Best way to trigger a phototransistor?

[nolan]

[removed to hide information from other teams that may find this lol]

[Phil1977]

The main questions I would like input on are:

• Which LED would be more effective in activating the phototransistor module? We have the option of running the lights off of a separate battery so power budget is not a huge concern.

If you´re using no colimation optics behind the LED then generally more power is better. The wavelength specific sensitivity of the PT for the green light may be higher, but the white LED provides so much more light.

• How would we maximize our triggering distance while keeping a relatively large projection area for the LEDs? We are not able to point the robot at each phototransistor module accurately, we would like to take a shotgun approach to activating them.

I think the key technology for achieving this goal is modulation and filtering. Either you modulate your PWM frequency with a much higher carrier frequency and filter for this in the receiver. Or you could even build a (digital) filter that passes only the base frequency and the harmonics of the PWM. The better your filter is, the better your distance will be. E.g. a simple infrared remote control reaches quite far without fancy optics but with a combination of wavelength- and modulation frequency filtering.

• Is there anything that I'm missing here? New ideas or suggestions would also be awesome.

You could also use the green LED together with a matched optical filter in front of the phototransistor. If you put a filter before the phototransistor that blocks as many wavelengths as possible that do not match the LED output, then your SNR and so your reachable distance will rise further.

[Xena E]

Is there any optical filtering on the receiver?

That may have some bearing on your decision in case ambient swamping is a problem.

If, the pwm required is low you may be able to jack the current to increase your range.

In a similar challenge, the acceptance frequency at the receiver being quite low was addressed by a mechanical scanner, the robot only needing to do a drive by shooting, rather than aim as in a tank turret, if that can be done that's a sure winner.

[Kleinstein]

With high power of the white LED about compensates for the lower sensitivity to white light. So one may have to try or do the math in details with the spectrum and sum things up.

The phototransistor with filter to get an about human eye response is not a good choice for high sensitivity. They get this with a filter that reduces the sensiticvity in the red / IR range. Chances are a more normal phototransistor or photodiode could be better.
The way the signal from the detector is amplified can make quite some difference. The output is usually not just on / off, but an analog signal with intermediate levels.

A big point can be the effect of background light, especially artificial light, as this often has hum and thus modulation.  If there is much background light, one could consider IR LEDs and phototransistors like in IR remotes, as the normal LED lamps emit only little IR / dark red and thus have less interference there.

[PCB.Wiz]

...
The main questions I would like input on are:

• Which LED would be more effective in activating the phototransistor module? We have the option of running the lights off of a separate battery so power budget is not a huge concern.
• How would we maximize our triggering distance while keeping a relatively large projection area for the LEDs? We are not able to point the robot at each phototransistor module accurately, we would like to take a shotgun approach to activating them.
• Is there anything that I'm missing here? New ideas or suggestions would also be awesome.

As far as we can tell, the strength of our light setup is the only good way to differentiate our robot from that of other teams, as it would allow us to complete the competition faster and with less risk. We really want to nail this so any criticism and advice is greatly appreciated.

It's a bit unclear - do you have control of both ends, or are you LED side only?
Is each team allocated a single frequency to send, as their ID  ?

Simple parts like TEPT4400 do not specify a response time, and you should check what the activate power/time and range are, in a real installation.

Items to check
 Do they manage background illumination in the receiver ?
 Can you overdrive the receiver ? ie do you need a two level TX burst ?

[nolan]

Hey everyone, thank you so much for all the responses! Unfortunately we have no way to control the receiver end, as this is a fixed part of the course. The PWM frequencies involved are very low, two teams are on the course at the same time and one uses 40Hz while the other uses 80Hz.

I really like the idea of using an optical filter to improve SNR! If these are available in large sheets one of my immediate thoughts is that we can have the robot drag/push one around on the floor, as the phototransistors are inside of an assembly set into the floor. We have not received the modules to test with from the prof yet, but as far as I’m aware there is little to no optical filtering or signal processing.

[Njk]

E.g. a simple infrared remote control reaches quite far without fancy optics but with a combination of wavelength- and modulation frequency filtering.

Concerning the application, keep in mind the price of that advantage. The light can reflect from the walls making such a remote control almost omni-directional when used in closed space.

[Xena E]

Hey everyone, thank you so much for all the responses! Unfortunately we have no way to control the receiver end, as this is a fixed part of the course. The PWM frequencies involved are very low, two teams are on the course at the same time and one uses 40Hz while the other uses 80Hz.

Dependant on the acceptance angle of the receiver, if the mechanical scanning idea is considered worth pursuing thats 2400 and 4800 rpm, fast but both doable.

(Think mirror drum, the duty cycle being proportion of the rotation).

Seems like a fun college you attend...

[ledtester]

If you are considering multiple ideas, just mount all of them on robot and activate them one at a time during the competition.