Automating the monitoring of Wheatstone Bridge Output

[qu1nn]

Purpose:
To automate the data collection of pre-wired full Wheatstone bridge
(bonded foil strain gage) load cells.
With an excitation of 10Vdc.

Strain Gage Resistance: 4500 Ohm or 1000 Ohm
Desired resolution: to be able to accurately be able to detect a
change of voltage of 1 uV

I would like to be able to put multiple load cells (Qty 25-50) inside
an environmental chamber and monitor the output of each load cells
through ambient --> hot (125F) and back to ambient.

I was thinking of using an Agilent 34420A.
Software: LabView
With a precision 10V power supply. It is desired to keep excitation
on at all times to eliminate any possible self heating issues.
I would keep the precision power & meter outside of the environmental
chamber & keep cable length as short as possible ( ~5ft)

Somehow the multiple cells would be switched with a multiplexing unit
(haven't researched yet)


Does anyone see any types of flaws


Thank you for your comments

Qu1nn

 

[Robert Baer]

Purpose:
To automate the data collection of pre-wired full Wheatstone bridge
(bonded foil strain gage) load cells.
With an excitation of 10Vdc.

Strain Gage Resistance: 4500 Ohm or 1000 Ohm
Desired resolution: to be able to accurately be able to detect a
change of voltage of 1 uV

I would like to be able to put multiple load cells (Qty 25-50) inside
an environmental chamber and monitor the output of each load cells
through ambient --> hot (125F) and back to ambient.

I was thinking of using an Agilent 34420A.
Software: LabView
With a precision 10V power supply. It is desired to keep excitation
on at all times to eliminate any possible self heating issues.
I would keep the precision power & meter outside of the environmental
chamber & keep cable length as short as possible ( ~5ft)

Somehow the multiple cells would be switched with a multiplexing unit
(haven't researched yet)

Does anyone see any types of flaws

Thank you for your comments

Qu1nn

Your 1uV spec is not too feasable; thermoelectric effects are large
compared to that, and RFi can also fiddle readings as well.
If you could arrange a method to modulate the bridge imbalance (ie:
alter a bridge arm in a regular manner), then the output will be
modulated, and a synchronous detector would give a chance of getting
better than normal resolution (but still will not take care of the
thermoelectric problem).

 

[GPG]

Your 1uV spec is not too feasable; thermoelectric effects are large
compared to that, and RFi can also fiddle readings as well.
If you could arrange a method to modulate the bridge imbalance (ie:
alter a bridge arm in a regular manner), then the output will be
modulated, and a synchronous detector would give a chance of getting
better than normal resolution (but still will not take care of the
thermoelectric problem).

Could use AC to excite bridge, amp and filter to recover signal,
or switch the bridge to produce an AC signal:





+10V
|
o------o-----o
| |
.-. .-. _/ sw1
| | | | o------o/ o-------o
| | | | | |
'-' '-' | |
| | | _/ sw2 o--------o
o-------------------o------o/ o----o |
| | | |
| | _/ sw3| | OUT
| o------o------o/ o----|--o
| | | |
.-. .-. | o------------o
| | | | | _/ sw4|
| | | | o------o/ o----o
'-' '-'
| |
o------o-----o
|
|
GND
sw 2,3 alternated with sw1,4

 

[Robert Baer]

Could use AC to excite bridge, amp and filter to recover signal,
or switch the bridge to produce an AC signal:

+10V
|
o------o-----o
| |
.-. .-. _/ sw1
| | | | o------o/ o-------o
| | | | | |
'-' '-' | |
| | | _/ sw2 o--------o
o-------------------o------o/ o----o |
| | | |
| | _/ sw3| | OUT
| o------o------o/ o----|--o
| | | |
.-. .-. | o------------o
| | | | | _/ sw4|
| | | | o------o/ o----o
'-' '-'
| |
o------o-----o
|
|
GND
sw 2,3 alternated with sw1,4

That is what i said....
But thermocouple effects will still give problems.
And it is very doubtful that 1uV resolution or accuract is needed.