Can I use an OP77 ultra low noise opamp on a single power supply?

Discussion in 'The Projects Forum' started by Spottymaldoon, Jun 10, 2016.

  1. Spottymaldoon

    Thread Starter Member

    Dec 4, 2015
    53
    1
    I am working on a trace gas analyzer that requires an extremely high gain, low noise input stage and I have had very good results using Analog Devices' OP77 with a huge feedback resistor (10G ohm) and a gain of 1000. My input is DC and always positive.

    Up until now I have powered this with ±5V. Can this device operate on +5V power alone and, if so, what is the disadvantage?

    If this can't be done can you recommend an alternative device?
     
  2. wayneh

    Expert

    Sep 9, 2010
    12,571
    3,375
    It's fine. All op amps operate on a single supply: They have two power pins with a ∆V across them. As long as the inputs stay within the common mode range, it works. Some op-amps have a CMR that includes both power rails. That's called "rail-to-rail". Some have a CMR that includes only one of the power rails. Those are called"single supply". Op-amps that cannot sense at either rail are often called "dual supply".
     
  3. bertus

    Administrator

    Apr 5, 2008
    15,959
    2,449
    Hello,

    The OP77 will not work for rail to rail.
    If you want rail to rail functionality, have a look at the following table.
    There is given I/O for rail-to-rail in and output and O for output rail-to-rail only.

    Single_supply_opamps.jpg

    Also have a look at appendix A of the attached PDF for single supply circuits.

    Bertus
     
    ronv likes this.
  4. Spottymaldoon

    Thread Starter Member

    Dec 4, 2015
    53
    1
    Weyneh and Bertus, thanks for your prompt and useful replies - I don't need rail to rail as I have flexibility on my V+. I plan to connect the OP77 output directly to a 16 bit AD converter which has a nominal 0-6V input range
     
  5. SgtWookie

    Expert

    Jul 17, 2007
    22,183
    1,732
    The OP77's output cannot get within about 1 volt of the negative rail, even with an output load impedance > 10k. If you absolutely have to use a single supply rail, the OP77 will not be suitable unless it's output will always exceed about 1.5v.
     
  6. crutschow

    Expert

    Mar 14, 2008
    13,873
    3,507
    What voltage swing do you need from the output op amp?
    Does it need to go down to zero volts?
     
  7. Spottymaldoon

    Thread Starter Member

    Dec 4, 2015
    53
    1
    SgtWookie and Cutschow, thank you for your inputs which are both sensible and useful. The nature of this type of measurement is that there is a moderate dc background level with the signal I am measuring which, so I hope, will translate to at least 1.5V of the total signal - a background that I'd need to remove in any case. The OP77, for all its limitations, seems a unique device with respect to its low noise, high intrinsic gain and high input impedance and I have used its predecessors with great success in the past with dual supplies. Of the multiple powered systems in the instrument, only the electrometer needs ± supply and so I'm hoping to eliminate that. However the proof of the pudding will come shortly when I run the thing up after almost a year of machining and building a set of circuits which are all new to me - I have had wonderful help here at All About Circuits: especially with a novel UHF driver which powers a special electrodeless lamp (photoionization) - I couldn't be more grateful for all you guys' help.
     
  8. Lestraveled

    Well-Known Member

    May 19, 2014
    1,957
    1,218
    Have you considered using a switched capacitor negative voltage converter like an LTC1144? This chip and two capacitors will give you a negative voltage almost equal to the positive supply at low currents. They are very convenient. There are a lot of these type chips around. Google,"switched capacitor voltage converter".
     
    wayneh likes this.
  9. Spottymaldoon

    Thread Starter Member

    Dec 4, 2015
    53
    1
    Lestraveled - no I hadn't and confess these devices are new to me. It could be a very good way to go but I need to check on the potential interference from the device's switching frequency with my super-sensitive opamp config - both through the supply rail and by radiation. Thanks for the idea.
     
  10. Spottymaldoon

    Thread Starter Member

    Dec 4, 2015
    53
    1
    I've recently returned to this project (using a ± 5 volt supply as Lestraveled suggested and thanks, that part works fine) which requires the amplification of a DC signal of the order of -10uV, embedded in a stable 100 MHz background. I am using, an OP77 op-amp with feedback resistors of 10M and 10G ohm on the inverting input and the non inverting input grounded. The 10M input resistor has two 10pF mica capacitors connected pi fashion to ground to remove RF and all the inverting input is isolated from ground using Teflon stand-offs. I should mention that a very similar arrangement has worked historically.

    Meticulously checking everything, when I switch on, the output immediately pegs at around 4.8V (this is not a rail-to-rail device of course) which in itself is not surprising with 1000 gain and when I touch the input with my finger, the output swings lower and negative seems to have good range BUT, the moment I release my finger, the thing pegs positive again. Something seems to be injecting a negative voltage - but the device offset cannot be so high.

    Furthermore if I ground the actual inverting pin on the OP77 itself the output remains pegged positive - I think this should not happen!

    And this is all before there has been any exposure to RF.

    I have applied Sherlock Holmes' rule: "Once you eliminate the impossible, whatever remains, no matter how improbable, must be the truth." but the snag is that I can't think of anything more improbable to try - other than ruling myself insane. Clearly I am doing something fundamentally wrong - please can you assist?
     
  11. crutschow

    Expert

    Mar 14, 2008
    13,873
    3,507
    It would appear to be the usual problem of not reading the data sheet. :rolleyes:
    Did you read this line in the data sheet upload_2017-1-5_16-23-24.png
    and calculate how much output offset that would generate with those values of input and feedback resistors?
     
  12. Spottymaldoon

    Thread Starter Member

    Dec 4, 2015
    53
    1
    It would appear that I've been sent to the corner - richly deserved too! Thanks, crutchow, you're right of course.
     
  13. dannyf

    Well-Known Member

    Sep 13, 2015
    2,196
    420
    it will depend on the sensor. If the sensor is only referenced to ground, you will need to make sure that the opamp, whatever you want to use, can pick up a signal that low.

    op77 for an opamp is quite ordinary (some may say obsolete). so there should be lots of replacement, unless your design uses something particular about that opamp.
     
  14. dannyf

    Well-Known Member

    Sep 13, 2015
    2,196
    420
    input offsets and limited common mode range would be my guess. I would use two stages for this kind of gain, and try to use an instrumentation opamp, or at least a differential opamp for this.

    Also, if your sensor allows some lifting off the ground, it would be helpful as well.
     
Loading...