View Full Version : Voltage divider or Bridge
04-24-2008, 12:31 PM
I am intending to read a Temp sensor through a 16bit A/D to the Propeller. The question is that i can not decide on a voltage divider or a bridge circuit. Can some of the more experienced people help me understand the concepts better.
04-24-2008, 12:38 PM
As usual, start with the Wikipedia: en.wikipedia.org/wiki/Bridge_circuit (http://en.wikipedia.org/wiki/Bridge_circuit).
If you want suggestions, you'll need to describe more what you're trying to do and, specifically, what kind of temperature sensor you want to use, what temperature range you want to measure, and why did you think of a bridge vs. a voltage divider.
04-24-2008, 01:07 PM
As Mike suggests, look into Wikipedia and refine the·goals that you want to accomplish in your project.
A Bridge circuit will provide immunity to any voltage variances on the supply lines while a voltage divider will not.
depending on other aspects of your design (a well regulated power supply) , a Bridge·may not be necessary.
Beau Schwabe (mailto:firstname.lastname@example.org)
IC Layout Engineer
04-24-2008, 06:36 PM
The output of bridge type sensor circuits is ratiometric to the supply voltage.·Any fluctuation of the supply will appear on the output unless the fluctuations are common mode(on each supply line).
The·traditional Wheatstone bridge·technique·uses an adjustable·element to null the bridge output;·the unknown(sensor) element·value can then be obained from the known value of the adjustable element - in this case the result is independent of the supply voltage.
04-24-2008, 08:52 PM
I think the bridge is useful if you are worried about common-mode noise due to long wires to the sensor or because the signal is weak.· You·want an ADC with a differential input to use a bridge.·
If you're using an active sensor, like the LM34, I think you can just send the output straight into the ADC...
04-24-2008, 10:40 PM
Thanks for the comments so far and Yes I have read wikipedia.
I intend on using a voltage divider or bridge circuit that will connect to a thermistor [I have not selected the perfect one just yet, but the temp ranges would be from -50 to 150c]. The accuracy will be around the order of .01. So if i am correct then i have a total rane of 200 degrees with a 16bit a/d i would be able to read to the .01 with no problems. There will be a 6' cable that will have other wires inside it, including the sensor at one end, and 2 of the wires inside will be running up to 5amps. There will be surrounding [out of the case and wire] noises that may pose problems.
This is why i am asking people that may have more experience using the 2 circuits. Currently i am running a device that uses an op-amp, but it is not the ideal solution for the new prototype.
*Edited I will be using a switching power supply
Post Edited (grasshopper) : 4/24/2008 3:42:37 PM GMT