The reason for the carrier is, as you guessed, to make it so that you can easily filter out all other sources of IR that might happen to make it into your detector. As long as they aren't strong enough to saturate your detector, you can simply ignore them -- unless, of course, the noise source is also modulated at the same frequency.

 

There was another post similar--Help with simple break beam SW by deuce01. Put deuce01 in find box this page upper right. Included sensors 1 ft apart, common IR driver. Might gleen a bit of inf.

 

The reason for the carrier is, as you guessed, to make it so that you can easily filter out all other sources of IR that might happen to make it into your detector. As long as they aren't strong enough to saturate your detector, you can simply ignore them -- unless, of course, the noise source is also modulated at the same frequency.

Having just built a quick & dirty remote handset tester with the IR sensor from the front edge of a scrap TV board, I notice that the LED lights a little bit if I point it at the room light CFL - Maybe the self oscillating half-bridge driver in the CFL isn't far enough from the pass band!

 

For the trailer problem, would blocking the count for say 300 ms after first detection be viable? If trailer is traveling below average speed, then it would still count as two. A second car riding bumper of first would escape counting.

 

For the trailer problem, would blocking the count for say 300 ms after first detection be viable? If trailer is traveling below average speed, then it would still count as two. A second car riding bumper of first would escape counting.

Besides counting the trailer as two vehicles, there is a good chance that you can count two vehicles as one with such delay. I'm going to try building a barrier which can detect the connection between the vehicle and whatever it is pulling.

 

Besides counting the trailer as two vehicles, there is a good chance that you can count two vehicles as one with such delay. I'm going to try building a barrier which can detect the connection between the vehicle and whatever it is pulling.

So now you want to detect something that might only be a couple of feet long and a few inches in width and heighth?

Just how many beams and detectors are you willing to throw at this problem?

 

Just how many beams and detectors are you willing to throw at this problem?

As I said before: I'm going to do some tests in order to know how many I need.

Right now I'm in the process of defining if I'm going to use a detector or a simple phototransistor. Detectors seem to be the best option to get rid of the ambient noise problem. But this comes at a price. I can't send a continuous signal, there must be a pause of 15ms between each burst. This is not a major problem in the emitter side. I could do that with two NE555 (or one NE556), but in the receiver side I can't figure out how to build a circuit capable of detect those burst without using a microcontroller.

 

In the post by deuce01, pages 66, 70, 73, &76, there were no micro controllers. As I remember the delay between Tx bursts was 1 ms.

 

As I said before: I'm going to do some tests in order to know how many I need.

Right now I'm in the process of defining if I'm going to use a detector or a simple phototransistor. Detectors seem to be the best option to get rid of the ambient noise problem. But this comes at a price. I can't send a continuous signal, there must be a pause of 15ms between each burst. This is not a major problem in the emitter side. I could do that with two NE555 (or one NE556), but in the receiver side I can't figure out how to build a circuit capable of detect those burst without using a microcontroller.

I've just discovered that the Dettol no-touch hand gel dispenser contains useful looking IRLED & sensor, just lately many discount stores have been selling the starter packs for less than just a refill - one local store currently has a buy one get one free offer for a lower price than just a refill.

Recently I visited a relative who uses these - they had a regiment of empties lined up on the loo windowledge, and had just bought the 2 for the price of 1.

The older model contains a 4.5V motor (6V in the current one) with a worm-gear on the shaft, there are various O-rings on the delivery pipe, most of the electronics are in a "black blob" chip - but there's a pretty tasty transistor; 2SC3279 to switch the motor, 2A continuous in a TO92 package - I wish I had a few more of those!

 

I extracted this from old post. Worked with several different IR modules, pin-out different. On receiver, Rx, R1 is adjusted so that C2 only charges to about 2/3 of ref V, V3-3 for one Tx cycle. If pulse group is missing for 2 or 3 cycles, output V3-1, falls. In origional ckt. out was cap. coupled to a 555.

 

I extracted this from old post. Worked with several different IR modules, pin-out different....

I need an IR receiver capable of doing that. The one I'm testing won't allow a continuous pulse train. It stops detecting after 137ms in this test, as you can see. The emitter is sending pulses of 600µs high and low, using a 38kHz carrier.

The receiver outputs a train as soon as I point the emitter to him. But its output stops after 136ms even when the signal is still present.

TMF5360 seems to support a continuous pulse train, but it seems to be discontinued. At least I can't find someone selling it.

 

The TFM5360 behaves like all the other IR receivers. If 3 or 4 cycles is too much delay, might go back to a continuous pulse train & use a IR diode or transistor-- believe I tested it last year, will check file. Just about any CMOS inverter can be used for the 4049. Must go for now

 

I've tried that too. But then I lost all the advantages of noise filtering I have with the receiver.

If TFM5360 behaves like the VS1838 then the circuit you proposed before will stop the output unless there is a pause in the pulse train. I could add a NE555 in monostable configuration to skip the pause. That will add an error in detection time of +Tp (where Tp is pause duration). Most datasheet don't speak about this issues. From the datasheet of the TSPO1738 I know "...After each burst which is between 10 cycles and 70 cycles a gap time of at least 14 cycles is necessary."

As I was writing this I tested this in the VS1838. I had to add a pause of 2ms after a burst of 12 pulses of 263µs Ton and 263µs Toff. I think I can bare with 2ms error in detection using the monostable. I will have to try that.

 

Digging back, ckt in post 32 gave a range of 20 ft. Tried VEMT2020 for TEMT1020 and blocks out all stray light except direct sun; range was shortend to 17 ft but gain could be increased.
A plastic lens was used for Tx, about 35 mm, FL 15 cm. shrunk beam to about 6 in @ 15 ft.