Post Go back to editing

ad5235 A to B resistance measurement not stable with Ohm meter

Category: Hardware
Product Number: ad5235B-250
Software Version: none


I'm am testing an ad5235B-250 digital pot.  I can write and read to the spi registers, and am powered by a clean 3.3V signal and the output pot is unloaded. 

When I take an Ohm meter and measure across A to B on either port, the resistance value changes.

I expect it to read the same value (~250K) every time as it's fully unloaded. 

For example, I program the port zero wiper to 0 using B10100000.  I can read this back and confirm the write worked. 

From a0 to b0 I get 187K (should be 250K).  a to wiper is 183K, b to wiper is 100.  So where did the 4K go?  

Then I program the port zero wiper to 512 using B10100000.  I can read this back and confirm the write worked. 

From a0 to b0 I get 208Kohm (should be 250K).  a to wiper is 105K, b to wiper is 108K.  So this is 203K, it should be 208K (should be 250K).

Why is the a to b reading changing with programming, and why does a to wiper to b add up to something different than a to b?

s there something I'm doing wrong to make this thing stable?

Is there some reason a benchtop meter would be messing things up?  Does it need a DC bias across the pot or something?



Thread Notes

Parents Reply
  • Hi Dave,

    The leakage current from the device can affect the ohmmeter readings, adding unwanted errors. You can put an op-amp as a buffer and measure the voltage at the buffer output. Another method is to use a source-meter where a precision current source/sink is fed through one of the terminals with the other terminal held at GND or some other level and voltage is measured across the same terminals.

    Best Regards,


  • Thanks for the reply.  To investigate further, I put a 3.08V supply in series with a 264200 Ohm resistor in series with the A port and grounded the B port.  The vout at A was 1.17V which means the series resistance from A to B was approx 160K.  So I think there is an issue with the AD5235 that I have, I would expect somewhere close to 250K.  

    Any tips to further debug this?  Same setup as I describe above.  I would think not matter what the programming, it should always be 250K across A to B.  

  • Hi,

    What is connected to the wiper terminal? have you tried shorting B to wiper then measure it through the wiper?

    Best Regards,