• Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

RF filter required for 36KHz

Status
Not open for further replies.

MikeBrady

New Member
Hi,

Can someone please provide me with a diagram of a circuit that I can use to detect the output from an IR diode which is being switched with an underlying frequency of around 36KHz. The output from the LED will be on/off pulses, each pulse having a period of say, 200ms. When the LED is 'on' it will be sending IR out at a frequency of about 36KHz.

What I need to do is detect when the beam from the diode is interrupted.

I've been trying to do something similar using a specialised IR receiver chip (ZD-1952) but without success. I thought that going back to basics, using an IR phototransistor, some kind of amplifying circuit, and a filter should let me detect the square wave that I will be emitting from the diode. Once I have the waveform reconstructed in the receiver, I can then use the 555 'missing pulse' circuit to detect when the beam (pulse train) is broken.

However, the part I don't really understand is how to filter the signal being received by the IR photo transistor, so that only the required frequency gets through. I need to eliminate all other spurious IR frequencies.

I have seen a few sites that provide filter calculators, but I immediately get lost, for example, I don't know what value to use for impedence etc. Any help in determining the component values for the filter is appreciated.
 

MikeBrady

New Member
The problem could well be that I don't understand the input requirements for the ZD-1952. It is a 'tuned' IR receiver, so I thought it would be a good/easy way to detect the signal being emitted from my IR transmitter.

I've already posted my transmitter circuit, and a variety of CRO displays taken from various points in the circuit. The transmitter circuit can be seen at http://www.electro-tech-online.com/attachments/circuit-jpg.29672/

The waveform being transmitted by the IR LED is shown in the CRO display http://www.electro-tech-online.com/attachments/crotracea-jpg.29674/

The ZD-1952 has 3 pins - ground, vcc and an output. I simply wired it up to 5v and measured the voltage at the output with/without the transmitter diode powered up, and with/without something blocking the signal. The results were quite inconclusive.

What I was expecting was that the ZD-1952 would give me a low reading when the expected signal was being received, and a high reading when it was absent. It is possible that I have misunderstood how this component works, but if you are interested, you can find the spec sheet here http://jaycar.com.au/products_uploaded/ZD-1952.pdf. You might have more luck deciphering it than I did...
 

Nigel Goodwin

Super Moderator
Most Helpful Member
Can't comment on that particular device, but ALL such devices I've used work like that, and work flawlessly - no carrier HIGH, carrier LOW.

However, they are intended for IR remote control, and aren't designed to work with a continuous carrier, which will make them drastically reduce their gain in order to overcome the error.
 

crutschow

Well-Known Member
Most Helpful Member
The ZD-1952 spec sheet notes that the carrier should be modulated with a pulse signal of 600±200µs. Thus if you apply a continuous carrier signal you will likely get no output.
 
Last edited:

jimmythefool

New Member
A lot of optical noise can be attenuated by the mechanical design of your sensor housing. ie using as small an aperture as possible, (narrow tube or similar) and setting the sensor inside, rather than flush.
 

Sceadwian

Banned
The one's I've used required a resistor on the VCC line along with a cap to act as a low pass from the power lines, I got a whole bunch of garbage without it. But you can't use a steady stream of carrier signal like that, it'll have to pulse. Nigel has pointed this out in a couple other threads and that's what crut is saying. You need your transmitter to send out strings of pulses rather than a solid carrier, and your receiver has to detect these pulses for the on/off function.
 

jimmythefool

New Member
The one's I've used required a resistor on the VCC line along with a cap to act as a low pass from the power lines, I got a whole bunch of garbage without it. But you can't use a steady stream of carrier signal like that, it'll have to pulse. Nigel has pointed this out in a couple other threads and that's what crut is saying. You need your transmitter to send out strings of pulses rather than a solid carrier, and your receiver has to detect these pulses for the on/off function.
I agree.. I think those tsop5038's I posted, have fixed gain, so should work without modulation, just a carrier
 

Sceadwian

Banned
Better find out, cause if you're wrong nothing you do is gonna work without a modulated carrier.
 

Sceadwian

Banned
It depends on the exact module.
 
Last edited:

MikeBrady

New Member
But you can't use a steady stream of carrier signal like that, it'll have to pulse. You need your transmitter to send out strings of pulses rather than a solid carrier, and your receiver has to detect these pulses for the on/off function.
Yes, that's precisely what I was trying to do with my transmitter circuit...i.e. there was a carrier with frequency of 36KHz, and I was trying to pulse this with a period of 1200us, however I couldn't quite get this pulsing part right. The screenshot from my CRO shows the waveform for the signal being sent to the transmitter IR LED http://www.electro-tech-online.com/attachments/crotracea-jpg.29674/. You can see that the period is about 1300us and the duty cycle is not 50%. I guess this is why the receiver did not seem to detect my signal.

However, having read some of the replies, I am now convinced that using the IR receiver in this way, i.e. to detect a beam being broken, is really not a good use of the device. The IR receiver is designed to be normally in the 'off' mode, i.e. no signal being received - what I was designing it to do was the complete opposite of this.

So, back to my original question. I have a pulsed wave being emitted from the IR LED. What would a receiver circuit look like that would be able to detect this...and recreate the waveform so that I could then use it to detect missing pulses?

I 'think' it will involve
a) IR photo transistor
b) an amplifying circuit
c) a filter circuit to remove all but the 36KHz signal, which should then leave just the pulses...

Once I've got the pulsed signal extracted, I think I'll be good to then use the 555 timer chip in its 'missing pulse' configuration to detect the missing pulse(s). This will in effect, tell me when the beam has been broken.
 
Last edited:

mneary

New Member
You're setting up, e.g., across a doorway? Do you have to deal with bright light (sunlight)?

I've been contemplating building a general purpose sensor system like this, but what I've been thinking is to send a bright pulse and look for a rising edge on the receiver. (Clock to a latch just a few microseconds after the pulse.) The receiver would use a transconductance amp configuration to give it wide dynamic range.

(Google for circuits and explanations.) Look up Rob Paisley for lots of simple sensor ideas.
 

MikeBrady

New Member
You're setting up, e.g., across a doorway? Do you have to deal with bright light (sunlight)?
This intent is to set this up across a driveway, but only activated at night, and using a mirror to reflect the beam, so transmitter and receiver are co-located. I think there is plenty of scope for using the same frequency generating mechanism, to also act as the 'timer' for detecting the reflected pulses.

Thanks for the tip on using a transconductance amp - I've never heard of one of these before, but I'm sure I'll find plenty using good old Google :)
 

MrAl

Well-Known Member
Most Helpful Member
Hi Mike,

Some of these devices, as others have pointed out, require a pulsing carrier
not a continuous carrier. They are made to work with signals that are typical
of real remote control transmitters. This means that if you can find a remote
control that you already have for something else in your home you can use
it to test your receiver and see if it works at all to begin with. The remote
you use for the test of course has to have a carrier that matches the
receiver, but sometimes even far different carriers will work at a closer distance.
You can hold it close to the receiver and look at the output of the device
and see that pulses are being emitted.
From what i see on that data sheet it looks like you can send a test signal
of 600us carrier and 600us silent and you should see something on the output,
if you want to do that.
The other thing is to make sure that the IR diode emits light that is close to 950nm
because that is also a requirement of your receiver.

These receivers are pretty simple to get working...you power them up,
shoot a signal at them, look at the output on a scope. If you use a digital
scope make sure you have the sample rate high enough to see the pulses
too.

Once you get the basic receiver working we can move on to a design for
your filter.

Good luck with it.
 
Last edited:

jimmythefool

New Member
This intent is to set this up across a driveway, but only activated at night, and using a mirror to reflect the beam, so transmitter and receiver are co-located. I think there is plenty of scope for using the same frequency generating mechanism, to also act as the 'timer' for detecting the reflected pulses.

Thanks for the tip on using a transconductance amp - I've never heard of one of these before, but I'm sure I'll find plenty using good old Google :)
Is this a 1 off Mike? if so, an Omron Photo-reflective sensor would be perfect. They're about £30 or so , over here.

If you want simple, then a Photo-Logic sensor is perfect. Like Sceadwian said, its not good practice to use a photo-sensor as a switch by just sending a carrier, but it can work. I think the TSOP5038 that I posted earlier is designed to work like this. If you're using a microcontroller, you can filter out spurious pulses in the software.
 

Martel

New Member
Ask around if anyone has a scrap VCR.

In all VCR's there is an infrared detector unit, already fit to detect the high frequency IR pulses from a remote. They are easy to recognize: Ther are made of a metallic shielded unit (around 1/2 inch) with three legs and an IR photodiode visible from the front. Check the PCB for connections: Ground, supply and output. Use an IR remote and check with a DMM/scope to identify the pinout and measure the supply voltage.
 

Sceadwian

Banned
No offense Martel but it's customary to read the posts in the thread before you make a new post =) The OP is already using such a module but his transmitter is only broadcasting a solid carrier which those types of modules will tend to ignore. They're designed to recieve pulses of the carrier, not solid carrier for long periods of time.
 

MikeBrady

New Member
Is this a 1 off Mike? if so, an Omron Photo-reflective sensor would be perfect. They're about £30 or so , over here.

If you want simple, then a Photo-Logic sensor is perfect. Like Sceadwian said, its not good practice to use a photo-sensor as a switch by just sending a carrier, but it can work. I think the TSOP5038 that I posted earlier is designed to work like this. If you're using a microcontroller, you can filter out spurious pulses in the software.
Thanks for the tip re Omron. However, I am doing this project mainly as a way of increasing my understanding of signal processing etc, so I would really like to be able to detect these pulses using 'first principles' and basic components. When/if the project is completed, I will need about 8 sensors, so price is also a consideration here.
 
Status
Not open for further replies.

Latest threads

EE World Online Articles

Loading
Top