
 

Offline Terry Bites

  • Super Contributor
  • ***
  • Posts: 2397
  • Country: gb
  • Recovering Electrical Engineer
Re: Any way to do DC voltage "cross-fading"?
« Reply #7 on: October 27, 2023, 08:32:39 am »
A dual VCA, eg AS3360,  Connect the inputs in anti-parallel and sum the output currents in a single resistor. Tie the CV inputs.
The output drift for this amp is about 0.36%/C.
You can do better at a similar overall cost with an LT1251 fader.



 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf