I'm pretty new to electronics and I bought a pt100 thermo-element to read in temperatures using my arduino. Therefore, I'm trying to setup a wheatstone bridge in this fashion:
This test setup contains 4x 100 ohm resistors, so I should have a voltage between the green and white cable of 0V. This is the case when I measure the voltage using a multimeter. However, when i feed the signal into my arduino (yellow cable, to A0) I get a voltage of 2.5V, which is probably somehow because of the voltage division. I tried to plug in an op-amp but that also didn't give me the result I was looking for.
I read somewhere I could feed, for example, the white cable to one analog in and the green to another analog in and use the white one as reference. This works of cause but I don't see the advantage in this because in this case I could also just use a simple voltage divider, or not?
could someone help me out here?
thanks a lot
matjes_
Also, it would help if you'd provide a complete schematics of your setup or at least a picture showing the whole setup, including the sensor and other parts.
Well I bought a pt100 with two wires and the idea was to set up something according to this illustration (of the wiki article):
The setup in the image I uploaded was just for testing purposes. In the end I plan to exchange the resistor in the upper-left with the pt100. Well, but since the bride is balanced at the moment and I'm reading for Vb (illustration above) 2.5V instead of 0V I guess I'm doing something wrong there.
Using a multimeter it seems to work, so my question is what is the right way to connect the Vb signal to my arduino?
Pure platinum has an alpha of 0.003925 ohm/ohm/°C in the 0 to 100°C range ...
So the output you get from the bridge will be very small. For example, if you power the bridge from 5V, the change in voltage across the bridge will be at most 9.8mV per degC.
Here are a couple of options:
Connect the thermometer in series with a fixed, low-temperature-coefficient resistor of about the same resistance between 5V and ground (or use the 3.3V pin instead of +5V and also use 3.3V as the analog reference). Connect the junction to an analog input pin. The analog reading will change by around 2 per degC, giving you a resolution of 0.5 degC.
If you need greater resolution, then you will need to amplify the signal first. One possibility is to use an instrumentation amplifier. However, a simpler arrangement using an op ap is possible, see attached schematic. The gain of the amplifier is 1 + (2*R2/R1). Use the 3.3V pin as the analog reference. R0 is a resistor with a low temperature coefficient and the same resistance as the thermometer near the centre of the temperature range of interest.
EDIT: both these suggestions assume that it is OK to put about 1.7V across the thermometer (assuming you power the half bridge from 3.3V). But this may be too much. What is the resistance of the thermometer?
Arduino is grounded, so the voltage measured is not across the bridge output, but rather across ground and the one output of the bridge, measuring half the the supply potential. Power the bridge from a second power supply and connect the Arduino's ground and analogue input to the bridge outputs, or use the difference of two analogue inputs across the bridge to map to temperature value
@dgar: Thanks a lot. That was the answer I was looking for.
Nonetheless, thanks a lot for the other suggestions! I guess I will try first to just put it in series with another resistor (0.5°C resolution should be enough). Anyways, the Kelvin bridge looks very interesting. I will take a closer look at it.
dgar:
Arduino is grounded, so the voltage measured is not across the bridge output, but rather across ground and the one output of the bridge, measuring half the the supply potential. Power the bridge from a second power supply and connect the Arduino's ground and analogue input to the bridge outputs, or use the difference of two analogue inputs across the bridge to map to temperature value
There's no need to use a full bridge arrangement, because the Arduino ADC is ratiometric; so all you need to do is make sure that Aref and the bridge are powered from the same power supply.
To put it another way: suppose you do use a full bridge, measure the voltage at the two outputs using two analog inputs, and take the difference as you suggest. Then the analog reading on the side that uses 2 fixed resistors will be constant. So there is no need to take any readings on that side at all.
OTOH the Kelvin bridge does have a significant advantage, which is that if the thermometer is of the 4-wire type (i.e with separate excitation and sense connections), then it cancels out the resistance of the connecting wires.
There's no need to use a full bridge arrangement, because the Arduino ADC is ratiometric; so all you need to do is make sure that Aref and the bridge are powered from the same power supply.
The point of using a bridge is to account for supply voltage instabilities and to widen the voltage range across the sensor. Could be running the bridge off 12V supply and arranging such that the sensor operates between 0 and 5v. Guess with Arduino supply of 5V and 4,9mV resolution A2D, single series resistor and stabilizing through Aref is best shot.
There's no need to use a full bridge arrangement, because the Arduino ADC is ratiometric; so all you need to do is make sure that Aref and the bridge are powered from the same power supply.
The point of using a bridge is to account for supply voltage instabilities...
The ADC itself accounts for the variation in the supply voltage, provided that the bridge supply and the analog reference are the same, and are stable over the period of the ADC reading - which is why I suggested using the 3.3V supply instead of the 5V supply, because the 5V supply is not very stable when the Arduino is powered from USB.
dgar:
... and to widen the voltage range across the sensor. Could be running the bridge off 12V supply and arranging such that the sensor operates between 0 and 5v.
You improve the resolution a bit (somewhat less than double) that way, assuming the sensor is OK with 5V across it - which I am not sure of, because of the low resistance (100 ohms) of the only one I have found detailed specifications for.
