Question on Infrared Modulation Frequency

Goeytex

Senior Member
I am measuring a modulation frequency of 41.6 KHz when using IROUT with a Picaxe 20X2. As I understand it, this should be 38 KHz and represents a 9.5% error from nominal.

My question is: Will this frequency deviation from nominal affect the sensitivity/ reliability of IR comms with a typical IR receiver looking for a 38Khz modulated signal or is this "good enough"?
 

nick12ab

Senior Member
ir.png

According to this graph from a data sheet, "responsivity" would be approximately half of what it is at the proper frequency (whatever responsivity is).
 

Goeytex

Senior Member
Thanks Nick,

An 08M2 also tested at ~41.5Khz.

Can someone else with a scope or frequency counter confirm what I am seeing ?
 

Buzby

Senior Member
That's interesting !.

A few weeks ago I did a lot of work changing the IR frequency to see how fast I could transfer some data.

See here for the full discussion : http://www.picaxeforum.co.uk/showthread.php?21531-How-fast-can-can-I-transfer-4096-bytes-between-two-PICAXEs-using-IR

I got reliable results up to 70KHz, but one thing I never tested was the actual IROUT frequency that a PICAXE generates !.

A variance of +/- 10% will have a minimal effect when sending over the normal utilisation range, i.e. sofa to TV.
 

PieM

Senior Member
On 2 14M2 firm: 6.A :
Frequency : 41.705 Duty cycle 33.5%
Frequency : 41.665 Duty cycle 33.3%

On a 08M2 firm 4.A :
Frequency : 41.667 Duty cycle 33.5 %
 

erco

Senior Member
XLNT observation & question, Goeytex. Per nick, typical receivers lose sensitivity off their center frequency. TV remotes blast a ridiculously strong signal, repeating after a pause as long as you hold the button. That intensity bounces off walls & ceilings and a saturates the IR modules to make bulletproof communications. Even 56 kHz IR receiver modules get swamped by a 38-40 kHz TV remote.

If you use a safe value of series resistor (max current=20 mA) to drive your IR LED directly from an IO pin, you'll have a relatively weak signal that might be affected by the wrong mod frequency. Options for a stronger signal include reducing the series resistor (OK for regular duty cycle, but risky for the IO pin if it gets locked on accidentally) or using a transistor switch to drive the IR LED.

Of course another factor is the wavelength of the particular IR LED you use, which should match the IR receiver. And there are dozens of different types of IR receivers, each has its own characteristics, some filter sunlight and noise, some are continuous signal compatible. They all look nearly the same but behave differently. Radio Shack sells two completely different modules (physically and electrically) in the same package, 276-640. The metal frame units are continuous signal compatible, the other is not.

I'll be doing some experiments with IROUT shortly, I'lll be interested to hear your findings.
 

Goeytex

Senior Member
I a preparing to do some "range testing" with some pretty high Power IR LEDs. These are 7watt IR LEDs (1 amp @ 7 volts) with an added 3 degree collimator lens to narrow the beam. I simply want to what kind of range I can get using IRIN /IROUT.

So I am wanting to make sure that the Picaxe being off spec @ 41.6Khz is not going to severely affect receiver performance when using IROUT. I suppose I could do the pulsed PWM thing to get an accurate 38KHz carrier, but I'd rather not.
 

Technical

Technical Support
Staff member
I am measuring a modulation frequency of 41.6 KHz when using IROUT with a Picaxe 20X2. As I understand it, this should be 38 KHz and represents a 9.5% error from nominal.
The Sony SIRC protocol uses a nominal 40kHz carrier, not 38, so IROUT is not as far out of tolerance as quoted. However in reality it will work fine with both 38kHz and 40kHz receivers.
The manual does incorrectly say 38, that'll be updated in the next version.
 

nick12ab

Senior Member
The Sony SIRC protocol uses a nominal 40kHz carrier, not 38, so IROUT is not as far out of tolerance as quoted. However in reality it will work fine with both 38kHz and 40kHz receivers.
The manual does incorrectly say 38, that'll be updated in the next version.
Will you update the PICAXE Store so that it sells 40KHz receivers too?
 

erco

Senior Member
Change the manual and call it a day? You can't find the old Panasonic tin cube 40 kHz sensors anymore, 38 is the new standard. Not sure why maintaining the traditional SIRC standard would be preferable to adapting to modern sensor technology. Sure seems like updating the PicAxe to 38 kHz would be the way to go. Even 39 is an improvement.
 

boriz

Senior Member
"...with an added 3 degree collimator lens..."

Just how transparent to IR is your lens? Might block some of the signal even if it does increase the range. Consider parabolic reflector, at both ends, RX being larger.
 

papaof2

Senior Member
Change the manual and call it a day? You can't find the old Panasonic tin cube 40 kHz sensors anymore, 38 is the new standard. Not sure why maintaining the traditional SIRC standard would be preferable to adapting to modern sensor technology. Sure seems like updating the PicAxe to 38 kHz would be the way to go. Even 39 is an improvement.
40kHz devices are readily available: http://www.goldmine-elec.com , http://www.sparkfun.com/ and there's always ebay.com

Because of the long history and the universal availability of Sony compatible remotes (not to mention working examples/projects) there's good reason to stay with the existing standard. Unless you want to answer all the newbie-to-IR questions of "I copied a project from the forum so why doesn't it work?"
 

John West

Senior Member
I a preparing to do some "range testing" with some pretty high Power IR LEDs. These are 7watt IR LEDs (1 amp @ 7 volts) with an added 3 degree collimator lens to narrow the beam. I simply want to what kind of range I can get using IRIN /IROUT. <snip>
If range is what you want, try adding a cheap plastic Fresnel lens in front of the receiver. A $1 lens will give you the most bang for the buck that you can get. Your range will go way up, perhaps to the point where it becomes hard to find a LOS location to do the experiments.
 

Goeytex

Senior Member
If range is what you want, try adding a cheap plastic Fresnel lens in front of the receiver.
The collimator I have is for the TX side and should work fine. My challenge is to find a Fresnel Lens or some other light
gathering optical solution with a relatively a short focal length so that the receiver is not obnoxiously large.
 

Attachments

Goeytex

Senior Member
Update.

The frequency does affect receiver "sensitivity" quit a bit. The closer to nominal the better so if using Picaxe IRIN/IROUT make sure receiver is specified for 40KHz as that is the closest standard frequency to the Picaxe's rather odd 41.6 KHz.

Anyway, I am not using IRIN/IROUT because I want to send serial data . I salvaged a 56KHZ receiver from an old Cable Box for testing. With a 56Kkz carrier generated by Picaxe PWM that is modulated by 2400 baud serial data I am getting > 300 meters with the current setup. That is the limit of my property here. The IR LED is a 7 watt device from Epitex (Japan) ..... DATASHEET HERE

I have the peak current limited to 800 ma because I am using a 2N2222 and a 2N2907 for driving the lamp, but could pulse it to up to 5 amps per the datasheet. Looks Like I need to change the driving configuration and find more space to do longer range testing.

Has anyone thought about or tried using RFIN/RFOUT to add Manchester encoding & error checking to Picaxe Based IR serial comms? Can it be done?
 
Last edited:

jedynakiewicz

Senior Member
@Goeytex, your project is most fascinating and I was very interested to follow this thread that now appears to have dried up for a few weeks. Do you have any further reports on its progress? You mentioned a need for infra-red transmissive lenses of a short focal length in post 16. I use f= 2 inch infra-red lenses on my carbon dioxide laser. I don't know their relative permeability at near infra-red wavelengths, but this may be a way to go.

Also, may I trouble you to describe how are you receiving and demodulating the 56KHz serial infra-red?
 

Goeytex

Senior Member
The project was a simple proof of concept for my own learning since I had never done much with IR. I ordered some Fresnel lenses for the receiver side to optimize range inexpensively. However, I have not yet fabricated the hardware (lens holder, etc).

The modulation / demodulation is simple and probably not optimal for optical comms. The 56KHz carrier is On - Off keyed with 2400 baud serial data using an AND gate. This is sent to a FET that drives the 5 watt IR LED. The IR receiver does the demodulation and filtering and the output is sent directly to a Picaxe 20X2 with background serial receive.

As soon as I get the Fresnel lens receiver done, I will post more info.
 

jedynakiewicz

Senior Member
@Goeytex, thank you for the update. I was particularly fascinated with the 7 watt infra-red emitter that you are using- that appears to be quite astonishing in its output. I can't seem to find it with any of my normal suppliers over here in England. I have used high power, but nothing like 7 watt, emitters for the remote control of indoor models and I am wondering about the applicability of the type of device that you are using. Thank you for the "illuminating" insight into the method of modulation; most interesting.

@Boriz, a fascinating alternative angle on things. Could you perhaps elaborate on what type of laser you use for this application and also how you modulate it? What frequency of infra-red are you using and what device do you use to selectively receive and demodulate the beam? I would have imagined that there would be a great need for optics aligned with sniper-rifle precision , unlike Goeytex's approach which would require only the rough aim of an Elizabethan cannon to hit the target with certainty.
 

Dippy

Moderator
That's very important point jed.
I had to do an IR device (ah those were the days..) and had the same problem.
Too 'collimated' meant alignment problems and the slightest vibration at the Tx end resulted in an early bath.
Agreed, in a really robust system laser would be better.
Though do keep an eye on laser wavelengths.

In my old system I used a flat faced IR LED pulsing at 1.5A via a ~25mm acrylic lens manufactured by CTP-COIL (basically the same as used in their 'jewellers' 7x eypiece magnifier that you can get from Farnell search 'COIL').
The reciever end used the same lens onto a BPW34F photodiode - then onto my own gubbins.
In the receiver I used some shrouding/baffles as found in some telescopes to reduce ambient light.
Signal:noise being the name of the game.

The lens manufacturer (being a proper one) supplied wavelength/attenuation graphs and information on the focal length change with wavelength/frequency.
Armed with that info I was able to make a small, low power link albeit only over 50m.
Alignment was still a bitch.

And do get the specs on wavelength shift and Vf in IRLEDs as you turn up the wick.

I did experiment with a 50mm lens by COIL and got greater ranges but alignment was almost as painful as laser.
As suggested by Geoytex a larger receiver end makes this easier but always pay attention to keeping ambient out.
 
Top