Decoding an ultrasonic parking aid unit.

Hi,

Rcecently, I purchased a car reversing/parking aid like this http://www.ebay.in/itm/ws/eBayISAPI....:PPOPN:IN:1123

It consists of 3 components; the display unit (gives the distance to the obstacle in meters, and on an LED bar), the 4 ultrasonic sensors and the main control unit into which you plug the display, and the sensors.

The signal from the main controller to the display is carried over a single wire serially. So, it carries data for all 4 sensors.

I presume it would have a sensor label/ID and the reading.

Each of these packets looks like the screen grab I got by using the Pickit2 logic tool. And, I believe each would have adresses AND the distance data for all 4 senors.

Now, neither the bit width, nor the space between the bits is constant. So, I couldnt use the Pulsin to grab the data. I thought an

RC5 / Manchester kind of an algorithm would work. BUT, the problem is, the minimum bit width, which is typically 100us, in reality varies between 100 to 140 us within the same frame, for different bits. As a result of which, I do not get consistent data, even when there are no sensors attached.

Has anybody played around with these before? Any suggestions on how to decode this packet, despite the variance in the bit widths?

I have seen some hack articles for parking units on the internet, but the most elaborate one I saw used a different protocol altogether.

Regards,

Anand
Attachment 6822

Okay, some more work done shows I was wrong about the data format. Each frame in fact contains the sensorID (0 to 3) and the reading indicating distance to object. Works quite well, easy enough to decode.

My question is, can this type of an ultrasonic trasducer ( https://www.google.co.in/search?q=au...ient=firefox-a ) be used in an 24x7 application? In its native purpose, it would be used just a couple seconds each day. Would the transducer degrade/fail if its pinging all the time?

That's a question best asked of the manufacturer.