IR Receiver/Demodulator

One idea, common to all TV remotes, hasn't been incorporated yet; you can add a start pulse as lead-in to the sequence, and look at the modulated received signal for a few milliseconds, and set an AGC level. Then, blink the light off, then back on for the timed duration you want to determine.

Any demodulation high-Q filter will need the long start pulse to lock to the phase of the sender, and the phase shifts (and associated timing errors) are negligible AFTER you make each time-critical determination on the light-turn-off signal (rather than determining both an ON and OFF time).

Or if you wanted to avoid the impedance matching stuff and layout issues use the RFM12B module. Similar price, good performance. That module is available in 868 and 915MHz variants as well with a minor firmware change. Actually this Microchip part seems identical to the Si4421 chip used on the RFM12B with just some signal name changes!

Mark.

Better to use an unbuffered inverter.

I'd use a TIA with an inductor across the FB resistor so there's no DC gain.

Two things: Best not to have too much gain before the filter. If a reflection from somewhere saturates the stages before the filter then the filter ain't doing you any good during that time. Also, someone (I think John) pointed out that you data rate may be a bit high for a watch crystal. The BW is usually not much more than 10Hz.

You've got to cut DC right at the beginning. In a TIA via an inductor across the feedback resistor, if you want to use the shunt method then via a coupling cap.

The cutoff should be as high as you feel comfortable with. Yes, ambient light shows up as DC but, for example, a hornet buzzing by, birds, reflections by cars and so on, that won't be DC.

If you need 20msec pulses to make it through you should have a few (very few) hundred Hertz of BW. Looks like LC to me.

I agree. That or one of the generic cheap model car control modules which claim ranges roughly in the right ball park. Without error correction on the data sent this thing is headed for a fairly major train wreck.

Meade stock was once pumped up in the .com (con) boom on the basis of high bandwidth building to building coms using 8" SCTs as the optics.

Regards, Martin Brown

Terabeam, RIP. I recently bought 75 of their beautiful 1.3 um APD/TIA modules for about 75 cents apiece--probably half a cent on the dollar.

Cheers

Phil Hobbs

The data sheet comes with a layout and BOM.

But frankly I don't understand why he's only getting 8 ft range. I've never tested mine in direct sunlight but I know at dawn/ dusk I get a helluva lot further then that.

Even if you Google IRDA ,IR BEAM etc. you'll find all sorts of examples of people getting hundreds of feet transmission distance with no special optics. A fresnel lens diffuses so I don't think that's helping for one.

As for an IR lens that filters out sunlight and other wavelengths just rip one out of an old TV,STEREO whatever. All IR remote devices in my house have a filter lens in front of the detector. I'm sure you could find a scrap TV or something to scavenge one off for testing.

I recall finding sources for the lens you can get it in strip form. I cant find the site anymore but its inexpensive.

You can go for miles and miles in the dark, if your aiming is good enough. During the day, even if you get rid of the DC, you're still buried by the shot noise of the sunlight. It's horrendous. Try calculating your expected noise floor--it isn't hard, and it's very instructive.

Cheers

Phil Hobbs

Sometimes this stuff seems to work though:

It does, but maybe you'd rather not do it. The RFM12B is available at about $2.10 in 1000off and has already been compliance tested, it lowers risk that the final product is compliant. You'd still need the BOM parts with the MRF49X in addition to the chip and get them placed on the PCB. In larger quantities the MRF49X solution might be more cost effective, but I don't think there's much in it.

Mark.

Well, at least he could get rid of some of that if he mounts a snippet of tube in front of the photodiode. I've seen hangar doors where a rather crude light sense scheme worked reliably. Ok, not 80ft but at least 40ft.

You posted that the last time this came up. It's a great read. I really don't know why it is so hard to get people to do simple photon budget calculations even when their livelihood depends on the results.

Anxiety, I suppose....but surely it's better to find a good design easily than beat your head against a brick wall (e.g. the shot noise of sunlight).

Cheers

Phil Hobbs

I agree if your doing it for a commercial product in quantity you are far better of get a fully assembled and compliant module. I don't think that's the case here though.

Theres been some discussion here about RF compliance testing I dont remember exact figures but it wasnt cheap. You would have to sell a lot of your product just to recoup the testing cost. Not to mention time and expensive RF test equipment....

I did get a couple of them though. I like to play;-)

Another option is to visit a store which sells sheet plastic (e.g. TAP Plastics, Professional Plastics), dig through their scrap bin, and buy as thin a piece of black acrylic plastic as you can find. This stuff filters out visible light, but the dye in it passes a significant amount of IR.

I've taken some interesting photos, using an old Fuji FinePix camera whose Sony sensor has a less-than-efficient IR-exclusion filter, shooting through a simple home-made filter made from black acrylic glued to the end of PVC tubing. The results have the usual IR-photo appearance - the sky is quite dark, green leaves on trees appear almost white.

Sometimes referred to as "sunlight noise", like on page 8 under "calculated noise performance":

I realize you aren't going to get hundreds of meters range in the day but you should be able to get better then 8'.

They use IR for laser tag. I don't know what the military uses I'm guessing IR as well for their war games. These are used during the day.So it can be done. They also use proper optics to focus the beam you would have to point the emitter at the detector for reliable

The only way you are going to get reliable 50-200m range daytime transmission is with optics and aiming at the receiver which is shielded as much as practical from the sunlight using the LENS mentioned and mechanical methods like a tube.

Frankly for daytime I think you would be better off using a 433MHz RF module. For nighttime line of sight IR would be simpler and cheaper but for transmission day and night RF would be the most reliable.

He needs 80ft.

It can be done. I know for sure ;-)

If you want to use pre-certed modules that's going to bump up the BOM budget, big time.

Then he is going to have to collimate the IR emitter and use an optical filter optimized for his wavelength on the detector. It would also be very directional.

Is that what it is for RF compliance testing I'd go broke;-).

Given the cost of compliance testing it still might be cheaper to use modules? Unless you have a really large mark-up or plan on moving them in the 100k quantities.Particularly when you factor in assembly cost, testing etc.

Yup, except those guys are using 10 uW/cm**2, which isn't easy to achieve at any distance unless you're using lenses. They also assume 5 W/m**2 for sunlight, which means they're using a narrow optical filter--full sun is more like 800 W/m**2.

Cheers

Phil Hobbs