Author Topic: Voltage Divider into ADC  (Read 12483 times)

0 Members and 1 Guest are viewing this topic.

Offline logictomTopic starter

  • Supporter
  • ****
  • Posts: 336
  • Country: au
Voltage Divider into ADC
« on: April 20, 2012, 04:08:11 pm »
I want to take a 0-10V signal and feed it into a 0-3.3V ADC, are there any general rules to using a voltage divider into an ADC.
I usually stick with values in the kohms range to limit current draw, should I go higher/lower - are there any specs to look up in the micro's datasheet to design against?
How do I go about designing the divider if I know it's xbits and y-zV range so I get sensible bits/voltage scaling in the software?

 

alm

  • Guest
Re: Voltage Divider into ADC
« Reply #1 on: April 20, 2012, 07:55:50 pm »
This recent thread is about almost the same topic:
https://www.eevblog.com/forum/projects-designs-and-technical-stuff/voltage-divider-on-the-output-of-a-voltage-reference-can-i/msg105854/

ADC datasheets will usually specify a maximum source impedance. You can scale the voltage however you like. You could divide it by ~3 to get the highest resolution. In the case of a 10-bit ADC, 1 bit would now correspond to 10 V / 1024 (3.3 V / 1024 * 3). Stability of the voltage divider will determine the effective accuracy of your ADC.
 

Offline free_electron

  • Super Contributor
  • ***
  • Posts: 8550
  • Country: us
    • SiliconValleyGarage
Re: Voltage Divider into ADC
« Reply #2 on: April 20, 2012, 09:18:38 pm »
keep in mind that some a/d have a pretty low input impedance. also , this impedance may change depending on what you apply.
it's good to always put a buffer amplifier between your sensing node and the input of the a/d
Professional Electron Wrangler.
Any comments, or points of view expressed, are my own and not endorsed , induced or compensated by my employer(s).
 

Offline Leo Bodnar

  • Frequent Contributor
  • **
  • Posts: 816
  • Country: gb
Re: Voltage Divider into ADC
« Reply #3 on: April 20, 2012, 11:26:04 pm »
ADC usually draws current in short bursts when their S&H cap is being charged.  If you are concerned about low ADC input impedance you can place a capacitor in parallel with ADC input.  In practical terms a divider and a good quality ceramic 0.1uF is all you need unless you have to maintain ADC's accuracy when measuring this 10V bus voltage.

Leo
« Last Edit: April 20, 2012, 11:27:58 pm by Leo Bodnar »
 

Offline logictomTopic starter

  • Supporter
  • ****
  • Posts: 336
  • Country: au
Re: Voltage Divider into ADC
« Reply #4 on: May 03, 2012, 10:13:37 am »
Thanks for the feedback - some food for thought.
Stability of the voltage divider will determine the effective accuracy of your ADC.
Would this be down to voltage supply stability, resistor drift over age and due to temperature? Any other major contributors?

In practical terms a divider and a good quality ceramic 0.1uF is all you need unless you have to maintain ADC's accuracy when measuring this 10V bus voltage.
Would you mind elaborating on this a little more? How does it effect the accuracy and if you wanted to maintain the accuracy what would you do differently?

Thanks.
 

Offline AlphZeta

  • Regular Contributor
  • *
  • Posts: 242
  • Country: us
    • Kerry D. Wong
Re: Voltage Divider into ADC
« Reply #5 on: May 03, 2012, 04:46:54 pm »
You will need to pay special attention to the input impedance of the ADC. Some ADCs (e.g. AD7706) have the option of configure the input as buffered or unbuffered. When you are using unbuffered mode, the impedance can be quite low (e.g. under 10K) and the divider resistors need to be chosen carefully to avoid losing accuracy.
 

Offline hlavac

  • Frequent Contributor
  • **
  • Posts: 536
  • Country: cz
Re: Voltage Divider into ADC
« Reply #6 on: May 04, 2012, 10:06:44 pm »
How does it effect the accuracy and if you wanted to maintain the accuracy what would you do differently?

The cap will create sort of a RC low pass filter.  Which may even be a good thing if you want less noise and the input voltage does not change too much (e.g. measuring a power rail or battery voltage).

It may be bad for fast changing signals like audio/RF - in that case I would recommend a buffer. You may still need a low pass filter to avoid aliasing (filter out frequencies above half the sample rate). I think you could do that filter/buffer using a single op-amp.
Good enough is the enemy of the best.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf