IR Receiver/Demodulator

Only if the house is the temperature of melting lead. Room temp stuff glows in the

This guy's face probably isn't more than 100F.

Looks like his server is down right now.

I'm not sure that the whole process is anywhere near sensitive enough for thermal IR, although it's certainly possible that a good digital SLR, with its internal IR filter entirely removed and an IR-pass filter added, might do the trick with a long enough exposure and a cold enough sensor.

Not unless the house is on fire. ;0

Cheers

Phil Hobbs

Seems to be working pretty well here:

He could even see where the veins run underneath the skin and they can't possibly be on fire :-)

Mighty good reading.

No, it's just that the skin is much more translucent in the IR. There's almost no temperature contrast between veins and skin, and if there were, the veins would look brighter since they're carrying heat from inside the body to the extremities.

Trust me, you can't see thermal luminosity of 300 kelvin objects with a silicon sensor. Even if you put a microbolometer array or a HgCdTe sensor there, the glass lens of that camera is completely opaque in the thermal IR.

Cheers

Phil Hobbs

PS: you get the same sort of view with an S-1 photocathode IR viewer. It's neat.

Yeah, but he's shining an IR flashlight at his face, horror-story style.

I wonder what the inside of an oven looks like? Might be hot enough to catch the very tail of blackbody radiation.

Tim

Informal energy audits you can more or less point at any single glazed windows and the roof as prime suspects...

No but at near IR wavelengths scattering in the skin is low enough that it is almost transparent. Photon energy for near IR CCDs is characteristic of a very warm house - about that of molten lead.

Pyrodetector based burglar alarm sensor and a rotating mirror might allow you to do some ad hoc long wave IR measurments on houses. They will reliably detect a 37C human against a background ambient of 25C so you could build an image up line by line.

Regards, Martin Brown

But it is being illuminated by the solid state IR led array he built as a part of the project. Yes you can film in the dark with suitable amateur kit without startling wildlife. But you are not seeing any temperature differences unless the thing is *very* hot. You are just seeing how well the object reflects near IR wavelengths.

Notably trees and foliage are white. If you have some near IR long pass filter and look through it for a while (about a minute) with all other light excluded your dark adapted eyes will see a little bit into the near IR with a sort of odd false colour effect.

Typical imaging CCDs with their IR filter removed will go out to around

Regards, Martin Brown

is

other

transmitter

which

path

Well, due to the nature of the cell you will be lucky to get much more = than=20

that.

message=20

filter=20

hands

Pierce

other

acting

they

capacitance of

=20

=20

you=20

effects.=20

for=20

but at=20

pulse=20

=46irst OP would need a battery that can deliver the current pulses, = cr2032 cannot.

First OP would need a battery that can deliver the current pulses, cr2032 cannot.

True, but a capacitor can. The CR2032 would only be topping up the capacitor at relatively low current (depending on mark/space ratio). I've no idea whether this would work though in practice.

Mark.

cannot.

So put a capacitor across the battery[1] and it will supply the current spike to LED. Duty cycle can be very low for LED drive, as it's the peak signal power that provides contrast (signal) at the receiver, not average power that gets swamped by ambient light.

If there's room you could stack a couple coin batteries for 6V to get more LED peak current from cap. Or, perhaps a voltage double charging the capacitor? Lots of options.

Grant.

Yes, I was thinking of something like putting a large cap (seveal hundred uF or larger) with a small series resistor just to limit the pulse curent and connecting it to the resistor/diode juction of the schematic posted:

The receiver photodiode has to be fast enough to respond to that small pulse. After that you could filter out the 32kHz, but the gain of that stage may not need to be as high so background noise is reduced.

Mark.

Other possibilites for the receiver might be to use a tube, cone or even a parabolic reflector made from IR reflective material. If you use the reflector you'd place the receiver pointing inwards at the focal point. This effectively produces gain as it captures power from a larger area. An IR transmissive filter on the receiver (worthless to ad that to the transmitter!). The LED already has a tight beam angle so no need to modify that.

Mark.

Make sure the capacitor has low enough ESR, else it might limit your pulse amplitude to some unknown value. Also, measure the voltage dip on the supply rail because if too deep you crystal oscillator might choke.

Photodiodes are plenty fast if the connected electronics are. You should have no gain at all at DC, this is very important. Not even in the first TIA stage.

The tube is an absolute minimum. I don't think the scheme will work at all during the day if the photodiode gets direct sunlight.

Bit more complex would be to use an inductor to transfer power to the IR LEDs. See figure 2 in

I found two PDs suit this: BPV23F and BPW41N both look plenty fast enough and are filtered to cut response to visible light. For ~950nm operation.

As far as a receiver directional filter goes, there's mention of a honeycomb type filter (I think it got a mention in one of referenced documents upthread).

I forget the name of the beast but there's a technique used for flash photography that uses a bunch of straws in front of the flash to limit distribution of light to a small circle. Something like that could be used to limit reception angle to the receiver? The shorter the straws, the wider is the reception angle.

Grant.

You still may need a DC loop to get the receiver from saturating on ambient. Here in AZ, that seems an absolute necessity. ...Jim Thompson