Our full technical support staff does not monitor this forum. If you need assistance from a member of our staff, please submit your question from the Ask a Question page.


Log in or register to post/reply in the forum.

Strain gages and multiplexers


draffard May 8, 2012 03:45 AM

We want to use a CR1000 with four (4) AM16/32B Multiplexer to measure 64 strain gages -120 or 350 Ohms) bonded on concrete.

16 strain gages would be connected on each multiplexer.

We can connect the strain gages in half bridge or in quarter bridge.

Is it possible to connect four (4) 4WFB120 (or 4WFB350) modules on the CR1000 and connect each multiplexer on one module?


Thanks per advance


aps May 8, 2012 04:17 AM

You could use the bridge modules like this, you may live to regret it. The reason is that with full bridge modules you will end up with multiplexer contacts in series with the strain gauges.

All mechanical relays have a small contact resistance (100 mohm for the AM16/32) which creates a small offset (not too significant) but more critically increases with time as the relays age, plus can vary each time the relay actuates. This will add some apparent noise and drift to the measurements which is undesirable.


draffard May 8, 2012 04:43 AM

For information, the monitoring duration is about 3 months (1 measure per hour).

Do you think It would be better to connect one 4WFB120 (or 4WFB350) module for each strain gage on the multiplexer?

Thanks


draffard May 8, 2012 04:44 AM

For information, the monitoring duration is about 3 months (1 measure per hour).

Do you think It would be better to connect one 4WFB120 (or 4WFB350) module for each strain gage on the multiplexer?

Thanks


aps May 8, 2012 05:07 AM

Yes one module per gauge is the best solution.


TweedleDee May 8, 2012 09:34 AM

The following is for 1/4 Bridge strain:

Placing relays internal a Wheatstone bridge strain system is discouraged. Any change in resistance of the relay contact would have a corresponding change in the bridge output voltage.

Changes in contact resistance can be induced by temperature fluctuations, oxidation, environmental conditions, and normal wear of contact surfaces. The specification for the relays that CSI uses state that initial contact resistance will be less than 50 mOhms (AM416) or less than 100 mOhms (AM16/32 series). There is not a specification for change in contact resistance for the relays because there are so many variables that affect contact resistance. I do have test reports for various test conditions that show contact resistance changing over time by 10 to 20 mOhms. These tests were performed using static test temperatures, so I would assume real world conditions would result in larger resistance shifts.

When strain gauges are used in the Wheatstone bridge, small changes in contact resistance result in large apparent strains. If two relays vary by 10 mOhms, the resultant apparent strain would be about 80 ustrain when using 120 ohm strain gauges in a one active element system."

For 1/2 Bridge strain (2 active gages per wheatstone bridge), the output is double of the 1/4 bridge strain, so the error would be reduced by a factor of two. An alternative solution for 1/2 bridge strain would be to use external excitation, so the power for the wheatstone bridge would not be multiplexed, only the output from the top side of the wheatstone bridge (in this setup the multiplexer's relay contact resistance does not affect the measurement). In this way, you could measure up to 72 half bridge strain circuits with one AM16/32B multiplexer and four TIMs. The external excitation would need to be able to source 5 VDC with enough current drive for all of the circuits. You would need to measure the excitation voltage each measurement cycle.

* Last updated by: TweedleDee on 5/8/2012 @ 9:39 AM *


Romans Jun 27, 2020 06:02 AM

This post is under review.

Log in or register to post/reply in the forum.