SRF08 vs. SRF10 => any performance differences??

I recently tried wall following with the SRF10. Also tried a SRF04. It was in an internally closed arena, about 1m x 2m. I found it too difficult (I was on a time line). I went to GP2D120 instead. The SRF10 was extraordinarily wide beamed, and both were prone to odd readings when the angle of incidence was high.

We did a side by side comparison with dpa's robot using the Polaroid type sensors, and his showed none of these problems.

-- Randy M. Dumse

Caution: Objects in mirror are more confused than they appear.

On my hexapod that went to Seattle-Robothon, I used 3 sonars for navigating the walker contest course. Originally, I tested an SRF04 unit, but discovered it has really nasty sidelobe pickup on the receiver end of the device. People on yahoo-SRS confirmed this, and recommended mounting the device vertically, with the receiver sidelobe pointing up into the air, where pickup isn't so critical.

Sidelobe pickup would have totally killed the ability of the walker to turn at the wall and point correctly for home. In any event, I scuttled the SRF04 and went to 3 Parallax Ping units, and these seem much better behaved. I do plan to go back and do some more testing of the devantech units one of these days, before reaching a "final" conclusion on these devices.

Also, a couple of years ago, I completely stopped using the GP2xx devices. Problem is, they take about 50-msec to produce an updated output, during which time they appear to average multiple triangulation readings. If the bot is moving during this time, especially panning around, then you can get extrememly weird results. Think about it - geometrical triangulation is extremely sensitive to changes.