Jump to content
UltravioletPhotography

Incandescent lamps as UV source, an attempt


baffe

Recommended Posts

I am looking for a source of uv. And I want it continously in time (not a flash) and in spectrum (not only spikes or lines).

The reason is, that some of my stones show fl only if they get a specific wavelength in uv. Some show (better) fl with Led illumination of 365nm, some with 395nm. And I don't know if some of them would show different fl at colours between.

So the Idea was to try an incandecent lamp as lightsource. And I made some tests. Here are only the essentials.

If the voltage of such a bulb is increased the colour of the emitted light changes. To check the usability and the degree of this I made a test. Using a 12V/20W halogen lamp and a regulated power supply. Beginning with 1V at a increasing voltage in steps of 1V to 12V. One photo at every voltage step. I placed the lamp, inside a tube of light ceramics. WB was set to sunlight, cam was Sigma SD10.

This are the Pictures. At low voltages one can see the light of my workshop in the background:

 

http://up.picr.de/20091685ci.jpg

 

To increase the amount of UV I didn't stop at 12V and made some shots with 13V, 14V... up to 24V. At a voltage of 25V the hopelessly overloaded lamp went defective. Here the results in pictures:

 

http://up.picr.de/20091689oy.jpg

 

 

I killed some of the 12V/20W bulbs but none of them went defective at a voltage below 24V! Of course they will not reach a long life under conditions of overload but they will not fail immediately either if voltage is not too high. So here is the complete set of shots:

 

http://up.picr.de/20091688py.jpg

 

A 1 inch UGxy Glass tube over the bulb and some uv active paint in a bottle and on a copper plate front to test for emitted uv. Here are test pictures at voltages of 0V 12V, 13V ... at 24V the filament fell apart.

 

http://up.picr.de/20091687kn.jpg

 

The contrast increases with increasing voltage on the bulb above nominal ratings. The lines on the copper plate do not need the full voltage at the bulb to get visible. Of course there is a lot of vis and ir and a lot of work to do...

 

What do you think?

 

! da baffe

Link to comment

Great work Stefan

I have looked at halogen lamps for UV before, but the ones I have looked at block UV :D

The original halogen desk lamps (40 years ago) were UV producing but got supersede.

I have not found a source of any tungsten lamps that don't block UV ?

Col

 

PS, I will comment more when I wake-up properly :P

Link to comment

A good presentation and the learning experience was probably beneficial but a quick google would have probably saved you time and money since you have re-discovered the wheel.

 

With regards to

Some show (better) fl with Led illumination of 365nm, some with 395nm. And I don't know if some of them would show different fl at colours between.

 

Fluorescence is dependent on excitation and emission centers i.e. the material has a region of the spectrum where it will absorb the "excitation" wavelengths and a region where fluorescence emission occurs (typically at longer wavelengths). The wavelength location of these centers is dependent on dopants or defects in the crystal structure. Many people think that excitation centers only occur at 365nm and the reason for this is that 365nm just happens to be the most common wavelength specified in LWUV lamps. While it is true that 365nm causes fluorescence in a lot of different materials it may not be the only excitation wavelength or the most efficient.

 

So the answer to your question

I don't know if some of them would show different fl at colours between

They might but only experimentation will determine that unless you can find a good reference paper.

 

An excellent overview reference is

Fluorescence: Gems and Minerals Under Ultraviolet Light by Manuel Robbins.

 

Excellent technical references are

Luminescent materials by Blasse

and

Luminescence of Solids by Vij

Link to comment
I have not found a source of any tungsten lamps that don't block UV

 

Most microscope halogens bulbs I have used in the past put out UV, not sure about the ones found in some bathroom lighting.

Link to comment

Shane I didn't "re-discover the wheel"! It's the diffeenc between "reading" something and "knowing" it.

 

Of course I could find and read a lot about halogen lamps and uv in the web.

 

Now I know that the uv-output of that bulbs on my desk really is enough to create fl under the conditions of overrating them with about 150% of the nominal voltage.

 

Further I know that there is not enough uv output at nominal operating voltage.

 

And I know that increasing the voltage from 150% to 200% of nominal value will only decrease lifetime and not increase uv output (very much). And everybody can see that on the pictures above.

 

>They might but only experimentation will determine that unless you can find a good reference paper.

 

As I do not have reference papers referring to the materials I dig from the ground experimentation will be the solution. But I needed a light source. Now I have a suitable light source right here on my desk. And I know it does work and not only on some peaks but on a broad bandwith.

 

I didn't only check the 20W bulbs. I checked a lot of different types and brands. "Bathroom" bulbs and automotive ones. The 12V/35W "bathroom bulbs" are best for me. Not the expensive ones but the cheap, packet together 3 to 5 pieces and noname. I found the bulbs that do not have sufficient output have a visible metallic "dichroitic view" of the glass or it's coating.

 

Another type that I know it works well is the 12V 50W bulbs mounted inside alu reflectors. I didn't do tests with bigger ones. The problem is that even with the 20W bulbs the filters get really hot and need reinforced cooling.

 

Of course it isn't possible to use halogen lamps instead of uv flashes, discharge lamps and uv led's but it's a special solution to my problem. And it is a solution that requires mains and a good power supply.

 

! da baffe

Link to comment

You might find these two references beneficial:

http://zeiss-campus.magnet.fsu.edu/articles/lightsources/tungstenhalogen.html

In particular Spectral Output and Colour Temperature

and

http://www.intl-lighttech.com/applications/light-sources/tungsten-halogen-lamps

 

In particular, the plot of the effects of increasing the design voltage with respect to output colour temperature, luminous output and lamp life.

Link to comment

Thanks a lot for the links!

 

I know about the effects. And I know my lamps will burn down like candles.

 

Similar to 1000W/230V bulbs for my photobeamers with a expected lifetime of about 3 hours (30 Euros each), -if one doesn't switch on/off too often. But I once bought a new beamer marked "made in GDR" (means it is about 25 years old now) threw away the original bulb and put an Osram in and never had to replace it. I use that bulb frequently for about 25 years now!

Link to comment
enricosavazzi
I remember that some 10-15 years ago, in Sweden at least, there was a brief scare in the media about the danger of UV emitted by small halogen lamps similar to the one you tested. Ikea and other sellers quickly modified their products for home illumination as a result. This probably means that much of the UV actually emitted within modern versions of these halogen bulbs is absorbed by the bulb glass.
Link to comment
enricosavazzi
The best continuous UV source with a smooth emission spectrum is probably a deuterium lamp. Because of this characteristic, these lamps are frequently used in spectrometers. Deuterium lamps are not strong, but their emission peak is around 220 nm, so they require filtering out the UV-B and UV-C for safety reasons if used "in the open".
Link to comment

Enrico i was considering bying a deuterium lamp but I was a little afraid of the uv-c. I think I will not be able to check a deuterium lamp if my Filtering is suffcient.

 

I read that a deuterium lamp is very expensive, sensitive in operation and handling and doesn't give very much output. (My halogen lamps do not either)

 

In my opinion the changes from "bare bulbs" to lamps with cover glass were to avoid burnings and injuries by exploding bulbs. But possibly I was wrong and that was to decrease uv emission like you describe that?

 

Colin I have some automotive and other xenon HID lamps including suitable inverters but I never checked them for uv. Perhaps should I do that...

Link to comment
enricosavazzi

I tried two or three different types of automotive HIDs. There is some UV-A, even enough to take pictures with an exposure time of several seconds. There is of course a lot more VIS than UV, but on the other hand far less NIR than from incandescent lamps of the same power.

 

The degrees K temperature specified by eBay sellers for automotive HIDs, at least in some cases, seems to be unreliable or even a fantasy figure. My best bet would be to buy one of the violet/purple ones. However, I have one that emits yellowish VIS but nonetheless a little more UV-A than a violet one. This does not seem to make much sense, but the difference I am seeing might be caused by UV absorption by the bulb glass or surrounding sleeve glass.

Link to comment

HID headlight, sparepart for my ATV. Modified halogen foglamp.

 

http://up.picr.de/20099968wf.jpg

 

The xenon lamp without housing. One can see bulb glass and surrounding sleeve glass enrico was talking about.

 

http://up.picr.de/20099969df.jpg

 

The lamp put into UGxy filter tube.

 

http://up.picr.de/20099970im.jpg

 

The xenon lamp switched on...

 

http://up.picr.de/20099965zl.jpg

 

...and the lights in the workshop switched off.

 

http://up.picr.de/20099966tg.jpg

 

Fl on uv active paint.

 

http://up.picr.de/20099967yj.jpg

Link to comment

Nice experimenting Baffe. I hear what you say about knowing something. Most, if not all of what I have done so far has been already done some other time before me but that will never stop me from finding out for myself!

 

-D

Link to comment

Usabillity of a solution depends on the problem.

 

Of course it is not suitable to generate uv with incandscent lamps for professional microskopy like shane needs it. No doubt about that!

 

But for me and my pictures it possibly is. I don't know yet. But it at least isn't expensive to find out...

 

Professionals built the titanic but an amateur built the ark! The ark solved noahs problem. But one should't forget that it was a disposable article for one-time use and not intended for professional use.

Link to comment
  • 7 months later...

Hi!

 

Some tests with a "Boehringer, Ingelheim, MSP12" Spectrometer.

 

Light source was a standard 12V/20W halogen bulb with "UV protection".

 

An adjusteable stabilized power supply delivered the operating voltage for the lamp.

 

 

At 2V quite "noisy" but one will not operate a 12V lamp at 2V to get uv:

http://up.picr.de/22326862fr.jpg

 

...4V:

http://up.picr.de/22326863ju.jpg

 

...6V:

http://up.picr.de/22326864qa.jpg

 

...8V:

http://up.picr.de/22326865we.jpg

 

...10V:

http://up.picr.de/22326866rj.jpg

 

...12V:

http://up.picr.de/22326867sl.jpg

 

...14V (here and below the lamp is overloaded):

http://up.picr.de/22326868wt.jpg

 

...16V:

http://up.picr.de/22326869va.jpg

 

...18V:

http://up.picr.de/22326870uj.jpg

 

...20V:

http://up.picr.de/22326871vm.jpg

 

Of course the 12V lamp was operated far beyond it's maximum ratings during the last tests. But it survived even the 20V Test!

 

Stefan

Link to comment

Quartz tungsten halogen lamps are used as spectral irradiance standards down to about 250nm depending on the wattage.

 

Not to expensive, perhaps worth a try are MR16 halogen photo optic lamps such as an OSRAM 54814 ELC.

 

As with any lamp of this type UV is less tan visible and IR but so is sunlight and the dichroic reflector helps mitigate this.

 

PS

Look for the ones with the UV safety warnings

Link to comment

Colin we have rainy weather today. But I'll make a shot when the sun is shining again.

 

My problem with the dichroitic reflectors is that they have different light on diferent angles (areas) of the beam. Simple bulbs do not show that effect.

Link to comment

Colin that was the sky this afternoon about 4:00PM :

 

http://up.picr.de/22345126gr.jpg

 

And that was the according spectrum:

 

http://up.picr.de/22345127yd.jpg

Link to comment

Colin if operating voltage doesn't exceed 16 or 18V the lamp should work for some time. Increasing the voltage slowly to the final voltage also increases lamp life. This works even if the lamp is not overloaded!

 

I still use the second lamp. The first 12V lamp went defective at 24V. So I didn't use the second lamp on voltages above 20V.

 

John following your hint I found a manual of a "Spekol11 spectrometer" from CZJ, unfortunately in german.

 

This page shows it in english:

 

https://www.mccrone.com/mm/a-poor-plm-microscopists-tunable-monochromatic-light-source/

 

I found that the instrument (Spekol11) uses a "Tungsten-Halogen-Lamp as standard light source" in the range of 340nm to 850nm.

Link to comment

Interesting use of an old single grating scanning monochromator. They are commonly incorporated into tunable sources but do not provide as clean a spectrum as a scanning double grating.

 

By "standard" source I mean a calibration standard one would use to calibrate a spectroradiometer. The OL 752-10 and 752-12 are what I mainly use to calibrate my instruments.

The plot on the second page shows the typical spectral output. If you pass it through ~1mm Schott WG-320 you have a "poor man's solar simulator".

 

Sorry to say but for comparison to sunlight your little diode array spectrometer probably lacks sufficient dynamic range and stray light rejection to measure into the UV, see: A Comparison of CCD and Scanning Systems for Spectroradiometry (particulary figure 3) so don't read to much into those spectra. :(

Link to comment

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...