EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: dave_robinson_022 on October 15, 2021, 02:23:22 am
-
I have set up a basic circuit as shown. I am measuring voltage using differential input on a TLE2141 with a gain of 0.033. As expected I get the correct readings through various voltages up to 30V.
However, when I try to change the voltage back to positive with a gain of 1 I seem to lose approx 68mV regardless of what the input voltage is. I can't for the life of me work out why this is happening?
[attach=1]
-
The input bias current of the TLE214 can be up to 2 microamps which into the 50 kilohm parallel resistance at the inverting input will produce an offset voltage of up to 100 millivolts.
The TLE214 should be used with much lower values of resistance, however placing a 50 kilohm resistor in series with the non-inverting input will cancel out this error reducing it by about 20 times from the input offset current.
-
You can also reduce it 100x by making the resistors 1kΩ or 2000x times by making them 1kΩ and putting 500Ω in front of IN+ ;)
-
Awesome thank you.
The original circuit came from a power supply design and the 100k resistors were to protect the inputs to the OP AMP should the current sense resistor and/or MOSFET short.
Keeping this in mind I tried 10K and 330R which would still limit current to 6mV at 60V, and this worked. Oddly though there was a slight variation between 0.8mV at 15V to 5.7mV at 1V.
I also tried 51K in series with ground as suggested and this fixed the error to within 1.1mV over the same range