Voltage drifting

[E2005]

I have a strain gauge connected to a metallic rod hinged at one end. The system detects the bending of the beam(when I press it with my hand, as seen by increase in output voltage) but even before touching the rod, as I bring my hand close to the rod the output starts increasing (offsetting from zero point) and decreases to normal when I move away from it. The change in output voltage due to this offset is greater than that caused by bending. It seems almost as if the grounding of the system changes. I tried insulating and shielding the circuit, wires and beam(with aluminum foils, tapes etc) but still the systems offsets as soon as my hand or body comes close to the rod. I am trying to understand the reason for it and if there’s a way to solve this problem.

 

[IRstuff]

Sounds like you're AC coupling to the circuit. You need to ground YOU. There are lots of non-linear components that will rectify coupled AC signals and create an offset in your system.

TTFN

 

[itsmoked]

Try grounding the rod.

 

[E2005]

IRstuff: my system is running through a DC battery voltage(no AC supply), the voltage supplied to the strain gauge is DC(5V), and final output voltage is also DC
istsmoked: I tried grounding the rod , that didnt help

 

[BrianG]

Some questions spring to mind:
Presumably you are measuring the drift at the output of the amplifier? Have you looked at the output on a 'scope?

As well as the expected amplified d.c. from the strain gauge, it could be unstable and oscillating at a high frequency. If this is the case you can have a problem with "hand capacitance" tuning the oscillation and producing a variable output as you get closer or further away.

 

[OperaHouse]

Is this a commercial readout or something you built. Since it is a DC bridge(?), you are free to place 0.01 ceramic caps to all the four inputs to a common instrument ground to reduce RF pickup. This will also make any capacitance from you insignifigant or totally drive the amp crazy.

 

[buzzp]

I am with IR stuff. Your capacitively coupling the junk (maybe from fluorescent lights) to the rod. Some solutions might be to ground yourself and/or perhaps use some better filtering on the output of the strain gauge (filter the AC).

 

[IRstuff]

E2005,

Just because your system is running DC does not mean that there is no AC, nor does it mean that there are no common mode signals running around. In some respects, a battery-operated device that is not grounded can be more susceptible to stray signals, precisely because it's floating.

Besides grounding, you might also check to make sure that your gauge wires are identical in length, dressing, etc. They should also be twisted/shielded, with a minimum of 4 twists per inch. Coax will not work correctly in this type of application, since the outer shield is acting as a signal line and behaves differently from the internal signal wire.

TTFN

 

[MrBananas]

Differential Amplifiers are your friends...

If you use coax cable you can use both the signal and shield as inputs to a differential amplifier. The "noise," whether coupled in AC, RF, etc. will then likely show up as "common mode" (present on both inputs). The differential amplifier will give you the difference between the shield and signal cables, which essentially cancels the noise. Depending on the op-amp/instrument amp and component tolerances used, this technique can reduce common mode noise by 40-80 dB.

 

[cbarn24050]

Hi, using DC to pwer the bridge is asking for trouble, especially if the amplifier is some distance away. Mount your amplifier next to the strain guage or use an AC excitation system.

 

[Buttersrq]

Hello, if the strain gauge is drifting a few things could have happened first the bridge may have become un balance between the excitation and signal lines, second if the power supply that feeds the strain gauge has become unbalanced if its a 5(V) system it needs a good ground check to see what the ohm reading on the signal lines on the strain gauge are. It will ohm out to 350 ohms or 700 ohms or 1000 ohms. Then ohm between the positive excitation and the positive signal then check the negative side this will be balanced also. This will tell you if the strain gauge has failed or not.

 

[felixc]

Your strain gauge readout system has become a Theremin.

How frequently should you read the bridge? Make a low pass filter to give a chance.

If I place my DMM in the DC millivolt range, leads floating, it will pickup any body movement around it.