I have the interest to build a UVB meter for measuring the ultraviolet radiation intensity in the wavelength range from 280 to 320 nanometer. It will be used in my reptile house. The reptiles need UVB light for making vitamin D for bone health, but over doze will hurt them.

UV meters are available as laboratory instruments and are quite expensive (hundreds of dollars).

I think I can meet my target with a simple circuit. What I need is a UVB detector diode plus a high gain current amplifier. May be a budget of $50 will suffice?

Any advice please, on the detector diodes and circuits?

startwith the datasheets of some photodiodes..

RS sis do some UV photodiodes perhapps you should start there.

About the diode try this:http://www.boselec.com/products/docu...0photodiode%22
http://www.scitec.uk.com/uvphotodiodes/
About the first link..it describes a UV photodiode with integrated amplifier which I've seen in a industrial made apliance.
U might manage to go under 50$ for the gadget. But u'l stil need to borow a industrial made one for comparing&adjusting readings.

I've also seen some more cheaper&available Osram photodiodes.

Hi every body

At last I have been able to find locally a supply of UV sensor diodes OS100 from Orion Semiconductor: http://www.orion-semi.com/orionsemi.html

I have tried it out, giving a few mV output when shined with a small UV lamp used for checking bank notes.

Will look for suitable circuits to make a handy UVB Meter.

Again any advice please with regard to the circuits, the curent amplifiers....and how to calibrate it evetually ....... ?????

One can classify methods to determine the amount of light into three categories: 1)Those that convert the light to heat and measure the temperature change (thermopiles); 2) Those that use a chemical reaction with a known quantum yield (i.e., molecules of product per photon of light)(actinometer); and 3) Comparison with a reference meter or reference source.

Since I don't know anything about the resources you have available, it is hard to advise. If there is a nearby medical university, you may check with the dermatology department to find out if they have a reference instrument you can use.

If you cannot find a reference instrument, then actinometry is considerably easier to use than constructing your own thermopile. The most popular actinometric system is based on the photochemical conversion of ferrioxalate to ferrous ion. You will need access to a spectrophotometer to measure the product yield.

Good luck. John

References:
1) Hatchard and Parker, Proc Roy Soc (London) A235, 518 (1956).
2) http://www.ncnr.nist.gov/userlab/pdf/E136uvpaper.pdf
3) The Iodide/Iodate Actinometer in UV Disinfection: Determination of the Fluence Rate Distribution in UV Reactors. Photochem Photobiol. 2006 Mar-Apr;82(2):611-5.

Originally Posted by Olihou

Hi every body

At last I have been able to find locally a supply of UV sensor diodes OS100 from Orion Semiconductor: http://www.orion-semi.com/orionsemi.html

I have tried it out, giving a few mV output when shined with a small UV lamp used for checking bank notes.

Will look for suitable circuits to make a handy UVB Meter.

Again any advice please with regard to the circuits, the curent amplifiers....and how to calibrate it evetually ....... ?????

U'l need a voltage amplifier and a opamp or a couple of transistors (I'd use a FET or MOS-FET cause of much higher imput impedance) will do. U'l need to amplify the voltage like 300X or so. The output of that amp goes to the analog imput of a uC (wich wil take care of the display to).
U'l need to compare the instrument to a industrial one.Just get some UV intensity (read on the ind meter) =f(V) ..voltage read at the autput of the amplifier. After that get de UVi=f(V) function (u can use a software designed for that) and let the uC do the math. U might get around 50$ if u use a LED display.
Keep me posted.

Originally Posted by Olihou

I have tried it out, giving a few mV output when shined with a small UV lamp used for checking bank notes.

Probably because it's a UVA/Visible violet source, but UVB. If it's UVA you're after then most UV LEDs will work quite well as photodiodes.

Thank you all for the wonderful suggestions.

I think I will use a high-input-impedance operational amplifier to amplify the tiny current from the UVB sensor diode (only 250 nA at 35 uW/cm2). TL081 is a possible choice but probably it will not work well at a supply voltage of 3 volt, which I intend to. I wonder if there is any better OP Amp for this purpose ?

For display I will just use a small microammeter or simply a DVM to start with.

As for calibration I will use a reference meter, if I can get one. The thermopile and the chemical methods are beyond my capability indeed.

"You should be able to compute the amount of UV based on latitude" --- that is very interesting, philba. Can you advise how this is calculated ? by a formula ?

Originally Posted by philba

sorry, I don't know the formulas. google might help. I just know that the angle of the sun (actually amount of air the sunlight passes through) will change the amount of UV. I've seen charts that show UV radiation at the surface by time of day.

Those charts are about average intensity (think!?)> The UV level is afected by how much smog, humidity, air temp etc. so that won't be much accurate.

BTW...Why only 3V..that low voltage only increases the dificulties. U could use a 9V bat or so because that meter uses only 20-30mA Max (with display) and isn't used continous....Isn't it?