Strain Gauges

[ljbeng]

The mechanical engineers I work with want to put some strain gauges on a trailer for testing. I want to read the values of 8 strain gauges into a 12 bit A/D and transmit into a pc for analysis. The A/D part and the computer part are no problem.<p>Where I need help is in the strain gauge amplifier. I would like the voltage from the strain gauge under no stress to be 2.5v. Then if the sg is in compression, I want the voltage to go below 2.5 and tension, above 2.5v. I have experimented with a simple opamp with gain but I don't get very much voltage change for the deflection. We test with a simple 1/4" steel bar that I try to bend in my hand. Temperature is not an issue so I do not want temperature compensation.

[Will]

LJBENG,
If you go back to the posts started April 9th by Ian Robinson under the tile `Multiplex Signal Conditioner' that may be of some help to you. When you talk about getting Volts and mV out of strain gauge circuits then it sounds like you are looking at the output of a strain gauge signal conditoner because starin gauges are (To the best of my knowledge) purely resistive devices. Your post should produce some good replies..

[Chris Foley]

Here it is...<p>Multiplex Signal Conditioner<p>[ May 14, 2003: Message edited by: Chris Foley ]</p>

[natcsparky]

Did you ever get the help you needed. Email if so.<p>Rick Capps
[email protected]

[7474back2backcrashes]

If you have tried op amps to boost the signal your problem may be that you used cheap ones.
Also use precision resistors along with them so your voltage divider outputs arn't lazy.
Get some modern good quality op amps <ie' not silicone base > ones and they will be more sensitive. btw - When using strain guages you always hafta be careful of the guages temp if you want any kind of sensitivity at all.
Another factor is the way the guages are attached.
Your bonding method has to be very consistent or you can give up the movement with to much glue or uneven bonding and will result in no consistincy between different sets.
Testing them to a calibrated and consistent standard is about the only way you can get anything meaningful out of any two of them.<p>Another more exact method is using a Renishaw laser or ballbar. You can detect movement down to 12 places behind the decimal point if you want to.
But for that kind of accuracy you MUST comp for the temp.<p>One more way to detect strain is by using muscle wires. They require a bias pull to get them started and balanced out but after they warm up temps don't effect them much. Muscle wires have a somewhat limited inline use and are easy to bump and damage but will span longer distances.
The amount of movement you expect to see may rule out one method or another.
Good Luck.
Chuck

[cato]

Are you using a wheatstone bridge in combination with your strain gauge? This will help remove the dc component of the signal and leave you with the difference between the nominal signal and the measurement signal. Then just throw a lot of gain at it.<p>Additionally, or alternatively uses a differential amplifier or insturmentation amplifier.<p>[ September 26, 2003: Message edited by: cato ]</p>

[upsmaster]

some transducer manufactures can give u different outputs ex 0-30 mv 0-5 volts 0-10 volts and current loop 0-20 ma. their are several manuf of amp modules that allow u to get what ever output swing in a window of hi lo voltages. one trick i know of is that most transducer outputs are not always zero when the bridge is at null... adding a 10 meg 10 turn pot across one of the legs of the bridge lets u tweak the output to zero from the transducer before sending it to a conditioning amp.if u have more questions i may be able to help.. it would help to know what type of measurment u are looking at and what type of transducer u want to use i can give u names of manufactures if u want to take this road