Single Element Ultrasonic Sensor

[cr0sh]

All - how does one go about designing a single-element ultrasonic sensor; that is, one that you pulse the sensor, then switch it to receive (after a short wait to settle the transducer), and it receives the echo (and you do your speed-of-sound calc to measure the distance)?

I've seen plenty of circuits on-line showing how to build a dual-element device (much like the Parallax PING, with transmitter and receiver elements), but I have never seen a circuit for a single transducer, other than the how-to articles for scavenging Polaroid camera sensors.

Now, I have several Polaroid cameras with the ultrasonic sensors, waiting to go under the knife (several Sun 660s and Spectras); I do plan on using them for robotics work in the future. I also have, though, several 24K "ultrasonic" transducers that I want to use, and rather than use them in pairs, I would like to build a similar single transducer system.

I understand that to properly pulse the transducer, you have to set up a dual-ended push-pull type amplifier, but how do you disconnect/reconnect the receive portion of the circuit? I am guessing some kind of transistor/fet switch or something similar, of course, but having never seen such a circuit - I am not sure what can be done here.

Does any one have ideas? Have you seen such a circuit? Have any links? Once again, I would want to use this in a robotics system, so TTL-levels for i/o would be ideal...

 

[colin55]

To use a single sensor, you don't pulse it then listen, but create a more-clever circuit.
You send out a frequency and at the same time listen for the "interference" the "interference fringes." From this you can actually work out the distance of the object.

 

[ericgibbs]

All - how does one go about designing a single-element ultrasonic sensor; that is, one that you pulse the sensor, then switch it to receive (after a short wait to settle the transducer), and it receives the echo (and you do your speed-of-sound calc to measure the distance)? I also have, though, several 24K "ultrasonic" transducers that I want to use, and rather than use them in pairs, I would like to build a similar single transducer system.

I understand that to properly pulse the transducer, you have to set up a dual-ended push-pull type amplifier, but how do you disconnect/reconnect the receive portion of the circuit? I am guessing some kind of transistor/fet switch or something similar, of course, but having never seen such a circuit - I am not sure what can be done here.

Does any one have ideas? Have you seen such a circuit? Have any links? Once again, I would want to use this in a robotics system, so TTL-levels for i/o would be ideal...

hi,
I see you want to measure distance using ultrasonics.
It is possible to transmit a 'pulse' at the transducer frequency.
A 24kHz frequency is on the low side for getting good distance measurement.
The method is to gate the 24kHz transducer on for say about 1mSec, this ideally at 24kHz generate will generate a pulse train of 24 cycles.
For a 1mSec frequency burst the 'length' of the burst in air will be about 0.33mtr also the ringing decay time of the transducer could be approx 0.1mSec.
This means that for this set up the minimum distance you could measure would be not less that 0.4mtr or so.
You could experiment by using short TX on times, use a scope to check the transducers response.

For the receiver section, its connected to the transducer via a resistor, value in the order of 10K, on the low side of the 10K, two diodes are connected in reverse parallel to 0v.
The echo signal to be amplified is taken from the junction of the 10K and the two diodes.
The purpose of the two diodes is to clamp the TX pulse signal at the receiver amplifier input to 0.7V, for the low level echo RX signal the diodes appear as an open circuit.

How do you plan to amplify and display the echo signal/distance.????

 

[cr0sh]

To use a single sensor, you don't pulse it then listen, but create a more-clever circuit.
You send out a frequency and at the same time listen for the "interference" the "interference fringes." From this you can actually work out the distance of the object.

Hmm - this is actually an interesting approach, and something I will consider - it certainly makes sense! Thank you.

 

[cr0sh]

hi,
I see you want to measure distance using ultrasonics.
It is possible to transmit a 'pulse' at the transducer frequency.
A 24kHz frequency is on the low side for getting good distance measurement.
The method is to gate the 24kHz transducer on for say about 1mSec, this ideally at 24kHz generate will generate a pulse train of 24 cycles.
For a 1mSec frequency burst the 'length' of the burst in air will be about 0.33mtr also the ringing decay time of the transducer could be approx 0.1mSec.
This means that for this set up the minimum distance you could measure would be not less that 0.4mtr or so.
You could experiment by using short TX on times, use a scope to check the transducers response.

Ok - that should be alright; it would be something experimental because I have these transducers to play with.

For the receiver section, its connected to the transducer via a resistor, value in the order of 10K, on the low side of the 10K, two diodes are connected in reverse parallel to 0v.
The echo signal to be amplified is taken from the junction of the 10K and the two diodes.
The purpose of the two diodes is to clamp the TX pulse signal at the receiver amplifier input to 0.7V, for the low level echo RX signal the diodes appear as an open circuit.

You know, I did some further searching, and I found these examples (for a different transducer, I realize):

Ultrasonic Sensors

I had wondered what the double reversed diodes were for; I thought it was for such kind of "protection", but I wasn't sure how it worked; your explanation makes sense.

How do you plan to amplify and display the echo signal/distance.????

Some form of op-amp, likely - not sure which - and its output would be fed into an analog input of my microcontroller (likely an ATMega328 - I'm an Arduino fan).

Thank you for your explanation!

 

[colin55]

You can't use a single sensor for both transmit and receive because you can't get it to settle down fast enough to read it.

 

[ericgibbs]

You can't use a single sensor for both transmit and receive because you can't get it to settle down fast enough to read it.

You can, I have designed and manufactured ultrasonics range finders/sounders for many years using this technique.

You do talk a load of rubbish, if you dont understand a topic, shut the heck up and stop misleading OP's.

 

[colin55]

Show me your circuit.

 

[ericgibbs]

Show me your circuit.

So that you copy it and post it on your website and claim it as yours, no way hosay.!

 

[Mikebits]

The bandwidth of a piezoelectric transducer does have the bandwidth to operate full duplex as Eric described.
**broken link removed**

 

[ericgibbs]

The bandwidth of a piezoelectric transducer does have the bandwidth to operate full duplex as Eric described.
**broken link removed**

hi Mike,
Just to show that it can be and has been done, I have attached a clip from one of my earlier designs, please note the document date.
I have deleted a major part of the diagram so that it would be useless to any plagiarist haunting these Forums.

 

[colin55]

I tried to use a 40kHz transducer in single mode and it did not stop oscillating in time to receive the signal. Show me anywhere where this has been done.
I searched the internet for hours some year ago and could not find where it has been done. Eric says he has a circuit but clearly has failed to produce.
I f he had a circuit he would not have replied in such vague terms in his first reply.
I am particularly annoyed at "armchair mechanics" who give opinions that they cannot substantiate and send unsuspecting newcomers in directions that are totally misleading.
Secondly, the transmitter is only half as good as a receiver.
Obviously no-one has done any experimenting with these devices as these points have never been mentioned.
I have and I know.
You diagram does not show the type of sensor and does not show the range. We are talking about distances up to 50cm for robot work. This is where they do not work.

 

[ericgibbs]

Obviously no-one has done any experimenting with these devices as these points have never been mentioned.
I have and I know.

Again its the case if you don't know how to do something, you consider it cannot done, what foolish arrogance.

This is what happens when you get someone who has no EE qualifications, who after tinkering with electronics for few years considers he is in a position to advise others.!

 

[colin55]

You talk about 24kHz transducers. Where can you get 24kHz transducers.
The circuit diagram does not indicate what distance is being measured and what transducer is being employed. I am talking about standard 40kHz transducers and detecting a signal from 10cm to 50 cm as is required by the original poster. None of these features have been answered. That circuit diagram is too vague to warrant investigation and secondly, if you had a realistic circuit diagram you have been the first to offer it to the original poster.
You have already said that the transducer will not work below 40cm and this is exactly what I have said, so I don't know what the confusion is all about.

 

[ericgibbs]

This what the OP posted.

Now, I have several Polaroid cameras with the ultrasonic sensors, waiting to go under the knife (several Sun 660s and Spectras); I do plan on using them for robotics work in the future. I also have, though, several 24K "ultrasonic" transducers that I want to use, and rather than use them in pairs, I would like to build a similar single transducer system.

Read the posts carefully

You talk about 24kHz transducers. Where can you get 24kHz transducers.
The circuit diagram does not indicate what distance is being measured and what transducer is being employed.

Nowhere has the OP stated a desired range/distance

I am talking about standard 40kHz transducers and detecting a signal from 10cm to 50 cm as is required by the original poster.

I did suggest a way of connecting the transducer to the TX and RX sections, had the OP come back I was prepared to expand on the technique.

None of these features have been answered. That circuit diagram is too vague to warrant investigation and secondly, if you had a realistic circuit diagram you have been the first to offer it to the original poster.

You didn't say anywhere or anything about the 40cm limit I mentioned.!

You have already said that the transducer will not work below 40cm and this is exactly what I have said, so I don't know what the confusion is all about.

Again for some reason you cannot accept what most of us already know.??

If you don't know how to do something, just say so, no one will feel less of you.



EDIT:
I see that you have used your usual trick of going back and Editing your posts so that they fit in with your version of the truth, get a grip

 

[colin55]

You are just running around in circles. You have provided absolutely nothing for the original poster. You say you have a circuit diagram suitable for robotics work and this obviously means detecting distances from 10cm to 50 cm. Instead of talking to me, help the person who wants help.
I am not here for an argument with you.

 

[ericgibbs]

You are just running around in circles. You have provided absolutley nothing for the original poster. You say you have a circuit diagram suitable for robotics work and this obviously means detecting distances from 10cm to 50 cm. Instead of talking to me, help the person who wants help.
I am not here for an argument with you.

You misunderstand, I am not arguing with you, I am trying to educate you in the art of electronics.!
Look at the link that the OP posted in post #5, it shows ways of using a single transducer.

The OP has already said thanks for my earlier explanation, especially the use of the two reverse connected diodes.

 

[colin55]

That circuit in posting #5 states the settling time is 5mS = 1.6 metres. We are talking about the use for a robot in the range 10cm to 50cm. Nothing has changed any of my statements.

 

[ericgibbs]

That circuit in posting #5 states the settling time is 5mS = 1.6 metres. We are talking about the use for a robot in the range 10cm to 50cm. Nothing has changed any of my statements.

Trying to help you is just a waste of time, you make things up as you go along..........................

 

[colin55]

You are not helping me. I don't need any help. I know it can't be done becasue I tried it over 12 months ago. And the circuit provided in posting #5 proved it cannot be done.
All you did was waste time going around in circles, proving nothing and not making it clear to the original poster that short distances cannot be detected.