Contents

Expectations

This is the third release of this form-factor from Devantech so our expectations are somewhat established compared to a completely new product.  The other, previous detectors are fantastic so this new version has some living up to high standards in front of it. 

We understand this detector to be an easier to manufacture alternative to the SRF04 with some improvements.  Since it has similar ultrasonic transducers, we don't expect the overall sonar cone pattern to change much.  It will be interesting to see if the range is different. 

Out of the Box

The sensor shares the same form factor as the SRF04 as we mentioned.  A couple of obvious changes are a small LED on the back and another wire hole.  Looking over the initial SRF05 documentation reveals that there is a new, optional one-wire interface that we will definitely have to check out.  This new mode could reduce the I/O pins by half in a sensor laden robot!

Test Rig

We used a BrainStem GP 1.0 to test out the SRF05 in both modes.  Since we had a SRF04 library already written, we just copied this and updated the names to test mode 1 which is exactly the same as the SRF04's INIT/ECHO interface with 2 digital I/O pins.  This library will go in our standard BrainStem software distribution so you can get it with the next release (build 23 or greater). 

This cloned SRF04 library worked straight away and the LED blinks each time it fires the sonar.  Now, we like blinking lights and this is no exception.  This helps when diagnosing a new interface to make sure your code is making it to the sensor. 

The wrapper program is just a simple main loop that takes a reading, displays the results in the console window, and then delays before repeating the sequence.  This is the same as the SRF04/BrainStem example on our website. 

Next, we tried out the one-wire interface which is really slick.  With mode 2, you use a single pin to first toggle the sensor (with your microcontroller pin set up as a digital output).  The pulse needs to be at least 10ms in length.  Next, you flip the same digital I/O pin to an input and measure pulse time on it to determine the distance.  Since the ouptput pulse on this pin is proportional to distance, you get the distance and initiation from one pin! Pretty fancy. 

We wrote a BrainStem library routine that manages all this digital I/O pin flipping and timing of the pulse.  The result is you can now put a different SRF05 on each Digital I/O pin and read all of them.  This is a great win if you are trying to conserve pins.  This library routine is also in the new aSRF05 TEA library for the BrainStem and it will work with any BrainStem Digital I/O pin.  This TEA library will ship with all post build 22 distributions. 

Performance

The sensor seems to behave in all respects just like the SRF04 in mode 1 (with the added bonus of a blinking LED).  No audible ping is heard and initial impressions showed an improved range out to around 4 meters (from the SRF04's 3 meters). 

In mode 2, which only uses 1 pin, the response seems identical, with the added benefit of halving the required pin count.  It took some time to get the library right for this mode, but that work is only done once and can be re-used. 

Conclusion

As we have come to expect and appreciate, Gerry has once again outdone himself with an all-around better design and yet a great price.  This is most definitely a product we will include in our product line and in our contract robot design work as well. 

Related Examples:

Devantech SRF05 Interface to BrainStem GP 1.0 Example