How can I effectively reduce the field-of-view of an infrared sensor?

Hi,

I have a GY-906 breakout board with a MLX90614 infrared sensor that I want to use on my robot to detect humans and animals at range by comparing the ambient temperature and the temperature of object in front of the sensor (both measurements are provided by GY-906).

GY-906 picture: https://ae01.alicdn.com/kf/HTB1mioEaGagSKJjy0Fgq6ARqFXaz.jpg

The datasheet for MLX90614: https://www.sparkfun.com/datasheets/Sensors/Temperature/MLX90614_rev001.pdf

Problem is, this sensor has a wide field of view, and it's object-temperature measurement is the average temperature of everything that is within that FOV cone. Meaning that it becomes unreliable to detect a heat source (such as a hand) more than 20-30 cm away, because it only takes up a small part of it's field of view.

What I want to do is reduce the field of view of this sensor, so that it takes measurements within a cone of just a few degrees, and in effect, would be able to detect changes in temperature much further away.

There is a similar sensor that has some kind of a tube on top of it:

Problem is, it costs 5 times more than the one I have, and I'm kind of reluctant to spend that much money when IR sensors are pretty damn expensive to begin with. I can't find a datasheet to determine what kind of tube this thing has on top, and I'm wondering if this is something that I could make myself. Something that would limit the field of view.

I tried out simple things like using a straw or a rolled strip of paper to make a tube and put it on the sensor I have, but it seems to raise it's object detection temperature, and the sensor doesn't detect much past the end of the tube.

Can anyone give any suggestions how to achieve this?

EDIT: I was trying to post this in the Sensors sub-forum... I have no idea how it ended up here. I can't seem to delete or move it myself. Can someone from the moderators please move this topic to Sensors section? Sorry :confused:

Metal reflects far-IR, so anything metal used as a heat mirror will work, so long
as its a clean metal surface with no paint, anodizing or other surface layer.

MarkT:
Metal reflects far-IR, so anything metal used as a heat mirror will work, so long
as its a clean metal surface with no paint, anodizing or other surface layer.

Is that all that is needed here? Just a metal tube? No lenses or anything?

No lenses or anything

Lenses are difficult to make in the sort of passive IR frequencies you have here. Look at the Fresnel lens that are used to make a wide angle field of view for PIR intruder alarms.

ebay has parabolic reflector cigarette lighters. get one, put a stick in it see where the stick burns. that is the focal point. put your sensor at that point, and put the reflector in a PVC tube to exclude IR coming in at an angle.

this is my plan for a long distance break beam sensor. I will tell you how well it works in April or May.

this is my plan for a long distance break beam sensor.

Different wavelength of IR for beam breaker than for

MLX90614 infrared sensor that I want to use on my robot to detect humans and animals at range by comparing the ambient temperature and the temperature of object in front of the sensor

IR for beam breaker 850nm or 940nm

IR for passive heat detection 8-14 micrometres, that is 8000nm to 14000 nm which is closest to the infrared radiation emitted by humans. So a whole order of magnitude out, things are very different there.

laukejas:
Is that all that is needed here? Just a metal tube? No lenses or anything?

You want to buy a lens, if better be germanium, silicon, calcium flouride or polythene because
glass absorbs far IR.

Thank you for your answers. But how can I determine what kind of lens I need to buy to have the range and field of view that I need?

But how can I determine what kind of lens I need to buy to have the range and field of view that I need?

Basically you can’t. I would be surprised if such a thing exists for you to buy anyway.

What has this to do with Audio? Why is it in the Audio section!?

A lens for passive IR for use in thermal imaging is very heavy and very expensive in the order of $300

Grumpy_Mike:
Basically you can’t. I would be surprised if such a thing exists for you to buy anyway.

What has this to do with Audio? Why is it in the Audio section!?

A lens for passive IR for use in thermal imaging is very heavy and very expensive in the order of $300

I mistakenly put this topic in the wrong section, and I don't know how to fix it... Sorry.

In my original topic, I posted a link to a product that already has such a lens... Problem is, it is very expensive (compared to the sensor I have). Which is why I'm hoping to mitigate the cost by making something similar myself.

mistakenly put this topic in the wrong section, and I don’t know how to fix it

You click on the report to moderator on the first post and ask him to move it.

That wavelength of light will not go through glass so how do you propose to make one yourself? Why did you think it was expensive? Yes the sensor is the cheapest part.

Grumpy_Mike:
Basically you can’t. I would be surprised if such a thing exists for you to buy anyway.

You could try Edmunds Optics, but expect to see prices from 100's to 1000's of pounds/dollars/whatever,
but they likely have a lens that would do the job well (and be totally over-engineered for the task!)