Completely Analog Ultrasonic Range Finder

[simply_me]

Hello,



I'm trying to design an ultrasonic range finder using only analog components and that its range will exceed 2m. So far I've designed the Wien Bridge (40Khz), bandpass amplifier, and a sinusoidal to rectangular wave form converter. So what I have in mind is-- The oscillator generates a sinusoidal wave sends it to the bandpass amplifier and from there one input goes to the transmitter and the other goes to the sine to rectangle converter. Now, I'll use a time delay circuit to pulse the rectangular wave (with a delay of about 12ms-- distance/speed of sound--4m/(343m/s), from there i thought to use a comparator to turn the receiver on to wait for a reading. Does anyone know how I can turn the receiver on for just 12ms? On the other side I'll have the sinusoidal wave pulsed all the time. At the receiver end, once it's turned on I want to put a threshold op-amp (not sure on how to implement it, but I'll do some research) and from there to amplifier. Once I have the output voltage readings, I'll use something similar to a voltmeter to read the distance (i.e. just a needle w/distance marks after doing manual measurements). Does that sound a reasonable design? Does anyone can point me to how to go about turning the receiver on/off every ~12ms?
Someone suggested that I'll avoid all that and use frequency sweeping and a frequency to voltage converter, the only thing is that I don't know how to do frequency sweeping in analog, does anyone have any reference for it?
I'll be happy to hear any input/different approach you might have.



Thank you,



A

 

[ericgibbs]

hi simply_you,
For V2F conversion, look at the attached pdf.
What accuracy in distance measurement and maximum range. ?

A circuit diagram of the project would be helpful.

BTW: did that 38kHz BW filter I posted in your other thread on this topic, work OK.??

 

[simply_me]

Hi Eric,

Yes, it works perfectly, I shifted it a little towards 40kHz. The maximum range should be at least 2m and accuracy ~10%.
Thank you for the f2v, I've looked it up earlier today. It's too early in the project to have schematics as I now try to decide on the approach:
1. send a delay rectangular pulse from the transmitter to the receiver to begin time counter, and basically go from the rise of the transmitter sinusoidal wave to the rise of the wave at the receiver end.
2. use frequency sweeping approach (sweep between 39.5-40.5kHz), f2v , and compare v or do the other way around.
The second method seems more elegant but it has a few caveats such as an analog frequency sweep (I can't find an example of it anywhere) and 2kHz band (limited by the transducers).
What are your impressions of the two methods?
Also, is it even possible to build an analog frequency sweep?

 

[simply_me]

I can create a block diagram as I envision it.

 

[ericgibbs]

I can create a block diagram as I envision it.

hi,
Ref the timing pulse method, be aware that the transducer will not reach full power output for the first few cycles of drive,
also on reception there is a delayed rise in transducer output.
You can allow for this delay in the program.

You could use a simple 'linear ramp' analog voltage to sweep a 555 timer output over the 40kHz band.
IIRC recall measureing distance in this way is sometimes called 'chirping'

 

[simply_me]

Chirping was exactly what was I referring to. I'm pretty confident about the timing method, but I think I'll try the chirping approach (for a couple of weeks and see how it goes) because it seems a more elegant approach. I'll look up the parts you've mentioned above.
Thank you for the input, very helpful.

 

[simply_me]

Hi Eric,

I've designed a linear ramp/chirp. The output isn't the best but I think it'll do.
I'm tempted to leave it as it is in a sawtooth form instead of converting it to a sinusoidal.
I might just use an amplifier for the transmitter end, although I'm going to to convert the freq at the receiver end it might give me a few more cm. If the oscillator-- transmitter is always on, should I leave the receiver always on too? I hope to upload the receiver end by tomorrow....

 

[ericgibbs]

Hi Eric,

I've designed a linear ramp/chirp. The output isn't the best but I think it'll do.
I'm tempted to leave it as it is in a sawtooth form instead of converting it to a sinusoidal.
I might just use an amplifier for the transmitter end, although I'm going to to convert the freq at the receiver end it might give me a few more cm. If the oscillator-- transmitter is always on, should I leave the receiver always on too? I hope to upload the receiver end by tomorrow....

hi.
If the TX and RX are always ON you will need some method of synchronising the timing circuits when want to take a range measurement, a take off point could be from the ramp gen.

EDIT:
This is a ramp circuit I used recently, you can take a synch pulse from the 'vo' point

 

[simply_me]

Wow that was a fast reply. Does my linear ramp looks fine to you? I guess I can use something like LM567. So I should pulse my outputs, for example-- 10ms at 40Khz and then 20ms at 0V, and then move to the next frequency?

 

[simply_me]

And thank you

 

[ericgibbs]

And thank you

hi,
Look at my EDIT with the image.

 

[simply_me]

Thank you I'll try that.

 

[ericgibbs]

Thank you I'll try that.

hi.
Look here for idea's.
Google

 

[joco000]

Hello Simply,

I want to build something like you. Does your circuit works?

regards