Finally!!! getting the numbers I wanted with matching cells

[Scenaria]

kind of curious about something...


when I check em the number isnt always exactly the same... 5 to 10 ms deviation durring 3 or 4 triggers...

I guess as long as I nail two that are within the same 5 to 10ms range all should be ok?

 

[Ethan]

I've noticed that cells will vary depending on the last time they were exposed to light (and the duration) in how they will read in a subsequent test (for static resistance and decay). So here's my suggestion FWIW, leave all your LDRS in the dark over night. Then, only when you're ready to test them take them out individually. This might be a pain with using your "reference cell" but it's just a matter of how precise you want to be. Once you've sorted out the ones you think fall into your "acceptable range" leave those in the dark over night and do the same with those once again. I probably haven't tested as many as CJ has, but I would guess that he's also observed that the cells tend to "remember" the last time they were exposed to light.

 

[Scenaria]

the reference cell is being stored in one of the channels of a storage scope

I just realized it would be cool to have 3 channels... one for the ref... and the other two for matching pairs agains the ref...

 

[CJ]

Your referance cell should be stored in exactly the same conditions as the cells getting tested.
That way, they all move around togeather.
Welcome to the world of drifting cells!
It will drive you nuts.
1 day in the dark is different from two days, etc.
Then, it depends on how much light you use while testing.
If you do on resistance checks at 8.6 k, the dark storage thing gets erased pretty quick.
I do the upper resistance check first, which requires less light.
This way, you get better numbers as the cell hasn't been blasted with light.
cj

 

[Ethan]

I wonder if there's a lightwave frequency that they're not sensitive to (like a photo dark room light) so that you could do all your testing with all the cells in that light (after a day/more in the dark)...

 

[Scenaria]

did you find any out of whack with the upper R?

I noticed in the notes that they didnt even check Roff

 

[Scenaria]

peak wavelngth is 550nm but it looks like most cant deal with anything over 900nm and under 400nm

perhaps infrared or uv and if that doesnt work maybe gamma rays :green:

perhaps a paint that is visable under infrared light

 

[CJ]

I usually work at the darkest level where I can still see what I am doing.
And don't have the TV on, when the screen changes from dark to bright, it can zap your cells.
Same thing goes for Lava Lamps!
Only a few cells were off so bad at the upper end where I had to toss them.
If they match up decay wise, the Roff will be about the same most likely.
I check them at a light level that makes them read 180 k average.
:guinness:

 

[Scenaria]

I can just toss everything in a box and keep it closed...

reach in through a flap to hook up the leads... and close it...

im sure I can keep em in total darkness...

the only thing thats a pain is having to narrow em down a few times... its gonna take a few days BAH!

 

[CJ]

I used the box they come in for an organizer tray.
Write down 6k, 6.1,6.2, etc all the way up to 9 k .
Then you can test potential pairs for matching etc.
It is best to get pairs the same day.
That way, they won't move around on you.

 

[Gus]

What color light are they most sensitive to? I would use a led array of a color that the cells don't "see" well to test them. Kind of like how humans can use red light and still still see well in the dark.

 

[FredForssell]

Aren't opto cells fun?

One thing you might want to think about is applying a bias to the light source of your opto so that it is always turned on to a given point. You can adjust the bias from cell to cell to produce some desired test criteria like static resistance, or measured attack (on) time, etc. While this may vary over time (aren't opto cells fun?) it can help to even out the cell to cell variations.

I know of one manufacter (and a pretty smart guy) who drills a hole in the side of the opto package and inserts a seperate LED (non-visual wavelength) to bias the cell to a specific point. He claims (and I believe him) that he can get faster attack times with the LDR optically biased on to a small degree.

Of course if you take this approach to match your cells, then the design in which they are used has to take the same approach. But none-the-less it is something the think about...maybe.