Post Go back to editing

What is the Input Resistance of the OPA184?

Hello community,

Title of the post says it all. I think I might be able to use this amplifier, but I have a requirement for Gohm+ R input.

Thanks

-Pete

  • Hi Harry, thanks for picking up my question.

    Sorry, yes, I meant OP184. I have been looking at many, many amps the last two days.

    I've got an existing device which I am trying to incrementally upgrade. I've got the relevant section below.

    For my inputs (RE1/RE2) I have a signal which can range from -5V to +5V, the 6V nets are actually +/-6.5V.

    RE1/2_SH are the cable shields and driven by the opamp. They are essentially antenna stubs.

    My input requirements are:
    +/-5V input (on 6-6.5V rails, this is proving to be difficult)

    Ghom+ Common mode impedance

    2Mhz+ GBW

    <= 1uA bias current

    Previously, the device was configured for +/-3V operation and used an LTC6242HV, which was suitable.

  • Peter,

    "OPA" is a T.I./Burr-Brown prefix;  did you mean OP184?

    The next generation part is out, the ADA4084-1, so let's use that one. 

    You didn't say which you wanted, but the differential input impedance and common mode input

    impedance is in the spec table on page 4 of the rev H data sheet.

    If you connect an op amp as a non-inverting amplifier, the input impedance gets bootstrapped

    by the amount of gain, so it's very high.  However, usually, rather than the resistive part,

    above 10 Hz or so, the input capacitance is of more concern.

    What are you trying to do?  Schematic?

    Harry

  • Hi Harry,

    I need to note that I've inherited the design the PCB exists and I'm working under a good number of constraints.

    It is not clear to me either why C180 and C181 are present. The original engineer did not leave a justification, I suspect he found it on a datasheet or app note somewhere. I would probably not have either the Cap or R109/R114. It seems to me to make more sense to connect it directly to the terminal. But, then again, I'm sort of out of my depth here.

    I would like to run +/-4 Signal. But the sales guys really want 5. I'd increase my supply voltage to +/-15, but then I need to re-engineer the entire rest of the system (which they also do not want to do). I'm really caught between a rock and a hard place.

    Generally, the source impedance is low. Typically it is a Fuel Cell, down to milli-ohm impedances, even at frequency. So, it is not clear to me why there is a Gohm imput impedance requirement, and realistically, the highest frequency we are interested in is 100kHz, so I'm not clear on why the GBW is so high. I think they are flexible on these points, but I need to be able to justify it, and sadly, I am not quite up to snuff here.

    Maybe more than recommending a specific op-amp, is there a good primer on how the opamp characteristics work in a circuit like this?

  • I've got the AD8610 in the circuit now, and it does well, as far as I've been able to evaluate it. Except that on the -5V input, there is a very noticeable 2-3% error. I think my input is too close to the rails.

  • In this area, I catch a break. The temp range is fairly mild, 5-35C Ambient. Inside case temp might get to 60C.

  • Harry,

    Thanks so much for all your help. Looks like I've still got a lot

    to address.

  • Peter,

    If it really is a cable shield that doesn't connect to anything, it's not clear why you want C180 and C181???

    Also, without knowing the source impedance of the +/-5V signal, it's hard to say what C208 and C209 do,

    if anything.

    For 2MHz, the AD8610 is the only op amp I found that is fast enough and runs on +/-6.5V.  We have

    a lot of high speed processes, but usually a 10 or 11 or 12V abs max.  So if you ran on +/-5V

    with +/-4V signals, you would have a lot more choices.

    With a perfect op amp, the CMRR of the difference amp stage is limited to 54 dB by the 0.1% resistors

    R112/113/115/117.  You may want to buy a four resistor array with guaranteed matching to 0.01%.

    With a few tenths of a pF difference on the U26 inputs, the AC CMRR will be less at 2 MHz.

    Harry

  • Yes, on +/-5V, the spec table only guarantees the Input Voltage Range (IVR) to be -2V to +3V

    so you are doing better than expected.  The ADA4084-1 is RR I/O.

    What temp range for the system?

    Harry

  • O.K., Leaving C208 and C209 in won't hurt anything.  With a very low  source impedance, the low

    pass filter corner is way up high.

    However, For the shield drive, With 1k and 1000 pF, you have a corner at 158 kHz, so if you have

    a 1 MHz signal, the shield is not being driven at all.  You really need to take C180 and C181 out

    or at least make them 100 pF and see what happens.

    You kind of have an InAmp with a first stage gain of one, so check out:

    A Designer's Guide to Instrumentation Amplifiers, 3RD Edition, 2006 | Education | Analog Devices

    A Deeper Look into Difference Amplifiers

    Harry