souped up ultrasonic range finder

ChargedUp

I'm working on a souped up ultrasonic range finder (40kHz). The transmitter and receiver are located 0.75 inch apart. The problem is crosstalk directly between the transmitter and receiver, see scope trace 1 (no target). The crosstalk signal hides the target detection, see scope trace 2 ( target at 1 inch) and scope trace 3 ( target at 2 inches). I would like to use a Javelin stamp. Does anyone have a software or hardware suggestion to separate the unwanted crosstalk signal from the target detection?

Phil Pilgrim (PhiPi)

Record the received (crosstalk) signal when there is no target in range. Then subtract that signal from subsequent receptions. (This is assuming you're using a fast A/D to capture the signal each time.) If you can maintain the same phase relationship, you should be left with only the echo. That said, however, ranging items within two inches is probably asking way too much from such a setup.

-Phil

Post Edited (Phil Pilgrim (PhiPi)) : 5/9/2008 5:11:12 AM GMT

ChargedUp

Phil Pilgrim (PhiPi) said...
Record the received (crosstalk) signal when there is no target in range. Then subtract that signal from subsequent receptions. (This is assuming you're using a fast A/D to capture the signal each time.) If you can maintain the same phase relationship, you should be left with only the echo. That said, however, ranging items within two inches is probably asking way too much from such a setup.

-Phil

Thanks Phil
You suggest recording the acoustic crosstalk and using a fast A/D. Is the Javelin Stamp or a PIC @ 20mHz fast enough?

The Ping Ultrasonic Sensor is capable of 0.75” to 10'. I don't know how they eliminate the acoustic crosstalk but they only drive the transducers at ±5V, I hope to use ±10V.

Beau Schwabe

ChargedUp,

Typically after you PING the ultrasonic transmitter, you wait for the echo to be received after the transmitter has turned off. This way, the receiver doesn't "see" the transmitter's initial signal.

If your transmitter is in continuous mode, then the phase relation between the transmitter and receiver is one way to look at the signal. Even if the transmitter and receiver are locked together on the same board, be aware, that ANY change in orientation between the transmitter, receiver, or it's surroundings will cause a phase change from then received reflections. Even a hand waving will cause beat-frequency disruptions from reflections - makes for a good motion detector.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

jmsiino

ChargedUp,
How did you supe-up your ping? I'd like to do the same.

ChargedUp

jmsiino said...
ChargedUp,
How did you supe-up your ping? I'd like to do the same.

Hi jmsiino

Th PING transmitter operates at +5v & -5v I raised this to +10v &-10v. This change increased both the sensitivity and the range. Unfortunately, it also increased the crosstalk directly between the transmitter and receiver. I haven't been able to solve the crosstalk problem.

ChargedUp

jmsiino

what was the maximum (increased) range? I have an application where I would like to be able to measure up to 25 feet.

I'll try boosting the voltage on my ping one volt at a time.

Is the cross-talk electronic or sonic? Do you need to filter the power to the receiver or do you need physically dampen the metal cylinders against ringing?
.

jmsiino

I just reread your original post. I originally thought you had two rangers and they were interfering with each other.

Your metal cylinders may be ringing/resonating.

First: Try wrapping a rubberband around the board to dampen the ringing of the board.

Second: Try wrapping a rubberband around one of the cylinder to dampen the ringing.

Third:. Try wrapping a rubberband around the other cylinder to dampen the ringing.

Fourth: Try wrapping a rubberband around each of the cylinders to dampen the ringing.

ChargedUp

Thanks for your thoughts

I blocked the receiver with my thumb - no crosstalk signal. I blocked the transmitter with my thumb - no crosstalk signal.
I shorted across the receiver – no crosstalk signal. I disconnected the receiver - no crosstalk signal.

Conclusion the problem is acoustic crosstalk.

I tried:
felt pads between the transducers and the circuit board
cotton balls between the transducers
paper, metal, rubber and plastic as dividers and cones around and between the transducers

None of these techniques were more then marginally effective

ChargedUp

jmsiino

The trace of the output "ping" signal appears to have gotten longer. Is it possible that the increased voltage has lengthened the duration of the output "ping"?

ChargedUp

jmsiino said...
The trace of the output "ping" signal appears to have gotten longer. Is it possible that the increased voltage has lengthened the duration of the output "ping"?

Thanks for the thought.
I rechecked, the transducer is sending 8 pulses at 40kHz and each pulse swings from +10V to -10V.