IR led and reciever module selection help

Hi guys , looking for input on the correct selection of IR LED and IR receiver module combination , in that i want to send a constant pulse stream of high at 800ns / low at 1.4us modulated to 38Khz , the requirement is to be a fine focus ir stream to IR receiver module over about 20 meters in outdoor open area during day conditions . where interruption after focus to stream is noted by system

Questions / assumptions .

IR LED selection:
1. IR led selection be one with a narrow degree angle so as to ensure a fine focus to IR reciever - 10deg or lower at 3deg
2. rise / fall time ability of the LED to need to be less than 800ns - typicaly thinking about 20 - 50ns or there about
3. peek wave length of IR LED should match IR receiver module selected or visa versa
4. high radiant intensity , typicaly 1000mw/sr or higher ? given distance and expected outdoor environment issues ?

IR receiver Module Selection
1. peek wave length of IR LED should match be the same as IR receiver selected or visa versa
2. transmission distance should be typically 20 meters or better
3, internal filter for a 38Khz Ir signal
4. visable light suppression by ir filter , given useage
5. Receptivity / angle of half transmission distance +/- 45Deg ( this seems to be standard for most Ir modules i have looked at )

Notes.
When looking at peek wavelength vs spectral sensiivity graphs of IR reciever , the " sweet spot" should be middle of top end of the curve - correct ???

cheers
sheldon

not look for help on which particular device to select , i can look those up , just if my assumptions of what is required is correct

I do not think you will be able to get IR working at that distance in the sun without a very high power transmitter. Think of a small laser pointer at that distance in the same conditions... The sun generates too much noise.

hi dave , you may be correct , still like to know if the approch in relation to peek wavelength matching selection critiria is the correct way to evaluate the selection , and if rise fall time selection of the led is valid on selection for pulse i wish to send

I used the TSHG6210 with a TSOP7000 and achieved up to 120 meters with a peak current on the transmitter up to 180mA. With lenses on both side.

Ioannis

thanks Ioannis

I see that its at 455khz and would be a good test what i am trying to do

you show some code on how generate frequency cycle

i see the reciever requires

The data signal should fullfill the following conditions:

The carrier frequency should be close to 455kHz.

The burstlength should be at least 22s (10 cycles of the carrier signal) and shorter than 500s.

The separation time between two consecutive bursts should be at least 26s.

If the data bursts are longer than 500s then the envelope duty cycle is limited to 25%.

The duty cycle of the carrier signal (455kHz) may be between 50% (1.1s pulses) and 10% (0.2.ms pulses).
The longer duty cycle may help to save battery power.

all timing are in microsecs not secs , copy n paist not work correct

i think tsop7000 may be a deleted part , as the manafacture does not list it anymore ???

I had not looking for this device long time now, and you are right. It is obsolete now and Vishay does not list any other 455KHz alternative. That is bizzare...

I am sorry for a mistake on the above post, my 120 meter application uses the TSOP1738/31338 device and with a carrier of 38KHz. The pulse train is 15 bursts of 26,3 usec every 400 or 800 usec (2 channels selected by jumper).

The version of TSOP7000 is for a low power IR beam setup and may cover up to 40 meters maximum. It sends 16 pulses of 400ns every 2.2 usec with a period of 17,5msec of pause.

An example of 455KHz carrier would be (on a 12F635):

Code:

---

For i=0 to 15
    @ bsf gpio,1
    pauseus 390    'this number may be lower or higher depending on internal clock freq
    @ bcf gpio,1
    pause 2200
Next i

---

I cannot post the full code of my application because it is a commercial one.

Ioannis

Thankyou Ioannis for you great help ,

i contacted Vishay engineer and he confirmed that the TSOP7000 ( 455khz ) receiver is no longer being made , and no item in there range suites the 455khz now.

Although i was able to get some good pricing for the receiver , still alot of them about and can be found
cant seem to find any other manufactures that now do a 455Khz receiver else id go with TSOP7000

if you know of another 455khz receiver in production let me know please, i do a bit of search today as well.

Also as it turns out there is no IR led in Vishay range at 10deg 940/950nm at 1000mw/sr or higher with a switching speed lower than 800ns. but if used at 38Khz should not be a factor.

but is far more preferable for the design to use a higher frequency IR receiver at 455Khz than at 38Khz as it takes less time to dectect a missing pulse

eg when using a IR reciever at 38khz then scan time = 26.3us x 11 = 300us ( min detect time for proposed IR reciever TSSP4038 )
so 20leds x 300us = 6ms per scan
3 scans of the array = 18ms

eg when using a IR reciever at 455khz then scan time = 2.2us x 12 = 26.4s ( min detect time for proposed IR reciever TSSP7000 )
so 20leds x 26.4us = 528us per scan
3 scans of the array = 1.58 ms

more scans can be done for shorter time using 455Khz


speed is everything sometimes lol

cheers

sheldon

also note that the TSOP1738/31338 is discontinued but an equivalent is TSOP4838 - agc2 / TSOP34156 - agc 1) if u need to find simular

If size is not that important for the transmitter you can put an external lens in front of the IR diode. An old magnifying glass will do a great job. You can see this if you use a RED visible led or a CMOS camera (mobile phone cameras usually work too).

Depending of the lens you will have to adjust the distance from the led to the lens to get the correct focal plane.

For the receiver you can put the TSOP in the end of a short black pipe so it can not get direkt sunlight. If you do these 2 things you can easily get 100m using a normal IR led and 100 mA in current indoor. If you also add a lense on the receiver you will get even more distance.

Dave , i am guessing that the IR leds that able to give from 1000mw/sr - 2300mw/sr cant be driven by the pic output directly and need a a nice Mosfet to sink the current , which i am guessing would sink about 100ma - 200ma.

Any recomended favorite N ch Mosfets suitable

i would guess be 20 -50v BVDSS , tdly on = <50ns or less , tdlyoff = <100ns , Ftime <50ns ,Ron <1ohm , current drain cont =>1000mA , Pulsed >4000mA

Also will i need a pull down Resistor of say 10K to ground on the gate or can the PIC internal pull down be enough ???

Cheers

Sheldon

sorry i ment last msg for Ioannis

For mosfets I am using the SI2302 (http://gr.mouser.com/ProductDetail/V...t414B%252b8%3d). According to the LED used, I am calculating the maximum current and then the series resistor in relation to the Power supply level.

Ioannis

will i need a pull down resistor on the gate for N channel , hear that they can sometimes not turn off clean when connected to pic i/o ??

No, the output of the controller is well driven high or low.

Ioannis

YES you will need a pulldown resistor if your current limiting resistor is choosen for short pulses and not continous operation.

If you choose to drive the LED hard during short pulses to get more light from it you must use a pull down resistor to make sure the LED stays off until the PIC has powered up and taken control over the output pin. Since the PIC will set all pins as input, without the resistor controlling the gate during this time there is a high risk that the gate of the FET can float a while and that can result in a burnt LED.

So if I were you I would put a pull down resistor just to make sure, a low cost to reduce a bigger risk.

There is a reason to place a resistor as Jumper says.

But with the LED I am using and the timing calculated there is no need to place a resistor for the time PIC is in power up state. Even with Power up timer enabled.

Ioannis

cheers guys ,

had not considered the time up state as a possible floating input on the MOSFET and then allow it to turn on , i would normally say what the heck add the resistor to ground for the gate , but if i have 20-30 to do for only that issue ill may let it pass ,

i guess ill build the prototype and see some measurements

For the emitter, couldn't a cheap ebay laser pointer be used? Or would that be too intense even in sunlight and overwhelm the receiver?

Or do we get into legal area of having a device that could harm curious kids?

(totally uninformed in this, just curious)

Robert

i have thought about lasers but given that persons even having the chance of damage to eyes , is not worth the risk or the possible cost , if it were to be used in a area that did have a less chance say in high roof area that is to be monitored perhaps , but there are those devices out there.

I have chosen to go with 38Khz IR to test , 455kz would have been better suited , but to design something on parts that are obsolete with no equivalents , and needing at least 20 per unit , is not wise

will see how it goes with the 38Khz

Cheers

Sheldon

HI guys , can i have some input on how to allow this code to send a shorter pulse than 7us , which mean some ASM changes , to allow multi i/o ports which is whats need i think to get a small pulse time
chips running at 32Mhz and best min time i get is about 7us

Cheers

Code:

---

For  T =  0 to 9                        ' Select IR LED for Pulses
        LOOKUP T, [1,2,4,5,16,17,18,19,20,21],PortOffset        ' Offsets from Port A to point to port a and port c I/o ports ( number of offset relates to bits offset in mem map of chip)
                Cycles = 0                    ' Set to 0
            for Cycles = 1 to Pulse_cnt        ' Each IR led increments the amount of pulses by 2 for IR LED ID use
                  PORTA.0[PortOffset] = 1      ' Set IR LED High
                  pauseus 2                        ' Remain on for 2uS
                      PORTA.0[PortOffset] = 0      ' Set IR LED Low
                      pauseus 9                    ' remain off for 9uS
            next Cycles
         
         
          Pulse_cnt = Pulse_cnt +2              ' increment pulse count by offset of 2 for each IR LED
          pauseus 320                            ' Wait 350us between pulse bursts and selecting next IR LED
        Next T

---