0
\$\begingroup\$

Hello — recently I have been working on an ultrasonic transducer project. The transducer simply works by transmitting pulses while the receiver waits for the echo. When there is water flow in a pipe, pulses from the transmitter can be read by the receiver; otherwise, when there is no flow, the receiver gets nothing.

My goal right now is to convert the echo signal to a logic signal that an MCU can read (3.3 V logic). The echo signal is 1 MHz with approximately 60 mV peak-to-peak. Echo signal

I created a simulation of cascaded stages with feedback using low-pass and high-pass filters (each stage has a gain of 10), followed by a comparator to convert the signal to 3.3 V. Simulation Plot

Then I built the real circuit to test it. It turned out the output differs from the simulation: the comparator output stays high even though there is no echo, and the filter output also remains high.

  1. What could be wrong with my schematic?
  2. Is there any alternative approach to achieve this goal?
\$\endgroup\$
1
  • \$\begingroup\$ Your har to read componnt name for IC appears to be a NE5532. If so a quick first trial could be one of those with unity gain at the input. \$\endgroup\$ Commented Oct 16 at 12:52

2 Answers 2

Reset to default
1
\$\begingroup\$

If you connected the transducer into oscillosope probe, it provides 60mV when loaded with rougly 10 megaohms and 20 picofarads.

Your circuit has much lower impedance of 300 ohms, so the weak transducer cannot provide much output into such low impedance.

Buffer the signal first, perhaps with amplification.

\$\endgroup\$
0
\$\begingroup\$

This is a bit late...

What's your power supply voltages? The NE5532 specifications state \$\pm5\$ to \$ \pm15\$ V for the power supply voltage. Your schematic is incomplete, or the V+ and V- power supplies are off the schematic view.

The NE5532 has a gain-bandwidth product (GBW) of 10 MHz. This is not a suitable opamp with a high-gain filter. You need a GBW of probably over 100 MHz.

Be careful of saturation issues regarding opamps. Some opamps don't do well coming out of saturation. SPICE probably won't model this well.

What's the impedance of the transducer at 1 MHz? If the impedance is too high, the first stage will load down the transducer which reduces the sensitivity. Generally, 1 MHz transducers are fairly low impedance devices, in the 100 to 1000 ohm range. You can test this out by monitoring the output of the transducer with an oscilloscope and connect/disconnect the input of your circuit to the transducer to see how much it loads the transducer.

The rectifier should be a small signal Schottky diode, not a slow rectifier diode. Even a 1N4148 would be better a better choice than a 1N4007.

It is questionable that you need a filter at all due to the high frequency of operation except a high-pass to suppress low-frequency noise (below 10 kHz to 100 kHz). Choosing the right input impedance, coupled with the simple capacitance (capacitance measured at lower than 5 times below resonance) of the transducer can be the high-pass filter, although, this doesn't give the best noise performance which probably isn't an issue for this design.

If you have other acoustic sensors at other frequencies and need a bandpass filter, consider using a simple RLC bandpass filter. Perhaps suitable values would be 1k ohm series resistor followed by paralleled LC components of 22 uH and 1.15 nF.

SPICE is only as good as the models. What you get in a SPICE simulation may not be reflected in reality unless you understand the limitations of the models.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.