1
\$\begingroup\$

I wonder whether I can make a small motion sensor using one piezotransducer with a feedback.

It should emit a low-frequency ultrasound (say 33kHz) and I expect to see some signal changes on a feedback pin when something moves nearby.

Does this sound real?
functional diagram

piezodisc

\$\endgroup\$
5
  • \$\begingroup\$ What is the nearest and furthest distance you intend to work over? What transducer are you considering? \$\endgroup\$ – gbulmer Sep 7 '14 at 18:24
  • \$\begingroup\$ I think about covering 250ft² (23m²) room. Nearest distance is 0.4in (1cm). I'm considering to use a low power transducer similar to one on the image above (say 0.2W of maximum power). \$\endgroup\$ – ivan Sep 7 '14 at 18:40
  • \$\begingroup\$ Definitely not.. The piezo will still be resonating from the transmit pulse long after the weak receive pulse has come and gone (for short distances). \$\endgroup\$ – Spehro Pefhany Sep 7 '14 at 19:11
  • \$\begingroup\$ @SpehroPefhany But what if I adjust a dsp ic to subtract a Tx signal from the input? \$\endgroup\$ – ivan Sep 7 '14 at 19:19
  • 1
    \$\begingroup\$ Mathematically it may work in a simplified simulation, but I don't think it will work in reality-- \$\endgroup\$ – Spehro Pefhany Sep 7 '14 at 19:45
1
\$\begingroup\$

At 33kHz, the emitted wavelength is:

340.29ms^1 / 33,000Hz ~= 0.01m, i.e. 1cm

"Nearest distance is 0.4in (1cm)"
The wavelength of one pulse is the same order magnitude as the detection distance.

The ultrasonic sensors I have seen, where I found adequate data, seem to take 2-3 cycles to get to full oscillation, and a similar number of cycles to stop (IIRC a little longer to stop than start). IIRC the active pulse duration was several cycles, so 8-11 cycles in total.

I can't find the link, and I can't remember if the 'stop emitting' was being actively driven, or it was just switching off the power. So you might be able to do a little better by driving the device more effectively.

Purely based on that, I'd expect the minimum detection distance would need to be slightly longer than the length of the pulses round trip journey to make it straightforward to do.

You might be able to characterise the system in such a way that it can detect reflected sound during the 'stop' phase, maybe halving the minimum distance. It might even be better to emit for longer periods, though you might then be unlucky in the interaction of emitter and room reflections.

I believe some bats emit at a wide range of frequencies to help detect target type and distance.

If the 1cm minimum detection distance is a key requirement, maybe look at higher frequency devices or multiple devices. The higher frequency emitters I looked at had much narrower 'beam angles' than lower frequency devices though. So that might introduce a different constraint.

\$\endgroup\$
2
  • \$\begingroup\$ What did you mean by "minimum detection distance would need to be slightly longer than the pulses round trip time" (distance/time)? \$\endgroup\$ – ivan Sep 7 '14 at 20:42
  • \$\begingroup\$ @ivan - Good point, I shouldn't mix distance and time! Hopefully the answer now reads better. The total pulse length (start+oscillate+stop) < round trip distance to object. \$\endgroup\$ – gbulmer Sep 7 '14 at 20:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.