555 timer for ic transmitter error

[bouvett]

hi, i am using a 555 timer at 51% duty cycle configured as an astable for a frequency of 38khz, for my ir receiver.

now the resistances i used are 180k, from Vcc to pin7 and 16k from pin7 to pin 2and 6. the capacitor i used is 100pf. these values were provided by a 555 timer software i have. now when i test it in reality it gives a frequency of 28khz using my digital multimeter.

when i tested this on my receiver, it never turns active... no matter what the intensity of light and no matter what distance. i am sure that the 555 timer works cause i tried it with a red led first and it did lit up. i then tested the reciver using a pic12f629(with a frequency read by the digital multimeter again of 18khz)(receiver rated at 38khz).. the result is a total succss, it worked perfectly using the pic...

now the only difference it seems that the frequency of the 555 circuit is better than of the pic so i wonder how it did not work at all....

i am very confused since both circuits seem to work and give a frequency but the 555 timer circcuit doesn't work...

(i used a momentarily high switch connected to the reset of the 555 timer so that it works only when the button is pressed, with a red led it shown that this function worked properly....)

any ideas of what might have gone wrong please? thanks a lot
regards

 

[MikeMl]

You are using a much too small timing capacitor. The stray capacitances in your circuit layout at the THRES and TRIG pin on the 555 are on the order of 15-25pf, which is 25% of a 100pF timing capacitor. Reduce the resistors, and make the timing cap about 1nF.
You also must consider the capacitor accuracy spec, which is likely +-5% at best. You need to trim the oscillator to the final frequency. I would do this by picking a timing resistor which puts the oscillator slightly below the desired freq, and then parallel much higher value resistors across the timing resistor until it comes in. It is much easier to get a final value resistor by paralleling one which is 10% too high with another that is about 10X the first...

 

[audioguru]

The values of your timing resistors are backwards which makes a duty cycle of 92% and also messes up the frequency.

Most IR receiver ICs have AGC that reduces their gain when they receive continuous 38kHz IR. The IR is supposed to be in bursts of pulses. There must be a certain number of pulses followed by a certain duration pause. It is explained in their datasheet.

 

[MikeMl]

The values of your timing resistors are backwards which makes a duty cycle of 92% and also messes up the frequency.

Good catch!

Most IR receiver ICs have AGC that reduces their gain when they receive continuous 38kHz IR. The IR is supposed to be in bursts of pulses. There must be a certain number of pulses followed by a certain duration pause. It is explained in their datasheet.

Not true! The receiver is most sensitive when it has been receiving a continuous signal for a long time, and then there is a short interruption of the beam! I proved this to my own satisfaction with the Sharp GP1U58Y IR remote-control receiver.

 

[audioguru]

The first line on the Sharp datasheet says that it is "Less sensitive to a fluorescent lamp driven by an inverter".
Compact fluorescent light bulbs are powered by a 40kHz inverter that flickers the light continuously at 40kHz.
But the block diagram of the sharp receiver IC does not show AGC that is used in other makes of IR receivers.

The datasheet shows the 833Hz bursts of 40kHz pulses and pauses between the pauses for the highest sensitivity.

 

[MikeMl]

The first line on the Sharp datasheet says that it is "Less sensitive to a fluorescent lamp driven by an inverter".
Compact fluorescent light bulbs are powered by a 40kHz inverter that flickers the light continuously at 40kHz.
But the block diagram of the sharp receiver IC does not show AGC that is used in other makes of IR receivers.

The datasheet shows the 833Hz bursts of 40kHz pulses and pauses between the pauses for the highest sensitivity.

OK, let me try this one more time: I wanted a beam-break detector that worked in sunlight (outside) No florescent lights within 25 miles!.

I built a 40KHz IR CW transmitter, took it outside, and determined that the receiver reliably detected interruptions of the IR beam caused by moving my hand through the beam at a distance of about 8 ft.

You posted that the range would be better if the beam was modulated with bursts; 600us on and 600us off. I tried that, and AS A BEAM BREAK DETECTOR, the range went down to only about 5ft.

I can only speculate, but at maximum range, the IR signal is barely detectable and the AGC does nothing. The max range is achieved by allowing the detector to see the biggest possible signal (i.e. CW) just before breaking the beam.

 

[electronspeed]

May be you are using TSOP1738.Now if you continuously provide 38khz then IR receiver will recognize it as noise.So better you use 2 555 one with 38KHz frequency & other which will be active for rest part(38/14)KHz you will get the result.

 

[AllVol]

You are using a much too small timing capacitor. The stray capacitances in your circuit layout at the THRES and TRIG pin on the 555 are on the order of 15-25pf, which is 25% of a 100pF timing capacitor. Reduce the resistors, and make the timing cap about 1nF.
You also must consider the capacitor accuracy spec, which is likely +-5% at best. You need to trim the oscillator to the final frequency. I would do this by picking a timing resistor which puts the oscillator slightly below the desired freq, and then parallel much higher value resistors across the timing resistor until it comes in. It is much easier to get a final value resistor by paralleling one which is 10% too high with another that is about 10X the first...

555 Timer Pro demo software calculates R1 as 1.29K, R2 as 18.43K with a 1nF cap and 9 V, resulting in 38KHz with a 52% duty cycle.

 

[MikeMl]

555 Timer Pro demo software calculates R1 as 1.29K, R2 as 18.43K with a 1nF cap and 9 V, resulting in 38KHz with a 52% duty cycle.

The 1.29K for R1 is getting too low. The current through this resistor will be about 10mA when the DISCHARGE transistor is turned on. Relax the 50/50% duty cycle a bit, make R1 about 3.3K, and then reduce R2 a bit to compensate.

Electronic design is full of trade offs

 

[bouvett]

You are using a much too small timing capacitor. The stray capacitances in your circuit layout at the THRES and TRIG pin on the 555 are on the order of 15-25pf, which is 25% of a 100pF timing capacitor. Reduce the resistors, and make the timing cap about 1nF.
You also must consider the capacitor accuracy spec, which is likely +-5% at best. You need to trim the oscillator to the final frequency. I would do this by picking a timing resistor which puts the oscillator slightly below the desired freq, and then parallel much higher value resistors across the timing resistor until it comes in. It is much easier to get a final value resistor by paralleling one which is 10% too high with another that is about 10X the first...

yes it will be a good idea and i will surely try it out, but what mixes me and which no reference was made by you in your post is the fact that the micro controller, produced a much smaller frequency and still worked with my receiver where as my 555 transmitter didn't work even though its frequency was better matching the natural frequency of the receiver ic

thanks a lot
regards

 

[Nigel Goodwin]

yes it will be a good idea and i will surely try it out, but what mixes me and which no reference was made by you in your post is the fact that the micro controller, produced a much smaller frequency and still worked with my receiver where as my 555 transmitter didn't work even though its frequency was better matching the natural frequency of the receiver ic

You haven't really checked the frequencies - the frequency range on a multi-meter is usually nothing more than a toy. Get a scope on it, and see what's going on.

But using a PIC there's no real need to measure it, you can pretty accurately program whatever frequency you want (to the nearest uS with a 4MHz clock).

 

[Boncuk]

R1=3K3, R2=17K31, C1=1nF / f=38.01KHz, duty cycle 54.3%

R1=3K3, R2=17K32, C1=1nF / f=37.99KHz, duty cycle 54.3%

Why not use a 4060 and a ceramic resonator?

Boncuk