21 replies to this topic

[Michel Trahan]

Please note that the analog-digital converter on the Netduino is 10-bit, but it is not uncommon to lose 1- or 2-bits of accuracy with integrated microcontroller ADCs (for a variance of 2-4 on the 0-1023 scale).

I have been trying to get stable data from a LM335A Temperature sensor and I do get a variation of 3 to 4 out of 1023. You say this is normal ? When I look at the oscilloscope data, it is quite stable (better using 10 pF capacitor) but still a variance of 3 to 4 ... you say this is normal ? The variation on the oscilloscope is does not even equate 2 out of 1023 so why more variance ?

A variance of 3 would make sense since the LM335A Temperature sensor is 1C accuracy which equates to 3 out of 1023, therefore a variance of 3 would be exactly the accuracy of the sensor. Am I making sense here ? But the oscilloscope seems to be more stable than what the Analog pin gives out ... why would that be ? The voltage to be read is more stable than the values obtained by the analog pin reading ... why ?

I use 3.3V with a 300 Ohm resistor and a 10 pF capacitor.

I did see that I have 3.250 instead of 3.3 when I put the oscilloscope on the power thread on the breadboard ... why I do not know ... shouldn't be 3.3 ? (was using USB power)

[Chris Walker]

I have been trying to get stable data from a LM335A Temperature sensor and I do get a variation of 3 to 4 out of 1023. You say this is normal ? When I look at the oscilloscope data, it is quite stable (better using 10 pF capacitor) but still a variance of 3 to 4 ... you say this is normal ? The variation on the oscilloscope is does not even equate 2 out of 1023 so why more variance ?

ADCs are funny things. If you go "expensive" you can get some very tight readings. ADCs in microcontrollers tend to require a bit more love and attention to get good readings. Or so the electrical engineers tell me.

I did see that I have 3.125 instead of 3.3 when I put the oscilloscope on the power thread on the breadboard ... why I do not know ... shouldn't be 3.3 ? (was using USB power)

If your USB port is running at a fairly low voltage, the power regulator may not be able to put out a full 3.3V. In our testing, we get pretty consistent 3.29V-3.31V outputs from them when 5V+ power is input into the Netduino. But there's a minimum voltage drop of 1V or so...

Chris