PDA

View Full Version : Question Op Amp Question


Hell-e-Guy
Apr 30, 2009, 10:38 AM
Hi,

I need to measure a differential voltage +/- 25V (floating) using an instrumentation amplifier and than feed it to an MCU's ADC.
I need to keep the input impedance as high as possible (hopefully above 10 Mega Ohms) and a voltage divider on the Instrumentation Amp inputs will degrade the input impedance.

Is it possible to find an Op-Amp that can accept voltages on its' inputs that are well above it's operational voltage (single supply op amp)? I looked for Beyond the rails inputs, but the ones I found go only 1/2 a volt above and below the rails.

Is using a high voltage op-amp a good solution?

I also need to attenuate the voltage from +/- 25V to 0-5V. In what way can can I attenuate the voltage without degrading the input impedance?

Thanks,
Hell-e-Guy

bridlewood
Apr 30, 2009, 11:29 AM
I do not know of any op-amp that can work that far beyond the rails. Of course if you place a 10 M ohm resistor between your 25v signal and a 5v op-amp you could measure the small current that is proportionate to the voltage difference. Do you have a high voltage power supply on the instrumentation amplifier that could be used to power a high voltage op-amp and divide the voltage down to 5v ? As for voltage dividing, there are amps with a variable gain control that may prevent decreasing your input impedance.

jeffs555
Apr 30, 2009, 08:31 PM
Something like this set up with a gain of 10.1 and then a 10meg/100k divider on each input(or gain of 1.1 and 10meg/1meg?) should give an output of +/- 2.5 volts. If you used a quad 5v op amp you could use the 4th op amp to create signal ground at 1/2 of the 5v supply which would give you the 0-5v output(actually VCC/2 +/- 2.5v).
http://www.allaboutcircuits.com/vol_3/chpt_8/10.html

tune by tito
Apr 30, 2009, 09:03 PM
Hi,

I need to measure a differential voltage +/- 25V (floating) using an instrumentation amplifier and than feed it to an MCU's ADC.
I need to keep the input impedance as high as possible (hopefully above 10 Mega Ohms) and a voltage divider on the Instrumentation Amp inputs will degrade the input impedance.

Is it possible to find an Op-Amp that can accept voltages on its' inputs that are well above it's operational voltage (single supply op amp)? I looked for Beyond the rails inputs, but the ones I found go only 1/2 a volt above and below the rails.

Is using a high voltage op-amp a good solution?

I also need to attenuate the voltage from +/- 25V to 0-5V. In what way can can I attenuate the voltage without degrading the input impedance?

Thanks,
Hell-e-Guy

Try Analog Devices
http://www.analog.com/en/amplifiers-and-comparators/current-sense-amplifiers/products/index.html

http://www.analog.com/en/amplifiers-and-comparators/current-sense-amplifiers/ad626/products/product.html

http://www.analog.com/static/imported-files/data_sheets/AD626.pdf

AD626 FEATURES
Pin Selectable Gains of 10 and 100
True Single-Supply Operation
Single-Supply Range of +2.4 V to +10 V
Dual-Supply Range of 1.2 V to 6 V
Wide Output Voltage Range of 30 mV to 4.7 V
Optional Low-Pass Filtering
Excellent DC Performance
Low Input Offset Voltage: 500 V Max
Large Common-Mode Range: 0 V to +54 V
Low Power: 1.2 mW (VS = +5 V)
Good CMR of 90 dB Typ
AC Performance
Fast Settling Time: 24 s (0.01%)
Includes Input Protection
Series Resistive Inputs (RIN = 200 k)
RFI Filters Included
Allows 50 V Continuous Overload

jeffs555
Apr 30, 2009, 09:32 PM
Main problem with that AD626 is that it has an input resistance of only 200k and Hell-e-Guy needs 10meg.

PS The AD626 gets its high common mode range by using an internal 6/1 resistor divider.

Hell-e-Guy
May 01, 2009, 06:19 AM
Thanks for all the replies.

I think the solution I will go with is to sample the voltage with a high voltage op amp with a 50V supply made with a boost+charge pump circuit. Reference the op amp with 25 volts and scale down to 0-5 volt.

This brings me to my next question... can an instrumentation amplifier serve as an attenuator (gain < 1) or is there a chance it will ocillate?

I remmember that op-amp generaly don't like gains less than unity (in the feedback path).

Hell-e-Guy

jeffs555
May 01, 2009, 02:31 PM
Most instrumentation amps are not able to set a gain less than 1. Simplest thing is just a resistive voltage divider on the output of the instrumentation amp.

Hell-e-Guy
May 01, 2009, 03:05 PM
Thanks jeffs,

I will implement the voltage scaling at the output with a buffered voltage divider.