Hi,
In a lap timer system I thought of equipping cars/boats/drones with IR LEDs, PWModulated at 30, 33, 36, 38, 40 and 56 kHz to match Vishay TSOP22.., TSOP24.., TSOP48.. and TSOP44.. series IR receiver modules. So a maximum of six contenders. At the start-/finish line I have the TSOP receivers, and the receiver that is triggered will tell which vehicle passed.

Is the separation of 2-3kHz between the modules enough to not cause multiple triggering from one contender (triggering modules on neighboring frequencies)?

From the datasheet of the modules I can see this:


If I understand this correctly, if I for example divide 36kHz by 38kHz I get 0.947, which is about 0.7 relative responsivity. But at which relative responsivity will I have no false triggering from neighboring "channels"?

Are these modules designed to co-exist without interference from each others remotes?

 

I believe your understanding is correct.
You are almost certain to get interference if the signal is only rolled off by 0.7 (-3dB).
So the modules are apparently not designed to work without interference from an adjacent frequency.
A rolloff of 20dB (0.1) may be sufficient but you probably want a rolloff of at least 30-40dB (.03 to .01) between channels to minimize interference.

 

All right then. I appreciate your reply.

 

The other way to do it is to use the same receiver and have each transmitter encode a different number as data. There could be collisions between 2 transmitters though, and you would need to deal with that.

Bob

 

The other way to do it is to use the same receiver and have each transmitter encode a different number as data. There could be collisions between 2 transmitters though, and you would need to deal with that.

Bob

Hi BobTPH (through part holes?), thanks for your reply.
I was hoping to make the transmitter circuits real cheap and low component count. Preferably not involve any intelligence.

My next hope is to do like you said but only transmit PWM signals and have a duty cycle detector on the receiving side. On the receiving side there would be an mcu either way so I can use timer capture and measure pulse length.

When it comes to collisions between signals, current solutions to this already use the same receiver on multiple transmitters and they just narrow down the field of view on the receiver to where it's so tight that it only receives when the contender is directly underneath it.

 

Notice in your diagram that the center frequency itself has a tolerance of 5% while most of your channels are only separated by about 10% (and 38 kHz and 40 kHz are only 5% apart). So you can pretty much forget using channel discrimination.

Your narrow field-of-view approach has to deal with two vehicles that are tied when coming across the sensor window.

Any system that is not highly directional (such as the narrow field-of-view) has to deal with multiple legitimate signals stomping on each other. This is not trivial, but spread spectrum techniques may offer an answer. If you want the transmitter circuits to be trivially simple, then one possibility is to use concurrent code spread spectrum. The receivers, however, will not be trivial (but not too bad, either -- an MCU should be able to handle the processing). There will be latency, but you will be able to tell who came across first without any problems.