help with opamp input

Discussion in 'General Electronics Chat' started by minkey01, Oct 2, 2016.

  1. minkey01

    Thread Starter Member

    Jul 23, 2014
    171
    0
    Hello. This circuit is almost working but has one problem. Can someone offer some advice. Also if you see anything that could be improved let me know.

    The power supply is single sided at +5.34 V. The Vin is a jack and will get plugged in from a different source and will be in the range of 0 to +5V. The voltage divider network then brings the input down to 0 to +3.3V. The diodes are there to protect against accidental over voltage and negative voltage. The opamp is running off of 3.3V power from the Teensy micro-controller. It is setup as a unity gain buffer and then outputs into the micro-controller pin 14.

    When nothing is plugged into the Vin jack the micro-controller is reading some low volts jumping around. I imagine this is because the opamp + input is floating when nothing plugged into the jack? How can I fix this? I don't want to use a switching jack either.

    Thanks!

    op amp input.jpg
     
  2. hp1729

    Well-Known Member

    Nov 23, 2015
    1,938
    218
    The op amp is RRIO, so the circuit should be good. I might consider powering the op amp from 5.34 V instead. Do you really need that high a resistances on the input circuit. I would think 194K would be enough to keep the input from "floating", but I could be wrong. Maybe go to 10K and 19.4K instead?
     
  3. minkey01

    Thread Starter Member

    Jul 23, 2014
    171
    0
    Hi. The op amp is powered by 3.3V because the micro-controller inputs are only 3.3V tolerant with analog read. I chose the 100K resistor because that is the recommended impedance from the Vin source specs. Hmm..
     
  4. tindel

    Active Member

    Sep 16, 2012
    567
    193
    The input isn't floating... Its connected to ground througg the 194k resistor. Which is kind of high impedance... Maybe reduce your decide impedance to 10k/19.4k and see if you get better results.

    The other thing that could be happening is that the output of the amp can't go all the way to ground so it is probably out of regulation. Its been a while since i looked at the 6001 datasheet but the lowest voltage it can go to is probably somewhere around 35mV.
     
  5. #12

    Expert

    Nov 30, 2010
    16,248
    6,744
    The input common mode range misses both rails by 300mv.
    Zero input voltage is not allowed.
     
  6. minkey01

    Thread Starter Member

    Jul 23, 2014
    171
    0
    dang.. seems like a lot. i thought this opamp was much closer to the rails.
     
  7. tindel

    Active Member

    Sep 16, 2012
    567
    193
    The min input voltage is 300mV below Vss. Ground is a valid input - but the output will only go down to 25mV - so it's response will be non-linear below 25mV - hence my comment about being out of regulation.
     
  8. minkey01

    Thread Starter Member

    Jul 23, 2014
    171
    0
    ok. thanks. i guess i'll write a dead band into the software for that 0 to 25mV range.

    last stupid question: my 100K & 194K voltage divider network is not really exactly a 100k impedance input jack, right? do you know the formula to make it exactly the 100k that is called for while still dividing the 5V down to 3.3V?

    Thanks!
     
  9. #12

    Expert

    Nov 30, 2010
    16,248
    6,744
    Oops. I read it wrong. Sorry about that.:oops:

    (I got distracted.)
     
    Last edited: Oct 2, 2016
  10. minkey01

    Thread Starter Member

    Jul 23, 2014
    171
    0
    i think an OPA342 could get closer to the rails. maybe i'll order some of those.
     
  11. tindel

    Active Member

    Sep 16, 2012
    567
    193
    Within 25mV of the rail isn't enough? I would think so - but I don't know exactly what you're doing either. Another thought - does the A/D on the micro extend to ground? Make sure you're within the proper input limit of the micro.

    As far as sizing R1 and R2 (100k and 194k) resistors properly to meet 100kohm impedance. You need to solve for two equations to get the values that will meet the input impedance.

    Eqn 1
    Vout = Vin * (R2 / (R1+R2))
    Eqn 2
    100k = R1 + R2

    Solve for either R1 or R2 of one equation. Put the result into the other equation and use algebra to find the answer you're looking for.
     
  12. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,281
    1,232
    Tell me again why you want such a high input resistance?
     
    #12 likes this.
  13. minkey01

    Thread Starter Member

    Jul 23, 2014
    171
    0
    the high input resistance is an analog synth spec convention.

    the micro-controller can go to ground. i'm doing 10 bit ADC with the 3.3V. 10 bit has 0-1023 values, so 3.2258 mV for each value. The 25 mV away from the rail is losing a good 8 values or more.
     
  14. Sensacell

    Well-Known Member

    Jun 19, 2012
    1,124
    266
    Can you buffer first, then divide? You can have any impedance you want at the input, just a resistor to GND to bias the opamp.

    Power the opamp from the 5.3 V

    The divider after the buffer can be sized to whatever voltage and impedance levels the ADC needs.

    Over voltage to the ADC won't be a problem if you choose the divider resistance to keep the input protection diode current below the max level.

    The not-going-to-the-GND-rail problem is then diluted by the divider ratio, the not-going-to-the-positive-rail problem goes away.
     
    Last edited: Oct 3, 2016
  15. hp1729

    Well-Known Member

    Nov 23, 2015
    1,938
    218
    Sounds good.
     
Loading...