AMP02 in loadcell application

I'm an engineering designer formally trained in mechanical engineering, with experience in digital electronics (although never formally trained in electronics), and now have a need to employ an instrumentational amplifier in a loadcell application.  Having just swated up on basic op amps., I have a couple of questions.

1.  I am given to understand that when connecting the inputs to a loadcell (and, in particular, a thermocouple) I need to provide a bias current path for one or other of the inputs to earth or one of the supply rails.  However, as the device in question employs FET (as opposed to bipolar) inputs, do I not need to concern myself with providing a bias current path ?  If so, what resistor value should I use to earth ? (the loadcell has an impedance of 1K across the bridge and will be excited between 4.7 - 5.6 V via a resistive potential divider arrangement, between the + and - power rails with a zener across the excitation input to stabilize the voltage).

2.  I am currently running off supply rails of + and - 12.3 V  and am only getting an output voltage swing of +/- 10 V.  If I have interpreted your data correctly, an output swing of up to +/- 12 V is possible; will I achieve this if I increase the supply rail potentials by a few volts ?

I apologize if this is basic electronics and would thank anyone for their kind assistance.


Graham Bruce.

  • 0
    •  Analog Employees 
    on Jul 18, 2012 8:56 PM

    Hello Graham,

    Is there a specific reason you're considering AMP02 instead of a more modern amplifier? Newer amplifiers typically incorporate the expertise acquired over the years to provide superior performance parameters, and they are a better bang for your buck. If you are working on a new design, I will suggest that you take a look at our selection table which can be found by following this link.

    I will answer your questions since they would be valid concerns independent of your amplifier of choice. But I will break them up to make them easier to read.

    1. Bias current path. Input bias path is always necessary, independent of the transistor used for the input. This is true for FET, CMOS or bipolar inputs. Even if there is no base current, there are leakages that will charge input parasitic capacitances. Such leakages or their resulting voltage levels are unpredictable.

    2. AMP02 is not a FET-input device; it only implements an over-voltage protection based on FET devices.

    3. If the load cell is a Wheatstone bridge, the bias return path is established through the bridge unless the bridge and the amplifier supplies are isolated from each other.

    4. Yes, you can expect the output swing to increase if you increase the supply voltages. Note that the data sheet specifications are provided with +/-15V supplies. So, for example, if the specified maximum output is 12V, that means the output can only expect to get as close as 3V from the rail. If you drop your supplies to 12V, then you can only count on the output swinging to 9V. In your case, you're getting to 2.3V from the rail, which is expected, but not guaranteed (if you take a different AMP02, you will get a slightly different result). Note that the swings are provided assuming a given load. Changes in loading conditions will modify the maximum output swing (smaller resistance at the output means reduced swing).

    4b. For best resutls, it is a good design practice to stay away from the output swing limits provided in the data sheet. This will improve your circuit's linearity and provide some necessary margin to allow for supply variations and such.

    In addition, you may find the following resource useful:

    Please let me know if you have any further questions.



  • Many thanks, Gustavo, for a very detailed answer which is extremely helpful.

    I'm just returning to engineering after a very long break and have lost touch with a lot of technology and have just started trying to develop a range of products.

    The AMPO2 rather chose me, being cheaply available on eBay, as I wasn't looking for anything particularly accurate provided the drift was low.  However, having looked as various distrubutors prices (bearing in mind that I hope this will result in a batch production project) I can see that I can get equally good deals on superior inst amps without having to rely on an inconsistent supply source.

    I'll certainly take a look at the link you have provided.

    Many thanks once again.

    Kind regards,