I'm trying to amplify a faint sine wave (+/- 100mv) from an ultrasonic transducer and then get the RMS into 0-3.3v range, so I can read it on a Netduino.

I'm a software guy by trade, so please excuse me if I use some terms incorrectly.

I have an ultrasonic transducer, which when I put it near the transmitter, gives a perfect sine wave centered on 0v at up to 100mv.

I used an LM386 and a coupling capacitor to to amplify that to a perfect sine wave up to the supply voltage (5v) where it clips, but I'm OK with that. My scope shows it as I expected.

I tried to use a true RMS chip (AD736JNZ) to get the RMS, and ran into problems - nothing works as expected. I simplified the circuit to test my assumptions, so I hooked it up as just a precision rectifier like on page 6 of their applications guide *PDF and got closer, but now things got even more confusing.

If my singal is low, the chip output flat lines at around +1.75v. If I increase my signal, it starts to output something that looks like a half-rectified version of my input, slightly out of phase, with amplitude inversely proportional to the input.

I'm stumped. I hope you all can help point me in the right direction so I can get on to the rest of my robot ;-} Here's a diagram of my breadboard.

 

Hello,

You have made some wrong connections on the AD763.
Pin 3 is not the input, but the bias connection.
Pin 4 should be connected to a - 5 Volts powersupply.

This is the drawing from your posted application note:

Bertus

 

Sorry for the confusion - my drawing was in error - here is a corrected one.

I'm using a battery input, so I'm not sure how to get anything but 0 and Vcc, so that's why I grounded pin 4.

Is +/- required?

 

Hello,

For a battery operated circuit, you can look in the real datasheet:

Bertus

 

Thank you - I think I understand. In English, that diagram is saying:
"Electronics is all relative to "ground", which is just a reference, not necessarily 0v. So, if you are on a battery, use voltage dividers to generate a stable Vcc/2 and wire that up to ground everywhere."

Is that correct? As I said, I'm new to this - for example, I just realized the other day that the load resistor is necessary for the circut to operate ;-}