I am working on a project that simply has two sensors(resistance change) in series. I want to use a wheatstone bridge to measure the change in resistance because the wheat stone "linearizes" the readings. But if I have two sensors in series, and I put a wheat stone bridge on each sensor, I would have 4 voltage references since each sensor requires a wheat stone bridge.
The problem is that my project has two sensors per finger and one for the thumb, but if I put a wheat stone bridge on every sensor....so... each finger except the thumb has two sensors making 4 voltage references. 4 voltage references times for = 18 voltage references + the two from the thumb. that is 20 voltage references!
Is there any sort of modified wheat stone bridge where I can use one wheat stone bridge to measure the resistance change of two sensors "Independently?" Or any way of decreasing the amount of voltage references points but keeping the readings still "linearized?"
I wouldn't have enough analog pins to read in the voltage change to calculate the resistance change.
I found the Kelvin double bridge, but it only has two voltage references and it looks like the two sensors are DEPENDENT on each other. I need it so they are INDEPENDENT of each other.
because the wheat stone "linearizes" the readings.
What do you mean by this?
The analysis of a Wheatstone bridge is simple ONLY in the case that the bridge is balanced. If it is not balanced, the resulting equations are not linear in the unknown resistance. The latter situation is almost always the case. See for example Wheatstone bridge - Wikipedia
jremington:
What do you mean by this?
The analysis of a Wheatstone bridge is simple ONLY in the case that the bridge is balanced. If it is not balanced, the resulting equations are not linear in the unknown resistance. The latter situation is almost always the case. See for example Wheatstone bridge - Wikipedia
Meaning that I will get a linear curve when reading instead of just measuring from a current source or voltage supply.
So is there a way I can fix the Kelvin double bridge so that both sensors are "independent" of each other? Looking at that picture i posted, there are only two reference points meaning the sensors are dependent on each other. How wold I modify this so it only has 3 reference points and each sensor is dependent of each other??
KenF:
This is an xy problem. Tell us what you're trying to measure and what sensors you have.
Ok, I basically have two sensors in series as shown in the picture(Ra and Rx). the sensors are just strain gauges that is being made in the nano department at my university. It is just a change in resistance.
Now I want to use the wheatstone bridge because that is the best method for measuring resistance sensors and it linearizes your graph instead of having it sloped. I can put a wheat stone bridge on both of the Ra and Rx, but this will leave me with 4 voltage references since you measure the resistance by taking the voltage difference between the two voltage references on the wheat stone bridge. If I have two wheat stone bridges, I am left with 4 voltage references.
Now this is going to go on every finger except the thumb. So 4 fingers x 4 voltage refrences....that is 16 analog pins I would need. The thumb will have two voltage refrences. So this is a total of 18 analog read ins I would need.
The problem is that I do not have enough analog pins on the microcontroller to do this.
Question: Is there a modified wheatstone bridge that can measure two resistances INDEPENDENTLY of each other I can use that I would only need 3 voltage references instead of 4 voltage references?
The picture I posted has two resistance sensors, but the wheat stone bridge only has two voltage references, so I am assuming that means they are DEPENDENT on each other. I need to get the resistance of each sensor individually.
You're still doing it! What are you attempting to do with these strain guages?
Do you want the total force on them both?
The average force between them?
The difference of the force between them?
You're jumping to the conclusion that some kind of bridge is going to help with your problem, but if we knew the problem, we'd probably be able to give a straight forward solution. Chances are it won't involve any bridge ( Wheatstone, Kelvin or otherwise )
KenF:
You're still doing it! What are you attempting to do with these strain guages?
Do you want the total force on them both?
The average force between them?
The difference of the force between them?
You're jumping to the conclusion that some kind of bridge is going to help with your problem, but if we knew the problem, we'd probably be able to give a straight forward solution. Chances are it won't involve any bridge ( Wheatstone, Kelvin or otherwise )
Wheatstone Bridge
They are simply polytype strips placed on your fingers. When you bend them, or stretch them as you would do in sign language, the resistance changes on these strips. It is a device to translate hand signals into text on a PC. A sign language glove.
So I have to measure resistance because the is how the strips work. Stretching them makes a change in resistance.
second, the answer is no. you get one sensor per bridge.
you could use a second, reverse of the first, one on top of a finger that bends forward, and one on the bottom of the same finger that bends backwards. those two inputs of the EXACT SAME movement can be input to a Wheatstone bridge to increase sensitivity.
but, you also should know that temperate will effect the bridge. you might need to put in a secondary sensor that is not bent, but only reports it's value as a change in temperature only.
then, you have the amp. if you need to add gain, or shift the slope, you will need to use an industrial op-amp,
In the early days of my PhD research (before I switched horses to a different project) I also looked at developing sign language gloves.
If I've read you correctly, you're trying to determine the angles of the individual joints on the fingers and thumbs of a hand. The fact that you say you need fewer sensors for the thumb than for a finger implies to me that you may have a redundant measurement there anyway. If you actually look at the way the top two joints of your finger move in normal use, you will see that the these joints never move independently. Ok you can force them to do so, but in normal use it doesn't happen. The angle on the topmost joint can almost always be determined from the angle on the next joint down.
I know it doesn't answer your original question but, if I interpreted correctly, you may need far fewer sensors than you think anyway.
you are misinterpreting the non-linearity of a strain gauge as the linearity of a simple resistor.
Scarlet in the video never discusses processes being measured.
the beauty of the linearity of the output / input relationship is that your device can be plotted on the curve for that device because the Wheatstone bridge does not add error.
the parabolic curve of your sensor output will have the same curve on the Wheatstone bridge output.
follows Thomas Crappers law of crap in equals crap out. The Wheatstone bridge will not lineralize the curve
Interesting. Learning a lot from here. Thanks guys. Can I post my schematic and explain what I am doing, and post questions about it? Maybe I will have to remodify a large part of it. IDK
Hi, please post your schematic.
If the only thing you ultimately want to do is measure the strain, then you can use one bridge with both resistance gauges.
You will not be measuring the resistance of anything, you will be measuring the current produced by the imbalance in the values of the gauges and using the current to interpret the strain.
please post the schematic.
we are here to help.
sometimes to offer suggestions
sometimes to clear up misunderstandings (often ours)
but mostly to help
.
