Linear light-sensing circuit required

May 22, 2005 5 Replies
K

Hi all.



I'm trying to make a circuit to sense the whiteness of a surface. Like white = 5V, black = 0V, and linear between.



I bought an SFH2030 photodiode which gave about 20 mV difference measuring the reflected light from a cheap red LED.



Googling the web for a good circuit, I found this one



formatting link



which is a bit overkill, I don't need to measure starlight.



Corresponding with the author, he said he still wasn't getting the circuit working satisfactorily.



Anyone got any good tips about working with photodiodes? Like whether they are best run in current or voltage mode, good transistor circuits, do they need particular op-amps (most circuits I see seem to insist on high-impedance CMOS op-amps not popular ones like LM741/LF351 etc.



Cheers, K.


Vote
Twitter Facebook LinkedIn
J

Hello Kryten,

AFAIK you should stay below 75% of saturation to avoid the more non-linear range. Infineon is pretty good in specsmanship even if their web site doesn't appear to be one of the best.

The 741 amp is really old. CMOS opamps aren't that expensive anymore. If it has to be rock bottom in cost you could use the LM324 with a FET in front and clamp away (auto-zero) the DC drift and offset. The 324 also comes in a thrifty low voltage version which is lean on supply current.

Regards, Joerg

formatting link

Vote
Twitter Facebook LinkedIn
S

Use in current mode to be linear. Just need a transimpedance amp. Basically, one op amp and one feedback resistor.

Steve J. Noll | Ventura California | The Used Equipment Dealer Directory: |

formatting link
| The Peltier Device Information Site: |
formatting link

Vote
Twitter Facebook LinkedIn
K

Yes, I felt this was the better way to go

I was thinking of tying cathode to + input, anode to - input, resistor between output and - input. Also tie + input to idle voltage (as a virtual earth).

Like so:

formatting link
from article
formatting link

AFAICS, this ought to do it: the op-amp should pull the output voltage lower as the PD tried to force current into the - input node.

I guess this is what you meant too, and not one of those unusual transconductance op-amp thingies.

K.

Vote
Twitter Facebook LinkedIn
J

On Sun, 22 May 2005 00:05:56 GMT, "Kryten" wroth:

You should have kept on Googling. Take a look at

formatting link
to see something that will work right out of the box. There's a bunch of good application data in there too.

Jim

Vote
Twitter Facebook LinkedIn
A

If you really mean "whiteness" you will need to use white light, not red.

~ Adrian Tuddenham ~ (Remove the ".invalid"s and add ".co.uk" to reply) www.poppyrecords.co.uk
Vote
Twitter Facebook LinkedIn

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required