Has anybody tried making a DIY colorimeter?

[Whitebeam]

I've just been searching the web for colorimeters and came across this page giving details of how to make your own:

http://www.home.zonnet.nl/rsetteur/aquarium/karel/colorie/coloriemeter_eng.htm

I'm just wondering if anybody here has tried this?

The scientific tecchie type in me is in charge just now ;-)

Peter

[Biollante]

Hi Peter,

Yes, I built a colorimeter a few years ago based on a PIC 16F88 microcontroller.

Took some serious messing around to get the thing to work right, finally a photographer type realized the need for a filter and a bit of a reposition of the cuvettes holder.

The other problem was the use of a light dependant resistors (LDR), this may have been fine for the original purpose but we require higher sensitivity. I ended up using TCS230D Color Sensor, photodiodes. :p

It worked pretty well; I still ended up buying a ridiculously expensive colorimeter.

Overall, for the basic tests most aquarists perform a good quality high color rendering index (CRI) lamp, makes as much sense.

Colorimeters in my ever-humble potted-plant opinion are great if you need to do a large number of tests with high repeatability or need to a very wide range of tests. Some relatively inexpensive colorimeters on the market are decent, for most hobbyists a good quality CRI lamp is a better investment. :gw

Biollante

[Oreo]

Just like with regulators, EBAY is your friend. I picked up a like-new Hach DR2000 for IIRC, $150 or so. I love that thing. But you have to really keep your eyes open for them. There will be long periods where there's not much on ebay or what is there is over-priced. Then one will show up out of the blue for a killer price. I've seen Hach DR890's go for $400 which is a hell of a deal on the most modern equipment.

My personal opinion: Don't try to build one. By the time you count all the parts & your time involved it's worth it to just spend the money on good equipment that is guaranteed to produce good test results.

Biolante, FYI in my DR2000 they use a prism instead of color filters. I think on the DR890 they use combinations of colored LEDs.

[Whitebeam]

I think I'll give it a go as, apart from the LED and sensor, I have all of the bits needed here and can fairly easily turn a chamber out of hardwood. I'll try the version that uses an amplified photodiode to give a better response - the Dutch text describing that enhanced version is available as a link ('prototypen') from the original site I quoted above. I'll let you know how I get on.

Peter

[Whitebeam]

Yes, I built a colorimeter a few years ago based on a PIC 16F88 microcontroller.

I'm curious what the microcontroller was for - was this to convert from voltage to ppm to prevent the need for a calibration curve?


Took some serious messing around to get the thing to work right, finally a photographer type realized the need for a filter and a bit of a reposition of the cuvettes holder.

I'm also curious about that filter. What sort of filter and what was it for please?

Peter

[Biollante]

Hi,

The microcontroller allowed me to run automated calibration routines as well program a number of routines. I used a 128X64 LCD screen and could store results and download or stream to a computer. :)

The photographic filter involved a compromise, if I recall correctly, I was having a problem with the 520-nm range. The filter corrected to give consistent accurate results.

As Oreo noted the superior method would be to use a prism. :cool:

I think were I to build another I would simply use a USB interface and let the computer handle the processing and allow for more routines. I would not try to make it portable. :D

Biollante

[Oreo]

Biollante, what kind of bulb are / were you using? Or, maybe the better question to ask is, does it make a difference? Will an incandescent bulb produce a full enough color spectrum that a prism or filters would provide enough of a signal to be measured? I know some spectrophotometers do require special bulbs but I don't know that much about the bulbs.

[Biollante]

Hi Oreo,

I used four 10 mm ultra-bright LED’s, I think I ended up with 30 degree viewing angle, I wanted 10 degree viewing angle LED’s but couldn’t find them. :(

The LEDs were listed as something like 3500-8000 Kelvin. :) My first attempts were with crappy LED’s. :(

I do not remember exactly but had a tough time in 510-520 nm range.

Biollante

[Oreo]

http://cgi.ebay.com/Hach-DR700-Colorimeter-/260650458693?pt=LH_DefaultDomain_0

DR700 for $75. Even comes with some glass vials & a 610nm color module. Only down-side here is finding any other color-filter modules you'd need but that's not the end of the world.

I use one of these where I work. Not as nice as my DR2000 but it'll do just fine for the aquarium hobbyist.

[Biollante]

Hi Oreo, Peter,

The reasons for wanting to diy something are not all practical or financial; :) sometimes it is a way to understand how or why something works. I know I learned a lot from my project. It also made me really, reeeally appreciative of my Hach colorimeter! I think my little diy colorimeter is still in use. :)

I saw a diy colorimeter awhile back that was really just a box with, I think four LED’s, definitely red, green, blue and I think perhaps white. The cuvette is inside the box it appeared to have a decent sensor (the fellow using it was not the builder or helpful for that matter). It had a simple rotary switch and used a voltmeter for the readout. It appeared you had to make some guesses and do a little figuring, but seemed workable and I would not think more than US$ 30 or so even if you had to buy everything.

{Rant Alert…Rant Alert}


Admittedly, lacking a coherent central nervous system and therefore a brain makes it difficult for me to understand things, but I really cannot understand why a cheap basic colorimeter is not available to the hobbyist.

{All Clear}

Biollante

[Oreo]

Hi Oreo, Peter,

The reasons for wanting to diy something are not all practical or financial; :) ...

...I would not think more than US$ 30 or so even if you had to buy everything.

Admittedly, lacking a coherent central nervous system and therefore a brain makes it difficult for me to understand things, but I really cannot understand why a cheap basic colorimeter is not available to the hobbyist.

BiollanteAs someone with an extraordinary amount of DIY experience, I totally support taking on the project for reasons other then practicality or expense. My weakness is fine electronics though so building a photospectrometer from scratch just seems daunting. Time involved would be the biggest expense by far, but if you enjoy the project that's not an issue.

I also agree that you'd think a decent hobbyist grade colorimeter designed specifically for use with aquariums would be available. They're so far superior to color charts & dip-strips it's hard for me to imagine how people get by without one.

[Whitebeam]

I saw a diy colorimeter awhile back that was really just a box with, I think four LED’s, definitely red, green, blue and I think perhaps white. The cuvette is inside the box it appeared to have a decent sensor (the fellow using it was not the builder or helpful for that matter). It had a simple rotary switch and used a voltmeter for the readout. It appeared you had to make some guesses and do a little figuring, but seemed workable and I would not think more than US$ 30 or so even if you had to buy everything.

Yep, that's the sort I'm making. The LED has all three colours in one diffused package and I'm using three seperate switches for these colours so that I can combine them if necessary. Even if it doesn't work perfectly, I'm having fun building it and it will also make me produce standard solutions and hence 'calibrate' the test kits that I have. The main challenge that I'm facing just now is how to ensure that I drill a hole through the hardwood cuvette chamber that I've turned and ensure that it goes through perfectly axially. Suggestions for 'better' cuvette chamber designs would be appreciated ;-)

I'll write this up as a DIY project once I'm done.

Peter

PS: One other reason for not buying one is that the wife will probably kill me for spending yet more money on the tank! A few electronics bits and an hour or so in the workshop goes in under her radar ;-)

[Oreo]

You need to search on youtube for a video of a machinist making a rifle barrel out of a piece of bar-stock. The trick is to rotate the piece of bar-stock, not the drill bit. This way, even if the drill bit is off-center, it still cuts a hole centered to the bar-stock.

[Whitebeam]

Yep - that's how I drilled the hole for the cuvette - bored it out on my lathe. It's the transverse hole that's worrying me - the one for the light path. I need to ensure it goes straight through a diameter of the central hole (if you see what I mean).

Peter

[Oreo]

I see what you mean but I don't see how it's that critical. Drill the hole over-size, fit the electronics, and then seal it up with some black silicone. Since you're building this thing from scratch you're going to have to calibrate anyway and calibration will compensate for the length of the light-path.

OOOooorrrrr...... You could consider a square sample cell & chamber. Makes the whole process a little easier.

Lots of options really, depending on your sample size. But going with hach's standard 10ml or 25ml sample cells would give the advantage of being able to use their high quality reagent sets.
http://www.hach.com/hc/search.productByKeyword.invoker/EntityType=PRODUCT/PageNo=2/RequestOption=ALL/NewLinkLabel=/Keyword=sample cell (http://www.hach.com/hc/search.productByKeyword.invoker/EntityType=PRODUCT/PageNo=2/RequestOption=ALL/NewLinkLabel=/Keyword=sample%20%20cell)

1" square polystyrene sample cells. Set of 2, or set of 12:
http://www.hach.com/hc/search.product.details.invoker/PackagingCode=2410222
http://www.hach.com/hc/search.product.details.invoker/PackagingCode=2410212 <----------------------- This 12pk is the best option IMO

1" square glass sample cells. Matched or UNmatched pairs marked with 25ml, or 10ml, or both.
http://www.hach.com/hc/search.product.details.invoker/PackagingCode=2095000
http://www.hach.com/hc/search.product.details.invoker/PackagingCode=1353702
http://www.hach.com/hc/search.product.details.invoker/PackagingCode=2495402

1cm square sample cells.
http://www.hach.com/hc/search.product.details.invoker/PackagingCode=2095100

[Oreo]

Another couple thoughts,

You could go on ebay & buy one of the cheapest colorimeters just to cannibalize the sample cell holder (and what ever else looks useful.) Usually, Hach & Lamotte single-chemical test meters are pretty cheap. If you find a non-working one you could probably get it for the cost of shipping.

Or, if you're going to be making this from wood, consider lining the sample cell holder with some black velvet. It will help to keep the light-path clean without a lot of internal reflection as well as to hold the sample cell snug without scratching the plastic / glass.