Single ultrasonic transceiver switching between Rx and Tx

[Shahar Levi]

Hi,
I am trying to build an ultrasonic detector with only one transceiver with switching between Tx and Rx. I am using MSP430 to generate 8 pulses at 40kHz and the switch control signal, CD4053 switch, mcp604 op-amp and Ekulit A-16PT transceiver with 1k Ohm nominal impedance.
While when using 2 sensors and no switch I have no problems at all, I get problems when I am trying to use switch and 1 transceiver. There are some sort of ringing right after the transmission which lasts a bit more that 1 millisecond and generate some noise in the receiving circuit and does not allow detection in the range of the nearest 30 cm to the sensor. after the 30 cm I am able to detect again but 30 cm is way too much of detection range to lose.
could someone give me a hint on how to remove this ringing noise and allow detection, or at least how to narrow it a bit down from 1 ms?

thank you very much.

p.s
1) so far the circuit was tested on a breadboard
2) the 2 npn transistors in series coming from P1.7 are there first to drive the transmission from 3v to 9v and then to invert the signal
3)I was trying to reduce the amplification gain or to use only 1 op-amp and it showed some improvement but not enough.

the Schematics:
**broken link removed**

no problems - ultrasonic with 2 sensors. the upper signal is the transmission and bottom is the measuring before diode D1 with object at 40 cm:

**broken link removed**

no problems - ultrasonic with 2 sensors. the upper signal is the transmission and bottom is the measuring at the OUT with object at 40 cm:

**broken link removed**


1 transceiver with switching. the upper signal is the transmission and bottom is the measuring before Diode D1 with object at 40 cm:

**broken link removed**


1 transceiver with switching. the upper signal is the transmission and bottom is the measuring before Diode D1 without object:

**broken link removed**

1 transceiver with switching. the upper signal is the transmission and bottom is the measuring at the OUT with object at 40 cm:

**broken link removed**

1 transceiver with switching. the upper signal is the transmission and bottom is the measuring at the OUT without object:

**broken link removed**

 

[jiripolivka]

Typically, transducer ringing is due to its vibration. You can try to damp it with a parallel resistor but then the range will be reduced.
Range meters require that transducers are selected from more units.
You can also try damping the transducer resonator by a drop of glue or a piece of cloth over the case.
Maybe you can modify the driver to reduce ringing, by a matching circuit in the output transformer.

 

[FvM]

None of the links is valid for me.

 

[Shahar Levi]

is it better now?

Schematics:



no problems - ultrasonic with 2 sensors. the upper signal is the transmission and bottom is the measuring before diode D1 with object at 40 cm:

no problems - ultrasonic with 2 sensors. the upper signal is the transmission and bottom is the measuring at the OUT with object at 40 cm:



1 transceiver with switching. the upper signal is the transmission and bottom is the measuring before Diode D1 with object at 40 cm:



1 transceiver with switching. the upper signal is the transmission and bottom is the measuring before Diode D1 without object:



1 transceiver with switching. the upper signal is the transmission and bottom is the measuring before Diode D1 without object:



1 transceiver with switching. the upper signal is the transmission and bottom is the measuring at the OUT with object at 40 cm:



1 transceiver with switching. the upper signal is the transmission and bottom is the measuring at the OUT without object:

 

[Audioguru]

The transducer is piezo so of course it resonates which causes ringing. Its datasheet says that has ringing for no more than 1.2ms.

 

[Shahar Levi]

thanks for your reply. I see it now on the datasheet and the ringing I get is indeed about 1ms or a bit more. is there a way to damp it? placing a parallel resistor didn't change anything. the thing is that I am interested in detecting objects in the range of 10-50 cm or so. the 0-30 cm loss would be too much

 

[FvM]

These sensors have small bandwidth and respective high Q and considerable "ringing". I see that you already terminate the sensor in receiving mode with a real impedance about equal to the nominal sensor impedance. So there's possibly no better damping available. You would determine the eaxct complex sensor impedance to see if a more effective termination can be found. The datasaheet doesn't have any impedance details.

You probably need a more sophisticated receiver circuit with dynamical varied gain (or a least a gain compressing characteristic) to detect echos near the sensor.

 

[Shahar Levi]

just an idea, what if I use RC circuit on the falling edge of the last transmission pulse, as one would gradually decelerate a swing from swinging further (something like it the pic below). Do you think it could possibly prevent the ringing or decrease it?

 

[FvM]

No, it does not reduce the ringing. The resonator Q is such that you need at least several periods to build up the full magnitude. The form of a single edge doesn't matter at all.

 

[D.A.(Tony)Stewart]

To suppress the ringing after Tx , you may need to short out the transducer with transistor active switch to ground for <1ms , rather than open circuit switch. The tail of the high impedance resonant energy will be absorbed quickly. ( Like a motor/generator brake effect) That should work using one shot or some equiv. cct.

 

[FvM]

Fastest oscillation decay will be achieved by terminating the transducer with a matched real impedance, as previously mentioned. It looks like the circuit has already a termination in this range.