Super sensitive light sensor

I am trying to detect light in a range from complete darkness to a tiny bit of light. Imagine my test set - a shoe box with a mobile phone, that way I can make it ring and get light. The room light is way more than present in the real situation. I could also use a water pipe and place a phone in the middle The light I want to detect is not natural, it will be LED and other light meant for humans (no IR or so). Also the space it will be used is very tiny. I cannot change the light provided in case some would suggest that.

The easy solution would be and LDR or photo transistor, some resistors and a comparator or ADC

I have tried with LDR's and found only a few that gives enough change on "a tiny change in light", however that work on my desk light on/off, but that is not the real case. I also found LDRs to react to light fast, but returning to dark is slow for some of them. This is ok, but I still found them too different to be reliable. I tried 5 of the same type and got 5 very different results. Once again, solvable in SW, ADC read min and max and react accordingly. But still....

I have tried one photo transistor which reads 860nm (~red), but it does not give any reading except 1-10mV (5V in series with 4K7), and a setup with an op-amp for amplifying it works with room light

Now, my transistor is at 860nm or red, and I am looking at present light, so I found another one that works around 550nm, which simulates the human eye: Datasheet:

formatting link
am considering ordering these for testing.

What can the community here suggest? I need a rest and new thoughts now

Reply to
sonnic...
Loading thread data ...

Photomultiplier tubes are very sensitive. They can count individual photons.

Or, you can drop the value of the resistor in your photo transistor circuit to a smaller value. What voltage level do you need?

Reply to
Ricky

The processing system runs on 5V. So getting something 1-4V would be useful, but the range is less important as long as I can get something reliable, 1-2V is fine, as long as I can process it down the line. I have played around with resistors from 1K to 2M2 and different setups. My brain needs a break and good ideas

Reply to
sonnic...

Actually, I told you wrong. A large resistor will give more gain. What voltage level did you see with the 2.2 Meg resistor?

Oh, only change one thing at a time to be sure of what you are testing.

Reply to
Ricky

LDRs are terrible.

A large-area photodiode would make a lot of signal, preferably back-biased current, not PV mode.

I have a bunch of First Sensor parts

formatting link
and I could send you a few.

Reply to
John Larkin

formatting link
is small and hard to beat for sensitivity, but they are noisy, and the electronics are trjcky, though they can be compact.

Reply to
Anthony William Sloman

A reverse biased LED is the starting point. Phototransistors are not sensitive, nowhere close to a photodiode. Please check out:

  1. formatting link
    formatting link
    formatting link
    formatting link
    Hope that helps.
Reply to
amal banerjee

This is nuts. A phototransistor is effectively a photodiode plus a transistor in which the photodiode current serves as the base current drive for the transistor.

You get more amps per lumen than you do with a photodiode, so it is more sensitive than photodiode. It's not great for measuring low light levels because the collector base leakage current adds to the light signal.

It probably won't.

LEDs are famously photosensitive, but they are not designed or specified for the job.

It would be interesting to see a LED used as single photon avalanche diode.

formatting link
offers a red LED with 12V reverse voltage which is easy enough to set up. Compact package too.

Reply to
Anthony William Sloman

What you'd really want, then, is a logarithm-scale sensor of visible light? Logarithm-of-current tricks with transistors can accomplish that.

Three or four differently-biased photodiodes, with different gains, and a bit of logic to splice the ranges together might work, too.

The phototransistor sensitivity to red and IR is good, but not to blue or green... some kind of filter is necessary to get insensitivity to IR, and a blue or green LED might be better sensitivity than silicon for the full 'visible light' range; TEPT5700 is a Vishay diode-filter that mimics human-eye sensitivity.

Reply to
whit3rd

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.