Longer(er) Distance Infrared Transmission

SRLM · 2009-03-31 05:23

I've been working towards building a simple beacon to track object, and decided to use an infrared LED (here or here, I used them alternately) on the beacon and matching receiver (here) on the mobile module.

Well, I got to do some testing tonight, and it didn't go as well as I hoped. I used two different LEDs that were suitable (correct wavelength) with similar results. I could only get reliable data out to about three feet, then it started to fall apart until at about 16 feet nothing more was sensed. I think (big step here) that the problem is weak LEDs. I viewed the LED through my digital camera, and I could barely make it out at about 5 feet, so it's unlikely that the IR detector could see it from there either. Granted, I was running at about 60 ma (out of 100 ma max), but I had hoped for better results of 20 feet at least.

The circuit that I used was a Darlington array (at 5v) pulsed by the Propeller. The Led has two 20 ohm resistors in series.

I'll probably give these high power LEDs a shot since they should be about 5 times as powerful if I run them at full tilt (300 ma).

I also tried shielding the receiver with a long tube, but it didn't seem to help (trouble aiming and holding probably contributed some).

So, the other option is a laser. I was thinking about getting the appropriate laser with a lens to make the dot a bar, but I'm wary of that option due to the danger (no blink reflex for IR). But with high power LEDs is it any different? I don't know...

Any suggestions?

Post Edited (SRLM) : 3/31/2009 5:41:00 AM GMT

Phil Pilgrim (PhiPi) · 2009-03-31 05:37

You should have gotten well over three feet. The detector you've cited operates at 56KHz. What did you modulate your IREDs with? What was your data rate?

-Phil

SRLM · 2009-03-31 05:45

The Spin transmitter program

CON
    _clkmode = xtal1 + pll16x
    _xinfreq = 5_000_000
OBJ
    BS2 : "BS2_Functions"    ' Create BS2 Object
PUB Start 
    BS2.start (31,30)        ' Initialize BS2 Object timing, Rx and Tx pins for DEBUG
    repeat
      BS2.FREQOUT(23, 1000, 56000) 'Freqout one second at 56 kHz
      waitcnt(clkfreq + cnt) 'Wait one second

And the BS2 receiver program

'   {$STAMP BS2}
'   {$PBASIC 2.5}

TX              PIN     12               ' serial output to LCD
T9600           CON     84
LcdBaud         CON     T9600
LcdCls          CON     $0C             ' clear LCD (use PAUSE 5 after)
LcdBLon         CON     $11             ' backlight on
LcdOn1          CON     $16             ' LCD on; cursor off, blink off
LcdLine1        CON     $80             ' move to line 1, column 0
LcdLine2        CON     $94             ' move to line 2, column 0

hightime VAR Word
lowtime VAR Word

HIGH TX                               ' setup serial output pin
DIR7 = 0
PAUSE 100                             ' allow LCD to initialize

'Prepare LCD (clear it, backlight, turn off cursor)
SEROUT TX, LcdBaud, [noparse][[/noparse]LcdCls]
PAUSE 10
SEROUT TX, LcdBaud, [noparse][[/noparse]LcdBLon]
SEROUT TX, LcdBaud, [noparse][[/noparse]LcdOn1]


DO
 'Find high time
  DO WHILE IN7 = 1 AND hightime < 1500
    hightime = hightime + 1
  LOOP
  'Display high time
  SEROUT TX, LcdBaud, [noparse][[/noparse]LcdLine1,"H:", DEC5 hightime]
  hightime = 0
  'Find low time
  DO WHILE IN7 = 0 AND lowtime < 1500
    lowtime = lowtime + 1
  LOOP
  'Display low time
  SEROUT TX, LcdBaud, [noparse][[/noparse]LcdLine2,"L:", DEC5 lowtime]
  lowtime = 0
LOOP

Basically, what my program does is to just 'toggle' the IR receiver alternately low and high. I know success by when the high and low times on my LCD match up, and failure when they are different. I've pulled out my Parallax oscilloscope and tested to make sure that I am indeed transmitting at 56kHz, and I am. One oddity that I noticed is that the oscilloscope software shows a peak to peak hight of <150 mV. It's odd, but I don't know what it means. I can't use a multimeter to check, since the meter is to slow (with the changes).

I just tried transmitting with the second led (LED55CF) that says its rated for 10A pulsed. When I tried (with about 500 mA) by removing my resistors, it got really hot. I'm afraid that my lack of practical circuit knowledge is showing through...

Post Edited (SRLM) : 3/31/2009 6:12:00 AM GMT

Phil Pilgrim (PhiPi) · 2009-03-31 06:30

The problem is likely your Darlington driver. Darlingtons switch too slowly to be used in this fashion. In your case, it's probably spending most of its time in its linear region, which would account for both the 150mV P-P readng and the overheating. Use a MOSFET instead. For interfacing with a Propeller, an IRF3708 should work well.

-Phil

Post Edited (Phil Pilgrim (PhiPi)) : 3/31/2009 7:09:59 AM GMT

SRLM · 2009-03-31 14:39

Would a circuit like this work? I'd replace the switch with the Propeller output and a series resistor (1K or so), and remove the R1 pulldown.

Phil Pilgrim (PhiPi) · 2009-03-31 15:35

That should work. Do not remove the pulldown, though. It will keep the transistor "off" during reset. I'd use a lower value for it, though — maybe 100K. Also, select a value for R2 that will yield the current you need for the IRED. (You can pretty much ignore V_DS in your calculation, since it will be so low.)

-Phil

Addedum: The transistor I recommended has a fairly high gate capacitance (2.4nF). You may want to add a small series resistor between the Prop and the gate to limit the transient pin current. 75 ohms will help and will still gve you a sub-microsecond rise time.

Post Edited (Phil Pilgrim (PhiPi)) : 3/31/2009 3:56:20 PM GMT

CannibalRobotics · 2009-04-02 02:13

What's the distance from the object, how fast is it moving and the desired accuracy and frequency of tracking?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Signature space for rent!
Send $1 to CannibalRobotics.com.

Longer(er) Distance Infrared Transmission

Comments