Post by thread:[EE] Multi-Touch Sensing through LED Matrix Displa

File this idea away for future reference. Use of a multi-LED display as a touch sensor simultaneously with its use as a display. Off LEDs are used as sensors of adjacent on LEDs with light paths being interrupted by fingers interrupting the light paths. Interestingly, cites a 1977 Forrest Mims comment. http://mrl.nyu.edu/~jhan/ledtouch/index.html Presumably this could be done as part of multiplexing a display with no visual effect on displayed information. Cries out for interactive display. Analog sensing possible. Use of IR LEDS (perhaps mixed with visible ones) could allow visually dark sensor. ____________ Upper level is a personal research page from Jeff Han at NYU. Some interesting material. http://mrl.nyu.edu/~jhan/ Ref: Matthew McMahon

Hi, this use of leds as sensor remember to me an Application related to Leds used as Optical Sensors (instead of a Photodiode) directly connected to a PIC. See this: http://www.merl.com/papers/docs/TR2003-150.pdf Some months ago I made a little circuit similar to this (on a 12F629 instead of 16F876) for use as an inexpensive photometer. Interesting results, if "cheap" is the first focus. regards Marco ---------- Initial Header ----------- >From : spam_OUTpiclist-bouncesTakeThisOuTmit.edu To : "PIC List" .....PICLISTKILLspam@spam@mit.edu Cc : Date : Thu, 13 Apr 2006 11:44:35 +1200 Subject : [EE] Multi-Touch Sensing through LED Matrix Displays {Quote hidden}> File this idea away for future reference. > Use of a multi-LED display as a touch sensor simultaneously with its > use as a display. > Off LEDs are used as sensors of adjacent on LEDs with light paths > being interrupted by fingers interrupting the light paths. > > Interestingly, cites a 1977 Forrest Mims comment. > > http://mrl.nyu.edu/~jhan/ledtouch/index.html > > Presumably this could be done as part of multiplexing a display with > no visual effect on displayed information. > Cries out for interactive display. > Analog sensing possible. > Use of IR LEDS (perhaps mixed with visible ones) could allow visually > dark sensor. > > ____________ > > > Upper level is a personal research page from Jeff Han at NYU. > Some interesting material. > > http://mrl.nyu.edu/~jhan/ > > > > > > Ref: Matthew McMahon > > > > > --

The only limitation is that it requires quite a bit of extra circuitry, and only works well in dark environments. The circuitry problem could probably be dealt with (simplified). The ambient light is going to be a more difficult problem (sense while no LEDs are on, sense while adjacent LEDs are on, detect difference, etc). -Adam On 4/12/06, Russell McMahon <apptechKILLspamparadise.net.nz> wrote: {Quote hidden}> File this idea away for future reference. > Use of a multi-LED display as a touch sensor simultaneously with its > use as a display. > Off LEDs are used as sensors of adjacent on LEDs with light paths > being interrupted by fingers interrupting the light paths. > > Interestingly, cites a 1977 Forrest Mims comment. > > http://mrl.nyu.edu/~jhan/ledtouch/index.html > > Presumably this could be done as part of multiplexing a display with > no visual effect on displayed information. > Cries out for interactive display. > Analog sensing possible. > Use of IR LEDS (perhaps mixed with visible ones) could allow visually > dark sensor. > > ____________ > > > Upper level is a personal research page from Jeff Han at NYU. > Some interesting material. > > http://mrl.nyu.edu/~jhan/ > > > > > > Ref: Matthew McMahon > > > > > -

More on LEDs as cheap sensors: http://www.merl.com/papers/docs/TR2005-097.pdf A low-cost chemical sensor with two SMD Leds and a PIC, used for colorimetric based gas detection. regards Marco ---------- Initial Header ----------- >From : .....piclist-bouncesKILLspam.....mit.edu To : "piclist" EraseMEpiclistspam_OUTTakeThisOuTmit.edu Cc : Date : Thu, 13 Apr 2006 14:11:31 +0200 Subject : Re:[EE] Multi-Touch Sensing through LED Matrix Displays {Quote hidden}> > Hi, > this use of leds as sensor remember to me an Application related to Leds used as Optical Sensors (instead of a Photodiode) directly connected to a PIC. > > See this: http://www.merl.com/papers/docs/TR2003-150.pdf > > Some months ago I made a little circuit similar to this (on a 12F629 instead of 16F876) for use as an inexpensive photometer. Interesting results, if "cheap" is the first focus. >

On Apr 13, 2006, at 6:21 AM, M. Adam Davis wrote: > The only limitation is that it requires quite a bit of extra > circuitry, and only works well in dark environments. To some extent, you should be able to sense shadows instead of reflected light, in a "bright" environment. There's been some discussion of the concept in general on the BEAM group at yahoo. Recent link: http://projects.dimension-x.net/technology-and-projects/ledsensors/ BillW

> The only limitation is that it requires quite a bit of extra > circuitry, and only works well in dark environments. Dark areas maybe, but I don't think it actually takes any extra circuitry at all... From http://www.merl.com/papers/docs/TR2003-150.pdf "Typically, with a photocurrent of 50 pA in normal ambient Laboratory lighting (LED off), and a capacitance (LED) of 20 pF, the voltage decay rate is approximately 2.5 V/s. Energizing the exciter LED increases the photocurrent typically by an order of magnitude. By means of a small program executing in the PIC 16F876 microcontroller we can accurately measure the time required for this discharge, and hence measure the diode photocurrent, thus indirectly measuring the incoming illumination on the LED itself. This obviates the need to directly measure these extremely small photocurrents using expensive instrumentations. By alternately measuring the decay times with the exciter LED on and off, we can make a differential measurement and compensate for the effects of ambient lighting. As implemented, the base sample rate is 14 Hz., alternating between measurements taken with the emitter LED on and off for differential measurement. Hence, the signal is measured in time units, typically microseconds." Even a chip as slow as a PIC can do it, so an SX should be able to see the most minute variations in light level. I wonder if this could be used with some of the matrix wiring systems e.g. http://www.piclist.com/techref/io/led/8x7s8pin.htm or http://www.piclist.com/techref/io/led/ucbimatrix.htm --- James Newton: PICList webmaster/Admin jamesnewtonspam_OUTpiclist.com 1-619-652-0593 phone http://www.piclist.com/member/JMN-EFP-786 PIC/PICList FAQ: http://www.piclist.com