Frequency Sweep using IR: How does it work?

Domanik

I was stepping through some IR distance detection tutorials and found something that puzzles me. How does increasing the frequency from 38.5KHz to 41KHz enable the IR detector be used for distance detection? Range is limited to a few cm but how?

The link is here: http://learn.parallax.com/node/335

I suspect it has something to do with the bandpass characteristic of the filter used in the detector PNA4602M, shown in the link. 38KHz has a period of 26us, which I think eliminates round trip delay. Other IR detectors triangulate for distance but that is not the case here. I read through the datasheet without seeing anything.

Does anyone have some insight into this?

Thanks,
Dom...

Sapphire

It detunes the receiver when the IR pulse frequency is changed. So you have to be closer to get enough signal through the receiver's filter to get a logic change on the output. Works pretty well actually. I've been able to change the detection distance by 3 feet using this trick.

Domanik

Still don't get it. By going off the center freq to 41KHz you slide down the sensitivity curve from 100% to 35% according to the datasheet spec. Okay, the filter has less signal to work with but that could be due to color or angle of reflection, not only distance. I don't see how signal strength can reliably be translated to distance. Dom..

Sapphire

I think the key word is "reliably". It can't. It's just a crude way of using different frequencies to approximate distance. It is dependent on color, angle and reflectivity, as you surmised.

Domanik

Seems like there is more going on in the circuit that needs to be understood in order to understand how it's measuring distance. After all it has an AGC, bandpass, demodulator, integrator, comparator, etc.

Somehow, in a close proximity, it will switch on the measure distance capability as the frequency goes off from 38KHz. The different frequencies picked up by the sensor are 940nm, 38KHz and 1ms.

Does anyone have a link to a good description of this? Dom...

Beau Schwabe

Think of it like a flashlight or a Star...... The Brighter the Star the further away (distance) you can detect the light, the Dimmer the Star, the detectable distance diminishes. Likewise if you are emitting a frequency from the IR that is closer to the "tuned" frequency of the IR receiver your IR transmitter is going to appear "brighter". If your IR transmitter is de-tuned, then it won't appear as bright to the IR receiver.

Domanik

Beau Schwabe (Parallax) wrote: »

Think of it like a flashlight or a Star...... The Brighter the Star the further away (distance) you can detect the light, the Dimmer the Star, the detectable distance diminishes. Likewise if you are emitting a frequency from the IR that is closer to the "tuned" frequency of the IR receiver your IR transmitter is going to appear "brighter". If your IR transmitter is de-tuned, then it won't appear as bright to the IR receiver.

Beau,
I completely understand and agree with what you're saying about stars and flashlights. The reflected light from the LED does not decrease with frequency as it enters the detector.

However other factors can enter into the intensity of the reflected IR signal, like color or texture. To counter-act this the internal AGC should prop up the signal to a fixed level inside the detector regardless of why it's attenuated. The AGC makes sure the rest of the circuit has enough signal to work with, but this happens before the BPF.

I'm sure the reflected signal is NOT attenuated in the least by changing the frequency. It will decrease as distance increases because the LED light spreads out over a larger area. Perhaps the AGC is out of the circuit when the reflected light is bright (due to small distance) and the Amp output is larger than normal. Therefore the bandpass passes a larger signal than normal but attenuates it when it's off frequency. As the combination of the DC signal due to the distance/bandpass product crosses the comparator threshold it provides something corresponding to distance. When a normal signal level from a normal distance comes into the detector it will amplify it (due to the AGC) as needed before feeding it to the bandpass filter puts out a fixed AC level and the demodulator puts out a fixed DC to the comparator. What think?

Dom..

Edit:
Second Thoughts: I think I understand what's going on now. Thank you for helping me with this.