EVAL-CN0343-EB1Z SENSE DISTANCE IS POOR

Hi Jdeca_727, and I apologize for the delay.

What is the physical situation? The Circuit Note document does not provide many specific details aside from:

" ...make  sure  that  the  U3 and U4 ultrasonic sensors are facing  the target.  The target  must  have a  large, smooth,  nonabsorbing  surface. Make  sure  there  are no  objects  within  the circular  cone angle of about 60° from  the  sensor. The  target surface  must  be  perpendicular  to  the sensor.  "

What is the display indicating for far targets? Does it read "Target too close", "Target too far", or an erroneous reading?

Can you provide a photograph of your measurement situation?

-Mark

Hi Mark,

Apologies to Analog Devices. I powered the EVAL-CN0343-EB1Z with an 8V Li-ion battery bank and a cheap 6V DC-DC converter to supply the 6V. The noise was horrid. I used a 6V wall wart and the range improved to about 8.5 meters (a very faint 10 meters as measured before the window comparator.) 

I am not looking for reflected signals nor am I range finding. I'm looking to detect an ultrasonic signal 15 meters away to trigger an event. My thinking is "buy the kit and experiment with the gains".

The situation: I pulled the transmitter and speaker from the eval circuit. I then wired the xmit transducer (bought separately) to a 40V DC-DC boost convertor and drove it with a mosfet gated by an Arduino Nano at 40kHz. I then tacked a wire to C31 and measured received signals.  I'm indoors in a hallway. I am encouraged but I'm out of time. Do you have something else I can try? Quickly?

Thanks in advance,

JD

I

Hi Jdeca_727,

So you're using the reference design as it's intended to be used, sort of. It is meant to be tested as-is to verify baseline performance, then picked apart and adapted to your needs if necessary. But that takes time, thorough testing under worst-case scenarios, and you really need to understand the fundamental principles behind the circuit, including the specific transducers - their gain, their sensitivity vs. angle, and probably the most important thing - their bandwidth - narrow bandwidth is good from the perspective of rejecting ambient noise (what if someone blows a dog whistle? What if there are bats in the area?)

As much as we want to sell Analog Devices parts - does something like this exist off the shelf? Optical solutions certainly exist:

https://new.abb.com/low-voltage/products/safety-products/optical-safety-devices (There are lots of lower cost options)

And a quick search on "ultrasonic data link" turns up some ideas, but it does look like you might have to do some exploration on your own.

Hope this helps a bit - but it does sound like you're doing all you can with the CN0343.

Oh one other thought - the reason I had searched on "data link" is that you could use some simple scheme to reject interference - like only accept received pulses of a certain width, or something more elaborate. But again, this is inventing something that's probably been done before.

-Mark