TSOP 1738 Noise Problem

[Suraj143]

I'm interfacing TSOP 1738 with PIC.When powering the design using a transformer power supply its working.
But when powering the design using a capacitive power supply the output timing of TSOP1738 is varying (Garbage values).For the PIC I use only 10uF decoupling capacitor.

Can I fix this by changing values of *R,*C near the TSOP sensor?I tried 100uF also but no luck

 

[Nigel Goodwin]

What do you mean output timing is "varying"? - output from TSOP's does vary, it's why you need specific coding systems to over come it (such as RC5 or SIRC's).

However, if you posted your COMPLETE schematic, there may be problems in your design (other than the usual capacitor supply problems).

 

[AnalogKid]

Also, wht do you mean by a capacitive power supply? Make? part number? Schematic?

ak

 

[audioguru]

I think he has a Mickey Mouse dangerous transformerless power supply that uses a capacitor's reactance to drop the voltage. Its voltage will vary with loading and the TSOP probably has a varying current.
Maybe the TSOP is picking up 38kHz IR from a compact fluorescent lightbulb nearby and its AGC cannot prevent it.

 

[Suraj143]

Ok thanks.

Audioguru came closer.Yes I have a power supply noise & some sort of current lagging so the output timings of TSOP will get affects. I want to eliminate here is the circuit that I use.

When I remove the red area & powered from a DC source its working fine.

 

[AnalogKid]

The AC power supply is a high-pass filter, excellent at bringing in outside noise. Add 0.1 uF and 0.001 uF ceramic capacitors in parallel to C2, C3, and C4. What is IC1?

ak

 

[Suraj143]

Hi
It is PIC12F675.

 

[ronsimpson]

Most people put a 0.01 to 0.1uF cap across the micro controller power pins. Big capacitors to not have good high frequency response. You could also try a small cap on the IR receiver.

Does the 12V work well when the relay is closed? That was the problem before. I think this problem is fixed.

 

[Suraj143]

Most people put a 0.01 to 0.1uF cap across the micro controller power pins. Big capacitors to not have good high frequency response. You could also try a small cap on the IR receiver.

Does the 12V work well when the relay is closed? That was the problem before. I think this problem is fixed.

Hi

I forgot to mention,Yes the relay worked nicely.Thanks.

Ahhhh that means this is lack of decoupling...!!!I'll try & see.

 

[ronsimpson]

The AC power supply is a high-pass filter, excellent at bringing in outside noise.

I also am thinking about noise from the power line. Here is a very simple circuit to help think about noise. I did not add ESR and ESL of C1 etc.
The green line is the noise reduction from line to +12V. Then the blue line is for 5V and then the TSOP 1738 supply. This might help with choosing a capacitor. When the Zeners are added in the noise should go down greatly.

Bottom line is that the TSOP 1738 is very sensitive to supply noise. I don't know if it is 100/120hz noise or 38khz noise. But it must have a good supply. The data sheet uses 100 ohm and 10uf which does not help at 100hz so they must be more interested in high frequency noise from a micro. BUT They probably think a 7805 voltage regulator is making the +5V so there is no 100hz noise.

We need to think about this a little.

 

[Nigel Goodwin]

Yes I have a power supply noise & some sort of current lagging so the output timings of TSOP will get affects.

What kind of coding scheme are you using?.

 

[audioguru]

If I bother to simulate the horrible transformerless power supply I betcha the "5V" for the TSOP has a lot of 100Hz ripple. Most of the rectified current pulses simply heat the 12V zener diode. Then the 220uF capacitor quickly discharges at 100Hz due to its load.

 

[ronsimpson]

Because the TSOP needs a very clean supply you could take out the 5.1V Zener and 1k resistor and try a LM78L05 voltage regulator.
There are MOS versions of the part that wast much less power. (quiescent current is typ. 2mA and max. 5.25mA) There are other small 5V regulators that are down in the 100uA range)

I have a 0.15V pk ripple on the +12V line. At 120/100hz. Which will not effect the relay.
I have 2mV pk ripple on the +5V line. Which will not effect the micro but will effect the TSOP.

 

[Suraj143]

Ok thank you guys.

Yes I have a power line noise affecting the receiver.The TSOP1738 is good if it has a well regulated power supply.Unless I have to add decoupling caps & change R C values near the receiver & check the output timings.

If I cannot eliminate the problem I must move to a high noise immunity series on vishay products over TSOP1738, something like TSOP2238.

 

[Suraj143]

That's a good solution to replace 5.1V zenner & 1K with a 78L05.I will try.

 

[Suraj143]

Because the TSOP needs a very clean supply you could take out the 5.1V Zener and 1k resistor and try a LM78L05 voltage regulator.
There are MOS versions of the part that wast much less power. (quiescent current is typ. 2mA and max. 5.25mA) There are other small 5V regulators that are down in the 100uA range)

I have a 0.15V pk ripple on the +12V line. At 120/100hz. Which will not effect the relay.
I have 2mV pk ripple on the +5V line. Which will not effect the micro but will effect the TSOP.

That's a good solution to replace 5.1V zenner & 1K with a 78L05.I will try.

 

[Suraj143]

What kind of coding scheme are you using?.

Hi I am making some sort of universal remote receiver.I save every bits timing in EEPROM & match them with the incoming pulse stream.I have one thing to ask.
If the stored value is 500uS then I must check it with a range like 400uS to 600uS.

I want to know whats the threshold point on this range?

Due to distance between remote & receiver does the timing affected? In your tutorials you also checking a range instead of exact timing...!!

 

[Nigel Goodwin]

Due to distance between remote & receiver does the timing affected? In your tutorials you also checking a range instead of exact timing...!!

Everyone does, because there is no 'exact timing', the output from a TSOP varies a considerable amount depending up on many factors, one of which is signal strength.