Guys, I have designed a signal conditioning circuit using Wheatstone Bridge + Instrumentation Amplifier + Low pass filter. But However the output I obtain was having a huge fluctuation of voltage waveform. The circuit schematic and output waveform using LabView was shown below.
In this circuit, i used MCP6002 Amplifier, it was recommended by someone earlier in this forum, I wonder this is the cause of the amplifier or there any mistake i done in my circuit diagram? The simulation result i get using Multisim was seems to be stable..

 

Why not use a real instrumentation amplifer? It is really hard to homebrew one out of three opamps and resistors because the resistors need to be matched to 0.1%.

 

I second Mike's suggestion. There are many IC instrumentation amplifiers that will work much better than your home-brew circuit and likely be cheaper.

 

Ok guys, so which instrumentation amplifier are suggested for my circuit here? because I'm really Less familiar with most the IC chips..

 

I'm not that familiar with all those out there but this unit seems to be popular and has reasonably good specs and price.

 

Shouldn't you be using bipolar supplies?

 

Shouldn't you be using bipolar supplies?

If all the voltages are properly biased above ground, it's not required.

 

If all the voltages are properly biased above ground, it's not required.

Maybe in another design, not this one.

 

Maybe in another design, not this one.

Okay, a cursory glance showed no voltages below ground. Where is the problem that you see?

 

The sensor and bridge are composed of variable resistances, which can change. The amp must be powered for all conditions if it is to work properly. I only point this out because if the OP doesn't know how to power an instrument amp, then using an integrated amp won't do him any good.

 

The common-mode range of the outputs from the bridge is constrained between 0V and 5V regardless of how badly the bridge is unbalanced. Why cant a rail-to-rail instrumentation amp handle that?

The bigger problem is setting the gain of the instrumentation amp, and biasing it's output relative to 0 and 5V...

 

Instrumentation amps don't work on common mode alone. For those who still have questions, I suggest they simulate the circuit under all conditions.

 

....................
The bigger problem is setting the gain of the instrumentation amp, ..

Why is that a problem?

 

Why is that a problem?

You have to put an upper bound on how badly the bridge can be unbalanced, i.e. constrain the resistance variation in the bridge arms...

Without constraint, the differential output can go from -5V to +5V. If powering the instrumentation amp on only 5V, why bother with an amplifier at all?

 

My sensor will only placed on the bridge arm of R9 (shown in the diagram). The one on R2 will be used a fixed resistor. I wanted to unbalance the bridge in order to measure the voltage output related to my sensor which varies in 10~20k offset due to vibration.

So Now should I still considering getting a in-amps for my case?

Appreciated for any helps =)

 

........................................
So Now should I still considering getting a in-amps for my case?

In my opinion, yes.

 

My sensor will only placed on the bridge arm of R9 (shown in the diagram). The one on R2 will be used a fixed resistor. I wanted to unbalance the bridge in order to measure the voltage output related to my sensor which varies in 10~20k offset due to vibration.

So Now should I still considering getting a in-amps for my case?

Appreciated for any helps =)

Look at the attached LTSpice simulation. All the voltages are shown as a function of the resistance of R9.

I assumed that you want the output of the instrumentation amp to go from 0V to 5V as the resistance of R9 goes from 10K to 20K. If this is a wrong assumption, say so...

Note that I tweaked R1 to the correct starting value to make this easier. Note that the instrumentation amp needs to have a differential gain of 33 with zero offset (the offset is taken care of by tweaking R1). In a real instrumentation amp, this is accomplished by grounding the Reference pin.

Oh, and throw away that toy simulator, and use a real one (LTSpice) Look how easy it is to get LTSpice to show how your bridge works...

 

Look at the attached LTSpice simulation. All the voltages are shown as a function of the resistance of R9.

I assumed that you want the output of the instrumentation amp to go from 0V to 5V as the resistance of R9 goes from 10K to 20K. If this is a wrong assumption, say so...

Note that I tweaked R1 to the correct starting value to make this easier. Note that the instrumentation amp needs to have a differential gain of 33 with zero offset (the offset is taken care of by tweaking R1). In a real instrumentation amp, this is accomplished by grounding the Reference pin.

Oh, and throw away that toy simulator, and use a real one (LTSpice) Look how easy it is to get LTSpice to show how your bridge works...

The resistance for my R9 value was actually 100k ~ 110k. But it's cool bro, I manage to understand what you did here haha. I wasn't really good with LTSpice, so I simulated with multisim.. So which type of in-amplifier I should use for my prototype in this case? Single-ended or rail-rail type? And May I know how did you add the E1 on your LTSpice Schematic? I cant find that component

Thanks in advance =)

 

Here it is again with 100K<R9<110K.

You should use a Rail-to-Rail Ina that will work well on Vcc=5V.

The LTSpice E component is a Voltage-Dependent Voltage Source. It is in the supplied library. You have to set the gain parameter before simulating. Look in the Help file for a description.

I am attaching the .asc file so you can run it yourself.

 

Here it is again with 100K<R9<110K.

You should use a Rail-to-Rail Ina that will work well on Vcc=5V.

The LTSpice E component is a Voltage-Dependent Voltage Source. It is in the supplied library. You have to set the gain parameter before simulating. Look in the Help file for a description.

I am attaching the .asc file so you can run it yourself.

Ok Thanks man Will try this out on my prototype first and see whether it works similarly to this simulation.

I will be using AD627 as the in-amp, I had attached the specification, hopefully this is the one you meant.