Ultrasonic transducers.

[solofrain]

I'm trying to build a circuit for ultrasound detection, and I'm asking for help to select ultrasonic transducers for it. The detailed requirements are:

. The distance between the transmitter and the receiver is approx. 30 ft.The transmitter should have low power consumption, so the receiver must have high sensitivity (-70 db minimum? I'm not sure).
. The receiver should be frequency selective (e.g. 40+/-0.1 KHz).
. The transmitter cannot be aligned with the receiver in real application.
. No data will be carried with the ultrasound. As long as it can be detected at the receiver side, that's OK.

Can somebody recommend ultrasound transducers that are OK for me? Thanks.

 

[BradtheRad]

I once got a couple of these at All-Electronics discount mail order.
$1.25 ea.
The customer comments give helpful tips on using them.
Can't say if it will perform to your specs.

 

[kripacharya]

Any basic transducer which works at 40Khz should be OK. But you DO need to have the proper Tx driver circuitry, and a good receiver circuitry to get the 30 ft range.

 

[Audioguru]

Ultrasonic transducers are VERY directional. They will not work 30 ft away unless they are pointing directly at each other.

 

[colin55]

"I'm trying to build a circuit for ultrasound detection"
With your limited knowledge of electronics you are best to buy the $4.00 HCSR04 Distance Measuring Module on eBay

 

[FvM]

I don't hear a clear word what you want exactly to detect. "ultrasound detection" literally means to detect the presence of ultrasound waves.

I rather assume you want to detect the presence of an object or e.g. a mechanical quantity like distance.

 

[solofrain]

I don't hear a clear word what you want exactly to detect. "ultrasound detection" literally means to detect the presence of ultrasound waves.

I rather assume you want to detect the presence of an object or e.g. a mechanical quantity like distance.

Thank you FvM for pointing this out. Yes, I want to detect the presence of the ultrasound wave. To be more specific, I need to capture the sine wave. What I want to do is to calculate the position of an ultrasonic wave source by calculating the time difference of arrival of the wave at three receivers. The receiver may be closer to the transmitter, but 30 ft might be the worst case. I can change the direction of the receiver array to search for the approximate position of the source, but a receiver with high sensitivity will help a lot, I think.

- - - Updated - - -

Ultrasonic transducers are VERY directional. They will not work 30 ft away unless they are pointing directly at each other.

Yes, this could be the most difficult part. I can change the direction of the receiver to search for the transmitter. I also saw someone used a diffuser with the transmitter to make it "non-directional" (something like Mirage used on its speakers), although this reduces "power per direction".

- - - Updated - - -

"I'm trying to build a circuit for ultrasound detection"
With your limited knowledge of electronics you are best to buy the $4.00 HCSR04 Distance Measuring Module on eBay

A distance measuring module may not fit my application, otherwise I will definitely go and grab one =)

 

[colin55]

A I said, it will the cheapest to buy 3 of the $4.00 HCSR04 Distance Measuring Modules on eBay.
If you didn't understand what I say previously, how of earth are you going to build anything?

 

[FvM]

A I said, it will the cheapest to buy 3 of the $4.00 HCSR04 Distance Measuring Modules on eBay.
If you didn't understand what I say previously, how of earth are you going to build anything?

I believe that the distance measuring module can be used by disabling the transmitter part and modifying the circuit if necessary.

I didn't hear however a specific suggestion of yours related to the OP's problem, so nothing that could be misunderstood.

 

[colin55]

I believe that the distance measuring module can be used by disabling the transmitter part and modifying the circuit if necessary.

I didn't hear however a specific suggestion of yours related to the OP's problem, so nothing that could be misunderstood.

The first thing you have got to do is buy the units. They are only $1.86 each.

You place the three in a triangle and use them in a line-of-sight situation.
You don't disable anything.

 

[solofrain]

The first thing you have got to do is buy the units. They are only $1.86 each.

You place the three in a triangle and use them in a line-of-sight situation.
You don't disable anything.

Thank you for your patient input. I read the HCSR04 datasheet as you recommended. It looks like a combination of an OSC, a transmitter and a receiver. My question is, if there are multiple objects around this module, and all these objects produce echoes, how do I pick the one I want to detect?

 

[colin55]

Thank you for your patient input. I read the HCSR04 datasheet as you recommended. It looks like a combination of an OSC, a transmitter and a receiver. My question is, if there are multiple objects around this module, and all these objects produce echoes, how do I pick the one I want to detect?

The whole thing has to be under the control of a program in a micro.

 

[solofrain]

The whole thing has to be under the control of a program in a micro.

Actually I have a PowerPC and an FPGA that can do precise control and complicated computation. To my understanding, HCSR04 can be helpful in measuring the distance between the module and a single (the nearest?) object. But I think it could be a mission impossible to differentiate the echo of the desired object (whose distance is unknown, of course) from the multiple echoes using this kind of method.

But your recommendation of HCSR04 is definitely valuable to me. If the datasheet gives accurate parameters, it can detect ultrasonic wave over a distance of 800cm when power supply is 5V. It makes me more optimistic to expect the result to detect over 30ft with amplifiers at transmitter and receiver.

 

[chuckey]

Because the ultra sonic transducers are mechanically resonant they have a limited band width, the cheap ones are bought in matched frequency pairs. Also for optimum results they need to have a resonant circuit associated with them for optimum power matching. To a make transducer "omidirectional", mount it vertical and place on its centre an inverted aluminium cone, this literally bounces the vertical beam out side ways, so it covers all directions. For the ultimate range you need to pulse the signal say on for 1mS leaving it off for 9mS, this way you can get all your .5W into the 1mS, giving 9mS for the transducer to cool down.
On the receiver side, if after detection you have a 1KHZ tuned circuit it will ring every time a pulse comes in, so providing a level that sits above the noise or a IC based band pass filter will do much the same, rejecting noise but not the 1KHZ.
Frank

 

[FvM]

Actually I have a PowerPC and an FPGA that can do precise control and complicated computation. To my understanding, HCSR04 can be helpful in measuring the distance between the module and a single (the nearest?) object. But I think it could be a mission impossible to differentiate the echo of the desired object (whose distance is unknown, of course) from the multiple echoes using this kind of method.

It seems to me that the problem specification has changed from post #7. So I'm unsure what you actually want to achieve.

A serious limitation of HCSR04 and similar modules is that it only returns a "digital" echo timing information ( a pulse edge) but no echo magnitude. In so far a multi object scenario will be most likely beyond it's application range.

 

[solofrain]

Because the ultra sonic transducers are mechanically resonant they have a limited band width, the cheap ones are bought in matched frequency pairs. Also for optimum results they need to have a resonant circuit associated with them for optimum power matching. To a make transducer "omidirectional", mount it vertical and place on its centre an inverted aluminium cone, this literally bounces the vertical beam out side ways, so it covers all directions. For the ultimate range you need to pulse the signal say on for 1mS leaving it off for 9mS, this way you can get all your .5W into the 1mS, giving 9mS for the transducer to cool down.
On the receiver side, if after detection you have a 1KHZ tuned circuit it will ring every time a pulse comes in, so providing a level that sits above the noise or a IC based band pass filter will do much the same, rejecting noise but not the 1KHZ.
Frank

Thank you Frank. Do you have any experience selecting such an aluminum cone? I only saw people using it, but have no idea where to find. Tried online electronic device distributors but no outcomes...
For the receiver, as the transducers are usually frequency picky, e.g. with +/-0.5KHz bandwidth, do I still need an additional filter?

- - - Updated - - -

It seems to me that the problem specification has changed from post #7. So I'm unsure what you actually want to achieve.

A serious limitation of HCSR04 and similar modules is that it only returns a "digital" echo timing information ( a pulse edge) but no echo magnitude. In so far a multi object scenario will be most likely beyond it's application range.

Sorry if my clarification made it more confusing. My application is to to detect the ultrasonic wave from a transmitter whose location is unknown. There may be several transmitters but no more than one will be active at a time. By mentioning multiple object differentiation, I meant that echo ranging doesn't fit my application because even if the transmitters are muted, they will echo ultrasonic wave.

HCSR04 datasheet doesn't give details of the circuit, but it seems the receiver is just the transducer. So I think the output of HCSR04 should be a sine wave since the transducer is a narrow band BPF?

 

[FvM]

HCSR04 datasheet doesn't give details of the circuit, but it seems the receiver is just the transducer. So I think the output of HCSR04 should be a sine wave since the transducer is a narrow band BPF?

No. The datasheet clarifies that the output is a digital signal.

 

[chuckey]

" frequency picky, e.g. with +/-0.5KHz bandwidth, do I still need an additional filter? ", If you use a filter at 1KHZ, this could have a bandwidth of less then 50 HZ, therefore attenuating the noise by an additional 20 DB, allowing you an extra 20 dB of gain.
ref. the ali cone. This comes from an article in "Wireless World" ,from years ago (mid 60s?), some one was making a radar simulator but using ultrasonics instead of GHZ RF. He also had directional launchers based on diffraction gratings. Nothing special about ali, any hard acoustically reflecting surface would do such as gloss painted wood.
Frank

 

[FvM]

The relative small bandwidth of US transducers is usually unwanted in time-of-flight measurements because it "slurs" the pulse response and effectively reduces the time resolution. It also involves arbitrary group delay variations between sensor exemplars.

 

[Audioguru]

The relative small bandwidth of US transducers is usually unwanted in time-of-flight measurements because it "slurs" the pulse response and effectively reduces the time resolution. It also involves arbitrary group delay variations between sensor exemplars.

The "slur" is caused by it ringing like a bell because it is resonant (small bandwidth). It takes time to ramp up its output when it starts then it keeps ringing for a while when it should stop.