Amplify 1mV. Ampop?

Thread Starter

Ironic

Joined Dec 23, 2008
18
I´m trying to build a DC current meter, using a shunt.

Due to design constraints i need the shunt to be of a low value, resulting in outputs of in the order of millivolts. I need then to amplify this value by 1000, to reach values of volts, that are easier to read.

My idea is to use an op amp to do the amplifying, but the problem is that most op amps have offset voltages of millivolts, causing therefore huge errors in the reading.

Does anyone have a suggestion for an readily available op amp to use? My requirements are: single supply and very low offset voltage(0.1mV or less). Since it´s a dc application, bandwidth and slew rate can be ignored.

In case this is hopeless quest, is there a better way to linearly amplify a voltage in the millivolts range?

Thank You
 

DedeHai

Joined Jan 22, 2009
39
Texax instruments TLE2022: offset voltage 100uV

There are lots of low offset voltage opamps, just use farnell's search engine. maybe digikey.com also offers searching for low offset.
 

rjenkins

Joined Nov 6, 2005
1,013
Use an instrumentation amp like an INA126.

These have very low supply current, laser-trimmed offset and the gain is adjusted seperately from the inputs.

Bias one end of the shunt resistor to half the supply voltage (using two equal resistors).

Also connect the shunt terminals to the op-amp inputs via 1K resistors and put a parallel pair of back-to-back 1N4148 diodes across the input to clamp the input voltage at the opamp so it cannot be damaged by spikes.

You can set the circuit gain with a resistor across the dedicated gain set terminals, or use a multi-turn preset to give adjustable calibration.

The power supply for the op-amp side must be isolated from the high current circuit (eg battery), or you will need to go to a dual +/- supply as the voltage at the inputs must stay within the power supply range.

If one side of the shunt resistor is at the circuit 0V, look for a different instrumentation amp that includes 0V in it's common mode range - an LT1101 looks like a possible choice, though it's only designed for gains of 10 or 100.

If that is not possible, look at using a hall effect current sensor (electronic 'current transformer') which only needs the high current cable threading through the sensor.
 

Thread Starter

Ironic

Joined Dec 23, 2008
18
I can sacrifice some of the gain.

To explain circuit a bit better, this is for a pwm. My idea is from 12V, load, mosfet, shunt resistor, ground.

For the low offset i want, instrumentation opamp seems the way to go.
I´ve looked up datasheet forLT1101, but the Common-Mode Range seems to be around 0,5V, or am i understanding this badly? What i need is rail to rail input voltage. Is that the same as rail to rail Common-Mode Range?

Using two separate power supply's would complicate design a lot so i would prefer to avoid it.

The other option is to place the shunt above the mosfet. This to me creates a problem: since the mosfet is turning on or off, the shunt potencial will be jumping from 12v to a few hundredths of a volt.This doesn´t seem to be a good way to do an accurate measurement. (i am using a rc filter to average out the changing voltage across the shunt)

Any more opinions? Tks
 

beenthere

Joined Apr 20, 2004
15,819
This is commonly done - amplifying signals by factors of 1000 or above without problems. Is this offset a real problem, or theoretical? What kind of absolute precision in the result must you have?

Can you post up the circuit so we can see what you are doing?
 

Thread Starter

Ironic

Joined Dec 23, 2008
18
If i have one terminal of the shunt at 0V and the other one at 1mv, if i use an opamp with an 1mv offset error, i will get and 100% error...


12v ----- LOAD----- mosfet-----shunt----- Ground
 

SgtWookie

Joined Jul 17, 2007
22,230
Last edited:

Thread Starter

Ironic

Joined Dec 23, 2008
18
Tks for your help guys.

Your suggestions pointed me in the right direction, and i have found what i was looking for.
 
Top