Author Topic: How to measure the voltage across a potentiometer ratiometrically  (Read 1374 times)

0 Members and 1 Guest are viewing this topic.

Offline RerouterTopic starter

  • Super Contributor
  • ***
  • Posts: 4705
  • Country: au
  • Question Everything... Except This Statement
Struggling with a concept, I am trying to convert a signal from a variable resistance sensor to a voltage with a way of compensating out the supply voltage, I am trying to keep everything analog,

The issue I face is, with most bridge differential amplifier methods, if you halve the supply voltage, you halve the difference between the 2 inputs, meaning half the output voltage,

Other methods require control of the ground node to compensate there, that node is where the sensor bolts into an engine, so left unable to change that,

So with all that in place, are there any op amp circuits that given a voltage divider like a potentiometer, to have a consistent output when the voltage across the potentiometer changes.

Linear would be ideal, but can work around it, and hoping for an output voltage in a 0-5V range, don't mind which direction, last point is, I only have a single supply available.

Anyone come across a circuit like this in the past?
 

Offline SiliconWizard

  • Super Contributor
  • ***
  • Posts: 15800
  • Country: fr
Re: How to measure the voltage across a potentiometer ratiometrically
« Reply #1 on: May 09, 2020, 01:38:06 pm »
Of course one of the easiest approach would be to insert a voltage regulator/reference of some kind between the input voltage and the divider. Would that be a problem here? I see you're looking for an output voltage in the 0-5V range, and your min input voltage seems to be 5V. Should be workable with any decent LDO. What did I miss?
 

Offline RerouterTopic starter

  • Super Contributor
  • ***
  • Posts: 4705
  • Country: au
  • Question Everything... Except This Statement
Re: How to measure the voltage across a potentiometer ratiometrically
« Reply #2 on: May 09, 2020, 01:41:31 pm »
There are a few edge cases where I don't control that, but I may just have to say tough luck. That is my fall back plan if nothing like this exists,
 

Offline Picuino

  • Super Contributor
  • ***
  • Posts: 1072
  • Country: es
    • Picuino web
Re: How to measure the voltage across a potentiometer ratiometrically
« Reply #3 on: May 09, 2020, 02:11:13 pm »
Search for wheatstone bridge.
Amplifiers and compensation.
 

Offline mikerj

  • Super Contributor
  • ***
  • Posts: 3382
  • Country: gb
Re: How to measure the voltage across a potentiometer ratiometrically
« Reply #4 on: May 09, 2020, 02:12:47 pm »
The other possibility is driving the sensor with a constant current.
 

Offline SiliconWizard

  • Super Contributor
  • ***
  • Posts: 15800
  • Country: fr
Re: How to measure the voltage across a potentiometer ratiometrically
« Reply #5 on: May 09, 2020, 02:30:18 pm »
There are a few edge cases where I don't control that,

Don't control what exactly? The input voltage range? Do you mean it could get under 5V? (If so, and you need an output that gets up to 5V, you'd need an additional supply anyway. In that case, you could consider a buck-boost converter.)
 

Offline Picuino

  • Super Contributor
  • ***
  • Posts: 1072
  • Country: es
    • Picuino web
Re: How to measure the voltage across a potentiometer ratiometrically
« Reply #6 on: May 09, 2020, 04:13:50 pm »
Look for LT3092 adjustable current source.
https://www.analog.com/en/products/lt3092.html

With constant current of 5mA, the voltage across R is proportional to its resistance and in the range 0-5V.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17428
  • Country: us
  • DavidH
Re: How to measure the voltage across a potentiometer ratiometrically
« Reply #7 on: May 10, 2020, 12:08:08 am »
So with all that in place, are there any op amp circuits that given a voltage divider like a potentiometer, to have a consistent output when the voltage across the potentiometer changes.

What you are suggesting is to vary the gain with the supply voltage which would require a multiplier.

Ratiometric circuits do this by using the same voltage for the sensor excitation and the converter reference so the multiplication happens inside the converter.
 
The following users thanked this post: Rerouter

Offline RerouterTopic starter

  • Super Contributor
  • ***
  • Posts: 4705
  • Country: au
  • Question Everything... Except This Statement
Re: How to measure the voltage across a potentiometer ratiometrically
« Reply #8 on: May 10, 2020, 01:16:33 am »
Ah ok a multiplicative circuit, that makes sense,

Going to play around with seeing if I can make it with a PWM multiplier, as I'm already using one elsewhere in the circuit.

Triangle wave vs divided supply voltage, double the supply voltage means half the on time, means half the voltage on the output on average, sounds like it should work in theory.
 

Offline RerouterTopic starter

  • Super Contributor
  • ***
  • Posts: 4705
  • Country: au
  • Question Everything... Except This Statement
Re: How to measure the voltage across a potentiometer ratiometrically
« Reply #9 on: May 10, 2020, 01:25:24 am »
SiliconWizard, The exciter voltage could be external, e.g. tapping in to another system currently using the resistive sensor, Its an edge case that is not that common, If I can keep the option, then means I don't have to compromise on it.
 

Offline RerouterTopic starter

  • Super Contributor
  • ***
  • Posts: 4705
  • Country: au
  • Question Everything... Except This Statement
Re: How to measure the voltage across a potentiometer ratiometrically
« Reply #10 on: May 10, 2020, 08:35:42 am »
Bit of an odd implementation, but a RC curve for the PWM reference gets the multiplier within 1-2% error over the range that I am after, If I understood the math better I could probably expand that range, but for now this more than suits,

Right now 10-30V is the stable range, anything less is not, but if used with a 5V supply it would be a stable supply anyway, If anyone knows any other methods I'm willing to give them a try, or any criticism to my weird circuit.

edit: technically only 1 input is constrained, the other could be bipolar if the mosfet right most mostfet was replaced with an analog switch.
« Last Edit: May 10, 2020, 09:00:52 am by Rerouter »
 

Offline SiliconWizard

  • Super Contributor
  • ***
  • Posts: 15800
  • Country: fr
Re: How to measure the voltage across a potentiometer ratiometrically
« Reply #11 on: May 10, 2020, 01:39:34 pm »
SiliconWizard, The exciter voltage could be external, e.g. tapping in to another system currently using the resistive sensor, Its an edge case that is not that common, If I can keep the option, then means I don't have to compromise on it.

That doesn't change anything much. I already figured the input voltage may not be fixed.

Whatever you do, the main thing to consider is the possible range of this voltage. Almost no solution will work for an arbitrarily low or high voltage, for instance.

You absolutely *need* to have a clear requirement on the acceptable voltage range, from which we can work out a solution.
If the min "exciter" voltage is guaranteed to be 5V (as your schematic would suggest), then an LDO should be appropriate.

If it can be lower than 5V, how low?
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17428
  • Country: us
  • DavidH
Re: How to measure the voltage across a potentiometer ratiometrically
« Reply #12 on: May 10, 2020, 01:49:34 pm »
Going to play around with seeing if I can make it with a PWM multiplier, as I'm already using one elsewhere in the circuit.

Triangle wave vs divided supply voltage, double the supply voltage means half the on time, means half the voltage on the output on average, sounds like it should work in theory.

That could work.  I have also seen it done with charge pump type circuits where the frequency varies with the excitation voltage.
 

Online jbb

  • Super Contributor
  • ***
  • Posts: 1265
  • Country: nz
Re: How to measure the voltage across a potentiometer ratiometrically
« Reply #13 on: May 10, 2020, 08:48:20 pm »
I think you can do something a bit cleaner. To check I’m on the right track, is this what you want?
- Vin 5V to 30V
- Sense resistor Rx 0 to 5 kOhm
- Vout = Vin * Rx / 5 kOhm (ratiometric)

If so, that wide voltage range will be a bit of a challenge...

Questions:
Will Vout need to provide any current?
How fast does it need to be?
 

Offline RerouterTopic starter

  • Super Contributor
  • ***
  • Posts: 4705
  • Country: au
  • Question Everything... Except This Statement
Re: How to measure the voltage across a potentiometer ratiometrically
« Reply #14 on: May 10, 2020, 08:59:33 pm »
- Vin 5V and 10.8-28V (I revised this a little when trying to make my own multiplier)
- sense resistor 3 - 1000 Ohm
- Vout = Vin * some constant, main thing is that it is consistant when the supply voltage changes

lets say no output current, as I can buffer the output
response time within 1 second.
 

Online jbb

  • Super Contributor
  • ***
  • Posts: 1265
  • Country: nz
Re: How to measure the voltage across a potentiometer ratiometrically
« Reply #15 on: May 11, 2020, 02:01:44 am »
Sorry, was that 1000 Ohm max or 5k max?

I think we can do this without analog multiplier chips or choppers.  The scheme I’m thinking of will drive the sensor with a constant current that is proportional to the supply rail.

Do you have any maximum current or voltage or power specs for the sensor (variable resistor)?
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf