Building a Light Integrator for accurate alt proc exposure - discussion of project...

[nick mulder]

Hello,

I'm going to build a integrating circuit for light - i.e. it will tell me how much light has fallen on its sensor over time, a kind of cumulative effect...

It will primarily be used to help with Pd exposures in a cloudy sky ...

You will see a bar-LED graph (like on your stereo) go up as the paper exposes, shine no light on it, it stops going up - shine bright light on it, it goes up faster etc... do this until some set mark and your exposure is complete (and the same as the last one even though the time might have been different) - get it ? yeh, cool huh

A quick issue right off the bat that I am yet to answer is that I will need a 'UV pass' filter - i.e. the opposite of a standard UV filter - I want UV and the rest of the spectrum to be gone so as not to add noise to the sensor readings with differing color casts in passing conditions (eg. less UV in clouds etc...) - A good example is that without such a filter it would 'expose' under your working tungsten lights which as we know doesn't affect your Pd coatings anywhere near like UV does.

Does such a UV pass filter exist at a cheap cost ? hopefully just a pull apart of some redundant trash-heap gear? (please!)

For those of you who know a little electronics - it will be an integrating op-amp with a photodiode sensor (more linear than LDR's) - the output display could be anything really, I was thinking a bar-LED system with one of those multi-comparator/LED driver packages.

Pretty simple to begin with...

As for non-linear/reciprocity issues I can 'fix' them with a subtle massage of the sensor output (maybe even use the non-linearity of an LDR if it suits) - many methods here and a lot of fiddling for exact results and that is where the bulk of the design will reside - but in reality most of the many many axes in the graph of Spock Chess will be set to a constant that is within the error specs of the application, which is pretty rough at this stage...

Maybe a uP version could be made with the math/translations done inside ? could then have an LCD for feedback - thats V.2 for now ...

I know these bits of gear exist already off the shelf - never seen one and am unaware of how they work operationally - any info appreciated...

We can keep it opensource too - build one for yourself

fun,
Nick

 

[Steve Smith]

For those of you who know a little electronics - it will be an integrating op-amp with a photodiode sensor (more linear than LDR's) - the output display could be anything really, I was thinking a bar-LED system with one of those multi-comparator/LED driver packages.

This is the way the simpler flash meters work. just a simple integrating circuit and a sample and hold circuit with a storage capacitor but working on a much shorter time scale. I started to build one myself but then bought one on e-bay which was cheap enough for me to give up my self build plans.

If you add a processor to it then the software can do all the integrating and the linearising as you point out.

I can't think of anything to use as a UV pass filter though (plenty of ideas to block UV).

Sounds interesting though. Keep us updated


Steve.

 

[nicolai]

I've been trying to design an integrating meter + shutter controller as well. I initially wanted to do it with a Palm PDA, but ended up ruling that out. My current thinking is that the Arduino board is the way to go. It's got both analog and digital I/O and has a simple C-like language and a sane API. And at $35, they're affordable.

My plan was to fire a solenoid to open the shutter, sample the voltage from a photodiode (which is easy, because it has a nice ADC, so you can programmatically read analog values as integers), integrate it, and close the shutter again when it had gotten enough light. It would be fairly trivial to add an ISO selector, which could be implemented as a sample multiplier. But I'm a programmer, not an electrical engineer, so this may not be as ideal a platform for you as it seems to be for me.

(As soon as I get a board, my first project is going to be a shutter speed tester.)

I have no idea about the UV-pass filter, though.

 

[Steve Smith]

Something similar (perhaps):

http://www.f295.org/Pinholeforum/forum/Blah.pl?b-cm/m-1154672610/s-0/



Steve.

 

[nick mulder]

I have been thinking about getting an Arduino for a while as I am now an owner of Max/MSP/Jitter which can use the Arduino as an I/O board to the real world, lots of interesting possibilities there ...

But the fact is I already have an OOPIC - although the languages are different, we can petty much have the same conceptual conversation. I thought it would be neat to attempt a analog circuit first, I like using the 'proper' stuff first before reverting to uP's .... Something of the APUG'er in me

But Steves link has revealed pretty much exactly what I wanted to build!

http://www.flickr.com/photos/92561604@N00/sets/72157594223607004/detail/

Looking pretty cool huh...

...except instead of the shutter I'd have a visual output like a progress bar - thinking more resolution in steps near the end of the exposure also and maybe an audible warning. Although user interface is usually simple stuff to activate, its endless in its possibilities, and therefore daunting.

Ever tried designing and coding a user interface/menu structure for a 16x2 LCD ?? turns out its not fun

 

[bobherbst]

Bulding a Light Intetgrator

Save yourself a lot of time, money, and frustration and purchase an Olix AI121/131, Douthitt 985 or other brand of integrator on ebay for $25-40. Purchase a photocell from Olec Corporation or on ebay and you are done. These units were designed for commercial print shops for the very purpose you describe. I even modified a photocell for use with a 100 watt bulb over a vacuum frame for use in making my AZO contact prints. I purchased a lot of three integrators for $75 on ebay and then sold one for the same amount giving me an AI121 and AI131 free.

Bob Herbst

 

[nick mulder]

Save yourself a lot of time, money, and frustration and purchase an Olix AI121/131, Douthitt 985 or other brand of integrator on ebay for $25-40. Purchase a photocell from Olec Corporation or on ebay and you are done. These units were designed for commercial print shops for the very purpose you describe. I even modified a photocell for use with a 100 watt bulb over a vacuum frame for use in making my AZO contact prints. I purchased a lot of three integrators for $75 on ebay and then sold one for the same amount giving me an AI121 and AI131 free.

Bob Herbst

Could be an option too for a basic integrator, do the photocells pass only UV ? Which products do you refer to at the Olec site ? I didn't see any photocells immediately in the products section

I do have all the bits here already for the project though, its pretty generic stuff - time would be the only cost...

The integrators from ebay, do they have reciprocity compensation functions ?

With a homebuilt uP version you could enter film characteristics directly from the manufacturers specs/graphs at whatever resolution the processor could cope with - that'd be neat, mind you for my application the film specs is my own coated paper and its not print-out so it will be a hassle to create the data

 

[Jon King]

Save yourself a lot of time, money, and frustration and purchase an Olix AI121/131, Douthitt 985 or other brand of integrator on ebay for $25-40. Purchase a photocell from Olec Corporation or on ebay and you are done. These units were designed for commercial print shops for the very purpose you describe. I even modified a photocell for use with a 100 watt bulb over a vacuum frame for use in making my AZO contact prints. I purchased a lot of three integrators for $75 on ebay and then sold one for the same amount giving me an AI121 and AI131 free.

Bob Herbst

I agree if you just want an UV integrator, ebay will be far easier. On the other hand, if you want to roll your own as a project, a photodiode like this:

http://www.advancedphotonix.com/ap_products/pdfs/PDU-G105A-SM.pdf

might be an easier starting point. If you want to go the analog approach, one way would be to use the diode as the input to a current to frequency circuit. Your uC is then just left counting pulses, driving the display and watching switches.