# Analog signal to modulate LED intensity

Discussion in 'LEDs and Optoelectronics' started by Samote, Dec 8, 2012.

1. ### Samote

6
0
Dec 7, 2012
Hello. I am very new to circuits, and jsut want to get my feet wet with some projects.

I was reading up on parts from a local electronics store, and I came across this part here: http://www.abra-electronics.com/pro...2d237S%2dLF-Light-to-Frequency-Converter.html

I was wondering if that could be used essentially as an analog-to-digital converter, where I would modulate an analog signal with the intensity of a light source, and I would pick that up in some sort of fibre-optics set up with the converter, or merely the two in some dark box, in proximity to one another.

I merely want to know
A. In your experience, are such converters consistent? (I can calibrate for different values, that is not the issue, as long as they are consistent)

B. If this idea is realistic, what type of light source should I be using? Would LED's suffice? Do I need a special light source?

Last edited: Dec 9, 2012
2. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,501
2,841
Jan 21, 2010
It's interesting. But you'd have to look at the datasheet for the device to determine its linearity.

Then you'll have to determine how you can convert the signal to a varying light level in a manner that is also linear.

Finally, you'll need to convert the frequency back to a linear signal in a way that it approximates the inverse process (to reduce rather than enhance non-linearity).

It's certainly achievable, but the questions are:

1) how much distortion can you tolerate?
2) How low can you get the distortion?

If you can get (2) lower than (1) you're OK.

edit: But I have to ask -- why not simply frequency modulate a carrier frequency? It's pretty much the same thing and the principles are well known.

Last edited: Dec 9, 2012
3. ### Samote

6
0
Dec 7, 2012
Would it not be possible to make a set up, then run known voltages through the light source, read the frequency off an arduino board and basically make a custom data sheet, approximating it with some function in the desired range?

4. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,501
2,841
Jan 21, 2010
Sure. But why do you want to do it the hard way?

Do you understand that manufacturers make datasheets so you don't have to do your own (possibly flawed) analysis?

5. ### Samote

6
0
Dec 7, 2012
Yes, I understand that, but I feel that I would get more accuracy if I calibrate the specific part I get, and also, I would have to cross-reference the datasheet for both the light source and the light detector, and compensate fro light lost in the transfer process.

Either way, I wanted to know if this would be a decent way of making an analog-to-digital converter-and-opto-isolator for feeding an analog signal to an arduino board.

6. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,501
2,841
Jan 21, 2010
Sure, it's "a" way of doing it.

As to whether it's decent, that is not so easily stated.

If you can tell us what you're trying to do (purpose of the signal, source of the signal, need for isolation, etc., than we may be able to assist more.

In general, I would think that isolating the "digital" signal would be better than isolating the analog signal.

Also, the arduino can read analog voltages directly, so the conversion to frequency seems a little unusual. Perhaps a simple optocoupler with some calibration would suffice?

7. ### Samote

6
0
Dec 7, 2012
Oh, an arduino board can read voltages...
Well then, I did not know that. In the long run, I am planning on making some form of EEG, but for now, I am just getting used to working with electronics, seeing if I can pick up signals, what kind of circuits are better than others, what can I do physically to the signal, what can I do through code, etc.

Last edited: Dec 9, 2012
8. ### KrisBlueNZSadly passed away in 2015

8,393
1,271
Nov 28, 2011
That's an interesting device, but not very useful for what you want to do.

If you're making an EEG, then you need to be extremely careful with isolation. You will need a very robust isolation barrier, and you will need to pass power across the barrier (to the circuitry on the "patient" side), and data across the barrier (in the other direction). Since you'll need significant circuitry on the patient side of the barrier to amplify the tiny signals and remove hum and interference, you could include the analog-to-digital conversion on that side, and pass the digital data across the barrier - serially through a high-speed optocoupler, for example. Or you can isolate an analogue signal using one of several techniques - linear optocoupler, capacitive transfer, etc. There may be safety issues to consider.

Google some relevant keywords, including isolation barrier, isolated, analogue/analog, medical, EEG, etc. This stuff has been done before, obviously.