Amplify small voltage signal

Hello

I have one troble. I am trying to amplify small voltage signal from a strain gauge that is approx -0,50 mV to 50 mV. I need to amplify to 0-5V so I can sample on the arduino bord.

I have connect 2 op-amp togeter (LM324N) and I don´t get the right result. What do u recommend? what ap-amp shuld I use and how to connect it?

Well without knowing how you've put your opamp circuit together how can we guess what's wrong.

Since you have a strain guage you will clearly do best with an instrumentation-amplifier setup - some are available on a single chip, there's a standard way to fabricate a high-performance instrumentation-amplifier from 3 op-amps too. A less performing solution with a single op-amp is the standard differential amplifier circuit. All of these you'll find on wikipedia these days, BTW.

What accuracy are you looking for BTW, that affects circuit design...

The LM324 will work from 5V single supply but you must keep all the signals in the 0 to 3.5V range, and best to keep them in 1.5 to 3.5V range.

Dewalt: I have connect 2 op-amp togeter (LM324N) and I don´t get the right result. What do u recommend? what ap-amp shuld I use and how to connect it?

post your circuit diagram.

off hand, i'd say even 1 op-amp is more than adequate... you need to however do some work to design the bias and gain within swing range... say bias your signal to 1.7V as the midpoint.... and put a gain of 60 odd... and feed 3.5v to the analog ref in arduino to get max range.

working with 2 opamps has its benefits but needs far more careful design work

srinathdevelopment:

Dewalt: I have connect 2 op-amp togeter (LM324N) and I don´t get the right result. What do u recommend? what ap-amp shuld I use and how to connect it?

post your circuit diagram.

off hand, i'd say even 1 op-amp is more than adequate... you need to however do some work to design the bias and gain within swing range... say bias your signal to 1.7V as the midpoint.... and put a gain of 60 odd... and feed 3.5v to the analog ref in arduino to get max range.

working with 2 opamps has its benefits but needs far more careful design work

Without knowing the details of the strain guage its premature to guess how to bias the bridge... We also need to know the accuracy requirement before designing in detail. The CMRR figure might be important for instance. The LM324N may prove inadequate for the accuracy desired too, it has a CMRR of 65dB worst case, offset of 5mV, and tempco of offset of 30uV/C

agree with you for most part.... a lot would depend on requirements and device specifics... comments more in the nature of what to think about to get started

on biasing.... etc, my comments were based on OPs specification that the gizmo is putting out +/-50millis volts signal... actually I've made an error in the gain suggestion :). LM324 swings GND to Vcc - 1.5V... so you'd want +50mV to be amplified close to 3.5V, -50mV close to GND and 0V to about 1.7V biased to bemid in the opamp working range... ie. voltage gain around 30 to 35 (not 60 as mentioned previous post)

I spent some time looking how to condition arbitrary voltages say from 1.5v to 2.5v to convert it to 0-5V for Atmega inputs. Is not the case for the +/- differential strain gage but the scaling part will works for you.

One of my last circuits was to scale a +/-5V to 0-5V but I repeat, is not differential, just bipolar to 0-5V.

To convert differential to single ended you must use a instrumentation amplifier circuit.

Could you post your schematic?

srinathdevelopment: agree with you for most part.... a lot would depend on requirements and device specifics... comments more in the nature of what to think about to get started

on biasing.... etc, my comments were based on OPs specification that the gizmo is putting out +/-50millis volts signal... actually I've made an error in the gain suggestion :). LM324 swings GND to Vcc - 1.5V... so you'd want +50mV to be amplified close to 3.5V, -50mV close to GND and 0V to about 1.7V biased to bemid in the opamp working range... ie. voltage gain around 30 to 35 (not 60 as mentioned previous post)

The range is -0.5mV to +50.0mV, not -50.0mV to +50.0mV

The AD620 looks good for your project.

http://users.ece.utexas.edu/~valvano/Datasheets/AD620.pdf

I will test in Proteus along with my voltage scaling circuit and let you know.

THe LMC662 is an OLD CMOS replacement for the LM324,358 series that will swing rail to rail, works well from a 5V and the inputs work down to ground on a single ended supply. It’s inexpensive, easy to use and draws about 10% of the supply current that a '324 or '358, further when I was using them in quantity (2008) everyone was sourcing them and at about $0.50 in 1K quantities (as I remember, could have been 10K) point is that they are easy to use and should certainly fill the bill. The only consideration would be input bias current, it doubles about every 10deg. C most of those breed did, Including the AD620 as I remember, The input transistors were J-Fet’s in a lot of the high input impedance op-amps. The AD620 is a full IA and used for many different applications. Extremely difficult to use properly as although the gain figures are high the components required to achieve accurate gain are costly and require special treatment AND a Bi-Polar +/- 12V PSU too. The output is not rail to rail and will swing to a volt or so of either rail… Cool 80’s era IA device but severely dated and still likely expensive as well as more education than you are likely to need… Yet… IMO.

Doc

Want Rail to Rail?

The OPAx342 I have used works with 5V single supply and is 0V to rail (5V) operation.

OPA2342 = 2 opamps, OPA4342 = 4 opamps.

the OPA335 can work with input voltages from 100mV below negative supply (V-)...

here is an in-amp circuit: http://de.wikipedia.org/wiki/Datei:Instrumentation_Amplifier_2OpAmp.svg (the ground symbol is the reference voltage, which should be about 200mV at a gain of 50... u would translate the range [-500uV..+50mV] to [+175mV..+2700mV]) (the resistors might be like these: R1=R2=100kR and Rgain=4.17kR: http://www.wolframalpha.com/input/?i=solve+for+g%3A+50*j%3D%281%2B100%2F100%2B2*100%2Fg%29*j )

No what I really think is that he want's something that is available, inexpensive and easy to use... He didn't say anything about rail to rail, I did, I have a great deal of doubt as to whether he at this time in his education even is aware of what that means, as to the inputs going to ground, that is really an issue with early op-amps like the '12 -324 and the '358 series and again I mentioned it not him. A simple low-drift op-amp nothing real special would get him working well enough to evaluate the future directions he might need to go and at least make some measurements. Just slamming into an Instrument Amplifier without at least some prior experience with both OP and Instrument amplifiers would surely be a recipe for a small disaster and certainly not a guide to certain success. In my somewhat experienced opinion.

Doc

I suggest AD620 amp chip for instrumentation amplification.

As you are probably aware, use this in the wheatstone bridge configuration.

Regards,