PDA

View Full Version : [News/Announcement] Oh noes! Sharp IR sensors now obsolete!



jwatte
02-18-2013, 09:28 PM
I just got a message from Digi-Key that Sharp is retiring their IR proximity sensor line. True to form, searching for IR distance sensors now just returns a bunch of non-stock items.
I imagine other suppliers will run out at some point, too. Does anyone know of an alternative that would work well in the future?

tician
02-19-2013, 03:42 PM
I had kind of wondered if that was going to be the case given how few were in stock a couple months ago. Part of me is really hoping they are just upgrading them instead of simply obsoleting the whole division/line (they are a pretty old design). It looks like the simple 5/10cm 1-bit digital sensors are still in production same as their many miniature proximity/ambient light sensors, so one can hope.

I decided to experiment with building a lidar partly because of the slow update rate of the analog output of the sharp sensors, but mostly to see if I could (still do not have it working - lazy, lazy, lazy - might just start over with a CM-530 instead of trying to debug my ATmega328 code).

jwatte
02-19-2013, 05:10 PM
Right -- the gap between a $10 "proximity" sensor and a $1000 LIDAR really needs filling with something other than ultrasonic sensors vulnerable to wind.

tician
02-19-2013, 08:25 PM
Parallax makes a $130 laser ranger using a laser pointer, a cheap camera cube CMOS sensor, and propeller IC. Unfortunately, it's only good for 6-48 inches at 1Hz output, and I'm not really sure about accuracy/precision. Really hope I get around to finishing my lidar attempt; it would be really nice to have done something useful for once.

DresnerRobotics
02-21-2013, 03:03 PM
This has been a long time coming. We got notice that they were discontinuing them in the US 2+ years ago.

With that said, we're working with Sharp & Digikey to reach a compromise, and as of now it's looking like we will have them available still.

tician
04-28-2014, 11:53 PM
So, are the GP2Y0E series really new products, or did I just not notice them until now. The 2014-03 lineup lists them as new; I can only find seven references to the series on the sharp website; and the only app-note/datasheets I can find list 2013 as the pdf's year of release. They are only good for 4~50 cm, but use I2C and/or analog with a purely linear distance function. I2C directly reports the computed distance, and the analog output appears truncated so there is no longer the 'two distances at one voltage' issue of the older PSD-based GP2Y0A series products.

Position Sensitive Device - basically a big photodiode with a common anode and two cathodes on opposing edges, and the two photocurrents give relative position of incident light. Was planning on using one iC-Haus ODL2C PSD for that laser range finder I never finished.

jwatte
04-29-2014, 12:08 AM
Those are new. And I really wish they'd come out with a 3 meter version, that works in sunlight :-)

DangerousThing
04-29-2014, 01:31 AM
It's a good thing I have a bunch of them from 10 years ago. :)

Though I was really hoping that they would have come out with more interesting and easier to use versions by now.

_ADAM_
05-06-2016, 04:30 PM
Is there a reason to use IR over ultrasonic? It was a long time ago, but I did some tests of a Sharp IR vs a Ping, and it wasn't even close, the Ping was way more reliable and accurate. I would actually say the Ping worked perfectly, even detecting narrow objects like a pen. I never got good readings from the IR, turned me off to them. This was the old Sharp GP2 that Lynxmotion sold.

tician
05-06-2016, 05:12 PM
Sonar usually have a rather wide forward view angle, so the sensors tend to have trouble differentiating between objects separated by a small angular distance from the sensor (i.e. one object [email protected] and another object [email protected]). They can also be affected by air currents and air temperature and may not sense certain types of objects like curtains that do not reflect sound well.

Triangulation based IR sensors, like those from Sharp, can get in trouble when the room is saturated with light and/or the reflectivity of the object is kinda low, but usually have a much narrower forward view angle permitting better differentiation. Time of flight and phase-shift/interferometer sensors are even better than triangulation as they are really only hindered by severe interference from external light sources and very, very low reflectivity surfaces (like windows).

jwatte
05-06-2016, 05:14 PM
Is there a reason to use IR over ultrasonic?

My experience: Ultrasonic is great when it works.

Drawbacks of Ultrasonic:
- Higher latency when measuring far away.
- Sensitive to wind, and certain kinds of environmental noise.
- If you want multiple separate readings, you have to time-slice (running many at once can see crosstalk).
- The MCU reading code is slightly more complicated, because measuring the duration of a pulse required a controlled "start pulse" and then two interrupts to time the interval. This is more cumbersome than reading an ADC.