Amplifying strain gauge output voltege using 741

First of all, thank you all for this forum - I really appreciate the help.

So, I've designed a very simple circuit using two strain gauges in Wheatstone bridges to turn an alarm on and off. Basically, when a load is applied to the bridges, the alarm turns on, and when the load is removed, the alarm turns off. A POT is added at the 741 op-amp to allow crude calibration for different desired weights.

My basic problem is with the op-amp. Since strain gauges produce very small voltage outputs, I obviously need to amplify the signal in order to trigger the attached transistor. Since I want to make this very sensitive, I went with an exponential op-amp setup using a loopback with a diode. However, I can't get anything out of this circuit. When I apply a load to the strain gauges, nothing happens to the output voltage.

My brother in law suggested that I need two op-amps. Since I am using two strain gauges, I need to use one op-amp as an adder and one as an amplifier. I don't see why since the op-amp isn't really adding - tying the strain gauge output together simply combines the input to the op-amp. My thoughts were that I should be tying one of the op-amp inputs to the negative of my power source (a simple wall-wart 9V inverter) instead of to the negative of the strain gauge output. Any ideas?

Lastly, I'm a ME, not an EE, and I never got very far in electronics. Is any type of diode okay? I just bought a pack at Radio Shack - I didn't know if there is really a difference between them.

Sorry the diagram is rough - the program doesn't allow rotation at 45 degrees.

 

1) the negative input of the op-amp is connected to positive 9 volts. The + input of the op-amp has to be more positive than the - input to get the amp to put out any positive voltage. Since there is no voltage more positive than +9 volts, the amp can never put out a positive voltage.

2) The diode is not in the feedback loop. It won't logarythmic anything, and it will make the signal level of the strain gauges be less positive by about six tenths of a volt. I don't think the strain gauges make enough voltage difference to overcome this, and you don't need the diode at all.

3) The beeper does not have a power supply voltage applied to it. No matter what the amp does, it can't turn on a beeper with no power.

Take out these 3 mistakes and try again.

 

Your schematic is confusing. And I think you are on the wrong track. But your project is indeed doable. In order to read a strain gauge you need a Differential Amplifier. Here is some info to start with. http://focus.ti.com/lit/an/sloa034/sloa034.pdf
Also the 741 type opamp is a relic. I would avoid using it.

 

In addition to the problems noted above, you might give us the reason for the strain gauges. What force is bring sensed, and why must it be done that way?

For a bit of a primer on strain gauges, try Omega - http://www.omega.com/prodinfo/StrainGages.html

 

Yes, my schematic is indeed poorly drawn. Half of that is the limitations of the software I used (random software I found that is in Ubuntu repositories) while the other half is my personal lack of experience. I apologize for the bad diagram. I was mainly trying to highlight the amplification setup - the alarm is a separate circuit altogether, so it's not complete. The biggest mess is the Wheatstone bridge setup since the program can't draw angled resistors.

Basically, let me rephrase my question. I want to use a non-inverting, summing, exponential amplifier to amplify the output of my bridge setup - non-inverting so I don't get a negative output (need positive 5V to drive a NPN transistor); summing to take input from each strain gauge bridge; exponential to make the output very sensitive. I used what I found at Wikipedia (don't give me grief, please):

http://en.wikipedia.org/wiki/Operational_amplifier_applications#Exponential_output
(Summing and non-inverting found above on the same page)

Does that clear up questions? Thanks for your help, all. If possible, I really want to use simple parts that I can get at Radio Shack (hence 741 use).

 

Use "lm741" as a search term in Google to find a data sheet. Notice that the 40 year old design is characterized for operation with +15 and -15 volts applied. Using an obsolete design like that will make your circuit much more difficult to realize.

Radio Shack is generally in bad odor. Their limited range of parts are overpriced and not at all up to date. Ordering from Digi-Key or Mouser Electronics at least gives several orders of magnitude more choices.

By all means, you can try to use a 741 with dual supply rails, but read and understand the data sheet - and the link to Omega Engineering.

 

Probably overkill for your project, but usually when you're monitoring a Wheatstone bridge with strain gauges, you'd use an instrumentation amplifier.

Here's a link to a good Application Note on instrumentation amplifiers:
http://www.intersilsemi.com/data/an/an1298.pdf

The use of an instrumentation amplifier solves a number of difficulties you may run into, like matching your external resistors precisely to obtain a high CMRR (common mode rejection ratio). There are a number of example circuits in the Application Note, including a Wheatstone bridge.

You might look at inst. amps like INA126/INA2126, INA128/INA129, INA118, etc.