Has anyone successfully implemented object detection using IR for a range greater than 5cm. I am building a robot project from scratch with the personal challenge of making all parts homemade. The brainboard and motor controller were succesfully built but now that i want to do object detection using IR (without getting the sharpe module) I have issues with ambient lighting and range.
I am trying to draw the schematics to add so i can show what i have done so far
Take an IR LED, modulate it
Take an IR LED, modulate it at 38khz, then use a 38khz reciever, you should be able to reach out a foot or so.
I tried that using a 555
I tried that using a 555 timer to do the osicilation but still I am only getting a few cm
yup
I have done 10cm with just 2 IR leds and one IR diode (regular 2-pin). I have gotten up to 50-60cm when modulated to 38khz with a (3-pin) 38khz sensor.
You just need some tweaking, I think. If you can get a hold of an O-scope, it would make you life a LOT easier.
O-Scopes don’t come cheap
O-Scopes don’t come cheap plus i was even trying to simulate the 38khz using the 555 to also skip the 3 pin sensor. this portion of course was until I could soruce the 3pin since solarobtics no longer offers and I am still nervous to use any other supplier to date. Been at this for a while now and now it seems I may have to give up on that challenge to use all homemade modules
Youre doing it wrong, the
Youre doing it wrong, the timing must have been sloppy or the angle isnt right.
Timing is everything
The receiver modules want to see a 50/50 duty cycle on the incomming 38khz to operate at peak efficiency. I have actually used this same basic circuit as a rangefinder accurate out to better than 6’ using the Sony version of the 38Khz IR receiver module.
There is a fairly linear relationship between the time the pulse is sent and when the module “sees” it. You can either create a “lookup table” to translate this delay into distance, or you can gate an oscillator and count the pulses, or, if you don’t need super accuracy, you can configure an L/C circuit and let an ADC pin on your board of choice convert it to a value.
The basic theory of operation goes like this: The receiver module needs approximately 3 good cycles of 38Khz, (or whatever frequency of module you’re using), to latch and produce an output. So, let’s say at distance zero, every cycle makes it to the receiver module, so there is no L/H transition. The farther the target is, the longer it takes to get 3 good pulses to latch the receiver and the more time has elapsed between the send pulse and the receiver output.
I did my research on this in the 90’s when I was working at the University of California. I’m sure, with the current availability of higher output IR LEDS, the range could probably be doubled. It was a nifty little circuit for the time and could be built in quanity for under $5.00.
I am currently building a new robot employing both Polaroid ultrasonic rangefinders and this IR circuit to produce an “averaged” output for better accuracy regardless of the surroundings.
Remember my personal challenge was everything from scratch
Salvage, I am not even using a standard 3 pin 38khz reciever.
What I did was to setup an astable circuit using a 555 timer producting a square wave at a duty cycle 50% (maybe it was sloppy). This pulsed on and off an IR LED. Just beside this IR LED I had a IR clear diode ( the pair came in a radio shack package) connected as reverse biasas acrosss the collector and base pins of a 3904 transistor. At the emitter end of the transistor I had a regular led showing signal when detected. Works beautifully for roughly 5cm at nights but during day time nothing. The only thing I can think of is that i fried teh 104 capacitor when soldering it in place (ceramic) . I must also point out that the detector section of the circuit works perfect when i aim my tv remote at it an press any key.
break out the multimeter and
break out the multimeter and measure volts and current across the emitter.
