I have a type of thick film resistors (3KOhm) which will output larger resistance under tension (2-10 times, say 6-30KOhm for my threshold force) and less resistance under compression (around 5-10%, say 2.5KOhm for the same force). They both have large drift with environmental temperature change, varying roughly from 1.5-5KOhm, but the force reaction ratio keeps almost the same. The two sensor resistances may have manufacturing tolerance, say 2.9 or 3.2KOhm. I want to use these properties to build a sensor to detect a tension type force/impact. I only need on/off signal output as certain force spike is detected (impact happening in millisecond).

I want to use two, one is set under tension and the other is under compression. I could put one sensor since I only detect tension force, but I think to use the second one for temperature compensation as well as redundancy (in case one fails). The mechanical design can be done, but how to detect the changes of two resistances and allow the ambient temperature change not effecting the detection.

At first I thought the electronic design is simple, but after some searching I am out of my mind. I thought to use a current limiting resistor (Rc) and the two sensors R1, R2 all in series connection to a 5VDC, and compare the changes of the two differential voltages, but I do not know if some IC can directly apply for this? For safety reason the adjacent capacitor, if any, is not acceptable if larger than 50pF. If this method is stupid, is there other simple circuit design can work this out?

I do not want other temperature compensation methods.

I appreciate any suggestions or comments.

 

Not knowing why you selected the thick film resistors, I am just going to suggest that there is an alternative with a mature support technology - http://www.omega.com/prodinfo/straingages.html

 

I have a type of thick film resistors (3KOhm) which will output larger resistance under tension (2-10 times, say 6-30KOhm for my threshold force) and less resistance under compression (around 5-10%, say 2.5KOhm for the same force). They both have large drift with environmental temperature change, varying roughly from 1.5-5KOhm, but the force reaction ratio keeps almost the same. The two sensor resistances may have manufacturing tolerance, say 2.9 or 3.2KOhm. I want to use these properties to build a sensor to detect a tension type force/impact. I only need on/off signal output as certain force spike is detected (impact happening in millisecond).

I want to use two, one is set under tension and the other is under compression. I could put one sensor since I only detect tension force, but I think to use the second one for temperature compensation as well as redundancy (in case one fails). The mechanical design can be done, but how to detect the changes of two resistances and allow the ambient temperature change not effecting the detection.

At first I thought the electronic design is simple, but after some searching I am out of my mind. I thought to use a current limiting resistor (Rc) and the two sensors R1, R2 all in series connection to a 5VDC, and compare the changes of the two differential voltages, but I do not know if some IC can directly apply for this? For safety reason the adjacent capacitor, if any, is not acceptable if larger than 50pF. If this method is stupid, is there other simple circuit design can work this out?

I do not want other temperature compensation methods.

I appreciate any suggestions or comments.

The bridge circuit to your rescue! Just be sure your reference resistors are of the same type as your unknowns.


Eric

 

Thank you both!

To Beenthere:

Due to my special application and harsh industry environment, there is no way to use strain sensor besides I donot need accurate measurement.

To: KL7AJ:

There is certain difference of my sensors than implementing a wheatstone bridge (e.g. for strain gage). I can only put two sensors inside the environment and can not use two extra such sensors as reference R. Also the temperature effect can not be absolutely cancelling out. Isn't a bridge output too sensitive for noise?