4
\$\begingroup\$

I have an analog sensor which outputs a value in range 0-3V depending on soil moisture. The Xbee's ADC requires a value between 0 and 1.2V.

I was thinking about using a simple voltage divider:

Simple voltage divider

Where R1 would be twice as large as R2 thus giving me effective 1/3 voltage division.

The part I do not get is what size resistors should I choose? While keeping in with the resistor ratio, I have tried a small resistor value first - around 100ohm for the R1. This did not give me any reading on my Xbee. Then I found an online tutorial for a different type of sensor (LM335) where they used 100Kohm for the same one and I found this to seemingly work a bit better in my case, giving me varied reading according to varying soil moisture.

How can I calculate what would the best resistor sizes for this application?

Some information about the rest of the circuit:

My power supply is 5V and the is powered directly from it, the Xbee also has a voltage regulator to bring it down to its required 3.3V (active XBBO). Grounds are connected.

EDIT

After bit of trial and error, I have determined the sensor website must be wrong because the moisture sensor behaves as it would have 10kΩ resistance (impedance) - I was able to half its voltage by simply using another 10kΩ resistor connected to the ground (and third it using two of those resistors). All the measurements seem to be working just fine so I may get away without using the amplifier. Either way, it would be useful to know what the parameters of Xbee's ADC are as the voltage requirements are different compared to the Series 1 which makes me reluctant to simply assume that the information is the same.

Anyway, that brings me to the basic question again - knowing the sensor impedance and other parameters of the circuit, how do I determine what resistor sizes would be a good match?

Lastly, thank you all very much for your input and help so far!

\$\endgroup\$
6
\$\begingroup\$

According to the product details, the output impedance is 100kΩ, so 100Ω would almost completely short the signal.
To get 1.2V from 3V, you need a ratio of 1.2 / 3 = 0.4

Assuming a steady 100kΩ output impedance, you will need a single 66.6kΩ resistor from the sensors output to ground to divide the range down as necessary:

3V * (66.6 / 166.6) = 1.2V

Looking at the xBee datasheet, the ADC input impedance is only 10kΩ, so an signal impedance of ~40kΩ (which is what you would have with the above) is not ideal, and you may lose accuracy.

Ideally, the best thing would probably be to use an simple non-inverting opamp placed in between this and the xBee ADC input to buffer the signal.

\$\endgroup\$
  • \$\begingroup\$ I also saw that 100 k output impedance, but that seems impossibly high to me. That's more for a current source than a voltage output. \$\endgroup\$ – stevenvh Sep 19 '12 at 11:32
  • \$\begingroup\$ I agree it seems high, but it's worth trying to see if it's correct. Also, just using a trimmer pot and adjusting empirically might be an idea if it's a one off project. \$\endgroup\$ – Oli Glaser Sep 19 '12 at 11:43
  • \$\begingroup\$ Of course, if the output impedance is this high, there is also the ADC input impedance which needs checking to make sure it is high enough not to worry about. \$\endgroup\$ – Oli Glaser Sep 19 '12 at 11:47
  • \$\begingroup\$ I could not find any reliable information but the "rumour has it" that the Xbee's ADC input impedance is around 100kΩ \$\endgroup\$ – petr Sep 19 '12 at 11:55
  • \$\begingroup\$ (just a side note - measuring the voltage out with my multimeter works just fine and I am getting the reading I would expect) \$\endgroup\$ – petr Sep 19 '12 at 11:58
2
\$\begingroup\$

The 150 Ω may have overloaded the sensor's output and have caused the voltage to sag, which would explain a wrong reading (though no reading at all is indeed very low).

100 kΩ on the other hand may be a bit high. It can be in the order of the ADC's input impedance, and the impedance would also distort the reading. 8.2 kΩ for R1 and 4.7 kΩ for R2 seems like a good compromise, which will give you 1.09 V out for 3 V in.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.