0
\$\begingroup\$

We're trying to buffer (for safety, ruggedness and scaling reasons) the input of a µC via some instrumentation-amplifier : INA826

We had our eyes on a schematic in TI INA826 datasheet TI app note The only problem is that we only have access to a asymetrical voltage supply (+5V) and also, we'd like to scale the op-amp output from 0-10V to 0-5V.

As I undestand it, we can just supply it with 0 / +5V and the output would be scaled to 0 / +2,3.

The one thing I don't understand though is that I thought the way to do it would be just to do a voltage divider followed by a unity op-amp circuit but that is not the case there...

\$\endgroup\$
6
  • \$\begingroup\$ just to do a voltage divider followed by a unity op-amp circuit Who says it has to be done like shown with the INA826 and that it cannot / should not be done with a voltage divider + unity gain buffer? I think it can be done that way as well and that would be simpler which I always prefer (never overcomplicate things). I see no reason to complicate things and use the INA826. \$\endgroup\$ Nov 23, 2020 at 9:17
  • \$\begingroup\$ Ok, I'll go for a voltage-divider + non-inverting buffer op-amp then. About your last sentence : no reason to complicate things and use the INA826 Do you mean that we should use the INA826 or we should not use it and go for maybe a simpler op-amp? \$\endgroup\$ Nov 23, 2020 at 12:03
  • \$\begingroup\$ I would choose to not use the INA826 as it complicates things. Using an opamp would be easier. I see no reason to use the INA826 but if there are good reasons please mention them. And indeed as Andy mentions, you will probably need a negative supply rail to drive the opamp's output all the way down to 0 V. \$\endgroup\$ Nov 23, 2020 at 12:32
  • \$\begingroup\$ I see, so best bet is to have a 0,5 to 4,5V voltage span (for example). The INA826 being an instrumentation-oriented op-amp (or in-amp as they call it) with good CMRR, is there any argument apart from cost in favor of swapping it for a LM358 or any other vanilla op-amp? \$\endgroup\$ Nov 23, 2020 at 15:29
  • \$\begingroup\$ with good CMRR Since (I think) your input signal is single ended (not differential) why would you need "good CMRR"? You're selecting on / mentioning criteria that do not matter. \$\endgroup\$ Nov 23, 2020 at 15:36

1 Answer 1

1
\$\begingroup\$

The one thing I don't understand though is that I thought the way to do it would be just to do a voltage divider followed by a unity op-amp circuit but that is not the case

The input voltage range on your INA826 schematic goes from -10 volts to +10 volts and so you need an amplifier that has a negative supply voltage as well as a positive supply voltage to handle the negative range.

The 2.5 volt reference voltage (REF3225) on the REF pin then centralizes a 0 volt input voltage to produce +2.5 volts on the outputs hence, the output maps an input of -10 volts to an output of +0.2 volts AND, an input of +10 volts to an output of +4.8 volts.

we'd like to scale the op-amp output from 0-10V to 0-5V.

As it current stands, the output range is +0.2 volts to +4.8 volts and would be acceptable for driving a 0 to 5 volt input range device.

The only problem is that we only have access to a asymetrical voltage supply (+5V)

Well, that's a problem if you need a negative voltage input range to map properly to the output.

\$\endgroup\$
5
  • \$\begingroup\$ Hi Andy, In our case, we only need a positive output range so the +5V power supply would do the job. Thanks for clarifying the 2,5V reference voltage, that's what I thought but wans't completely sure. One point stays blurry for me : why the 0,2V difference with the rail voltages? Is that an inherent property of the op-amps? What would be the name of that characteristic? Thanks, \$\endgroup\$ Nov 23, 2020 at 12:02
  • \$\begingroup\$ @VoltsAndNuts the INA826 can swing its output from the -Vs + 0.1 volts to the most +Vs - 0.15 volts so, if you are running from 5 volts and 0 volts then the output can only swing between +0.1 volts and +4.85 volts. So that's the best you can get from the INA826 on a single 5 volt rail but, the circuit design goes from +0.2 volts to +4.8 volts and that is a span that the INA is capable of with a comfort factor AND, also it has to feed some unknown circuit attached to the output AND, if it's an ADC then there's no guarantee it's input is 0 to 5 volts (despite what p1 of the DS might say). \$\endgroup\$
    – Andy aka
    Nov 23, 2020 at 12:11
  • \$\begingroup\$ Hi Andy, One thing I don't quite get is that the part states "rail-to-rail output" but the voltage swiing is +0.1 / -0.15V Are these two different things? \$\endgroup\$ Nov 24, 2020 at 13:04
  • \$\begingroup\$ Page 1 of a data sheet is reserved for the marketing hype. Most non R2R op-amps are lucky to get to within 2 volts of power rails so calling an op-amp R2R that gets to within 200 mV is par for the course in marketing hype @VoltsAndNuts \$\endgroup\$
    – Andy aka
    Nov 24, 2020 at 13:08
  • \$\begingroup\$ @VoltsAndNuts if you are happy with this answer, please consider formally accepting it. If you still have queries then leave a comment. You've asked 14 questions now and you should get into the habit of formal acceptance of relevant answers (one per question where applicable). This helps other users focus on the answer that worked best for you. It is important BTW. \$\endgroup\$
    – Andy aka
    Dec 7, 2020 at 17:08

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.