Author Topic: Resistor Imitation  (Read 1619 times)

0 Members and 1 Guest are viewing this topic.

Offline vonorfasyexelaTopic starter

  • Newbie
  • Posts: 2
  • Country: ru
Resistor Imitation
« on: October 09, 2019, 09:04:16 am »
Hi guys!
This is my first post here so please pardon mistakes etc.

I have a device which can measure the resistance of a conductor using 3-wire measurement.
To check that is works correctly we need a some kind of resistance imitator.
For simplicity let's say that the device drives 1mA current into this imitator and measure the voltage across it's terminals.
Then it calculates the resistance dividing voltage into 1 mA.

I have developed this schematic as a resistor imitator:
851056-0

But when I power up this schematic (and NOT pushing any measure current between OUT and ground) I got this:
851044-1

Zoom in:
851048-2

So the gate voltage increases over time until it reaches around 4.2V. Then we can see some oscillation.
This process repeats with some more or less stable period.

Maybe someone can explain why gate voltage changes this way?
If you need some extra info about the schematic etc please ask.

 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: Resistor Imitation
« Reply #1 on: October 09, 2019, 12:46:23 pm »
The circuit won't work. The op-amp will attempt to adjust its output voltage so that both inputs are the same voltage, which is 2.2*5(68+2.2) = 0.157V in this case.

Why not just use a variable resistor?
 
The following users thanked this post: 001

Offline SiliconWizard

  • Super Contributor
  • ***
  • Posts: 15797
  • Country: fr
Re: Resistor Imitation
« Reply #2 on: October 09, 2019, 05:21:36 pm »
Yep. I didn't quite get where the "terminals" of this "virtual resistor" (which it's not) would be either?
 

Offline 001

  • Super Contributor
  • ***
  • Posts: 1170
  • Country: aq
Re: Resistor Imitation
« Reply #3 on: October 09, 2019, 05:35:05 pm »
U muddle with resistor vs load conception
 

Offline SiliconWizard

  • Super Contributor
  • ***
  • Posts: 15797
  • Country: fr
Re: Resistor Imitation
« Reply #4 on: October 09, 2019, 05:46:27 pm »
Couple thoughts:

- Is your measurement device supposed to do precision measurements? If so, checking it would be best done with a series of precision resistors. Possibly a precision decade resistor box, or something.
- If you don't care much about accuracy, but only want to do a rough check, using a variable resistor is obviously much simpler.
- If you thought about implementing an active resistor just to be able to control the resistor value from a voltage/programmatically, you could just use a digital potentiometer instead (just check its specs carefully, especially voltage range.) I doubt you'll do better accuracy-wise implementing an "active resistor".
- If you're still interested in implementing "active resistors", just look that term up in your favorite search engine. The most basic active resistor can be implemented with just one MOSFET. Will that accurately mimic a true resistor? Hardly. Will it be easily tunable? Not so much. But it's often used in CMOS design, as they take up a lot less area than "true" resistors. You'll also find more sophisticated implementations,  but you'll have a hard time getting any to be even as accurate as a simple digital pot.

 
The following users thanked this post: 001, vonorfasyexela

Offline mzzj

  • Super Contributor
  • ***
  • Posts: 1298
  • Country: fi
Re: Resistor Imitation
« Reply #5 on: October 09, 2019, 07:23:54 pm »
https://www.eevblog.com/forum/projects/voltage-to-resistance-converter/

read teslacoils post.

You could do it also with ADC-DAC combo.

 
The following users thanked this post: vonorfasyexela

Offline vonorfasyexelaTopic starter

  • Newbie
  • Posts: 2
  • Country: ru
Re: Resistor Imitation
« Reply #6 on: October 10, 2019, 08:49:45 am »
Thanks for the feedback!

Forgot to say that the range of imitation should be from 0 to 200 Ohm with 1 Ohm step and ~0.5% tolerance..
So the overall idea is to set the reference voltage from 0 mV to 200 mV on the opamp input.
In my simplified scheme it is 157 mV.
Then when the measure current (1 mA) will flow through the mosfet the opamp will keep the reference voltage drop over the mosfet.
And the measure device will "think" that it measures 157 Ohm, for example.

Why not just use a variable resistor?
The problem with a variable resistor is it's "wiper" which has non-zero resistance itself. At least I never see pots with less than 50 Ohm resistance. Maybe you know such?

Couple thoughts:
Actually the previous version of the imitator worked with a resistor box. And it worked great. But took a lot of place.
And I did some tests with just one MOSFET as well. It has horrible dependance from the temperature. Even the warm of a finger changed the resistance for several Ohms..
That's why I started to think about opamp that will "track" any changes in resistance and regulate them.

https://www.eevblog.com/forum/projects/voltage-to-resistance-converter/
Thanks for the link! It has several ideas, need to think about them.
« Last Edit: October 10, 2019, 08:54:30 am by vonorfasyexela »
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: Resistor Imitation
« Reply #7 on: October 10, 2019, 03:06:41 pm »
Thanks for the feedback!

Forgot to say that the range of imitation should be from 0 to 200 Ohm with 1 Ohm step and ~0.5% tolerance..
So the overall idea is to set the reference voltage from 0 mV to 200 mV on the opamp input.
In my simplified scheme it is 157 mV.
Then when the measure current (1 mA) will flow through the mosfet the opamp will keep the reference voltage drop over the mosfet.
And the measure device will "think" that it measures 157 Ohm, for example.
So do you really need a variable resistor or just a 0mV to 200mV voltage source?

Quote
Why not just use a variable resistor?
The problem with a variable resistor is it's "wiper" which has non-zero resistance itself. At least I never see pots with less than 50 Ohm resistance. Maybe you know such?
I've not heard that before. The one linked below has a minimum resistance of 1Ohm, which will be the same as the wiper resistance.
https://www.mouser.co.uk/datasheet/2/54/3590-776468.pdf

Quote
Actually the previous version of the imitator worked with a resistor box. And it worked great. But took a lot of place.
And I did some tests with just one MOSFET as well. It has horrible dependance from the temperature. Even the warm of a finger changed the resistance for several Ohms..
That's why I started to think about opamp that will "track" any changes in resistance and regulate them.
Or are you trying to make a voltage controlled resistor and your comment about wiper resistance referred to digital, rather than mechanical potentiometers?

Making an accurate, stable, linear, voltage controlled resistor from analogue parts is not easy and it's darn near impossible to meet the requirements you've stated.

An op-amp and MOSFET can be used to scale down the resistance of a digital potentiometer, which can be controlled using an ADC.
https://www.eevblog.com/forum/projects/ptc-resistance-output-from-0-10v-analog-input/msg2454789/#msg2454789##

What's your budget? A ready made solution exists.
https://www.kele.com/Catalog/24%20Transducers/PDFs/RES-1E%20Series%20Catalog%20Page.pdf
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf