IR sensor tuning?

[MikeK]

I was at a line-following robot competition recently and one of the contestants mentioned that he tuned the IR detectors to match the emitters. I asked if he meant that he was pulsing the emitters, but he said no.

I know about tuning for a carrier frequency, like 38kHz, but had never heard of tuning for the light frequency. What was he doing?

[Soeren]

Hi,

I was at a line-following robot competition recently and one of the contestants mentioned that he tuned the IR detectors to match the emitters. I asked if he meant that he was pulsing the emitters, but he said no.

I know about tuning for a carrier frequency, like 38kHz, but had never heard of tuning for the light frequency. What was he doing?

Hard to say, but a few possibilities might be:
- He was a lying rat 
- He made sure each emitter and detector pair pointed to the same spot.
- He made sure that both emitters and detectors had the same angle of half intensity.
- He bought IR LEDs with a wavelength matching the peak sensitivity of his detectors.
Either one (or more) of the above, or nothing at all. While you can change the wavelength of an IR LED to a minor degree by pushing it, only a moron would try that stunt and the only way of determining the result would be to compare amount of received light (unless he had access to a pretty accurate spectroscope).

Personally, I think he was either yanking your chain or trying to make himself sound like a genius by ridiculous and wrong techno babble.

I have met quite a number of such persons over the years - frightening as it sound, even in a Hi-Fi shop that's trying to portray itself as the über audiophile shop in the country - way above any other - where a sales-guy, when I asked him about the power output of an amp I wanted to buy, claimed that it was impossible to say, as it depended on the speakers, the cable (and quite possibly the color of your wallpaper).
Even telling the guy that I was an electronics engineer and have designed and built several amps over the years and asking him to cut the consumer B.S. and find some documentation didn't rock him a bit, Ohms law didn't bite either, what a moron... Then he found a boxed one for me and when I opened the box, the power output was in big red letters on the front of the user manual (as well as in the specs section of course) - That sort of knocked the I'm-so-bright out of him and he could barely look at me when he made up the receipt and took the money.
All that "fun" just because we were having a double round birthday party that year (40+50) and I hadn't got time to fetch a working amp for the party, so had to throw money at it (in between preparing for 120 guests arriving two hours later.

The morale of that - If you don't believe what people tell you, or if you know that what they say is impossible - ask for an explanation and if you don't get it, write it off as hot air. And always make sure, in good time, that the amp you are gonna use is working.

In case you meet that person again, do ask him what he meant, so it won't puzzle you any more.

[MikeK]

Yeah, I supposed.  He didn't seem like the bragging type, though.  I've talked to him before...he builds the robot with his young daughter.

If he was just mistaken and was actually pulsing the emitters at 38kHz, You only get a signal or not, right?...You don't get a reception intensity, like when you read the analog signal strength.

[Soeren]

Hi,

If he was just mistaken and was actually pulsing the emitters at 38kHz, You only get a signal or not, right?...You don't get a reception intensity, like when you read the analog signal strength.

Intensity will change some with duty cycle. Not proportional though and peak intensity will usually be at above 50%, but depends on the sensitivity and speed of the sensor (and somewhat on the ambient parameters).

Intensity will change with distance and reflectance of the target as well, but if integrated (TV type) receivers are used, you won't be able to get that info, as it's just used to control the built-in AGC.
With circuits like the IR receivers from around the late seventies (eg, TDA4050) you could access the AGC voltage, but you needed a stockpile of resistors and caps to filter to whatever carrier, so for most purposes, the integrated receivers are a huge step up.

If you need something very sensitive or with other special characteristics, you need to make a discrete build that address the parameters needed.


BTW. Why try to tune anything for a line detector, which must be considered pretty close range and unproblematic?

[MikeK]

I didn't want to tune anything, I just wanted to see if he was using a better sensor system than me. 

Right now I've got 8 QRB1134's, which I may change in favor of separate emitters and detectors, which is what he was using.  The packaged QRB's allow only a narrow reflection range.  And this causes a problem since the white-line course is outside on a concrete path and the sensors bounce up and down as the robot moves.