Mercury arc rectifier

[ppppenguin]

That photo is from the Hulme Hippodrome: https://www.dailymail.co.uk/news/article...s-ago.html

I can only see two anodes but there might be a third round the back. I would have expected this sort of installation to be three phase. There seem to be multiple connections to the cathode mercury pool. In many MARs the cathode is connected to an electrode in the bottom of the bulb. This probably gives better contact but it a mechanical weak spot. In this one, and mine, the cathode is an electrode dipping into the pool. I think I can make out the starter solenoid on the left MAR, again very much like mine. An electromagnet pulls an electrode into the mercury pool to start current flowing. This then operates a contactor that releases the solenoid and hopefully the arc is struck. An alternative start method, used on some MARs is to tilt the whole bulb to dip an electrode in the mercury.

This MAR is only a few miles from me. I wonder if it's still there.
https://www.28dayslater.co.uk/threads/be...21.128571/

[PerdioPal]

       
From "Electric Lamps" circa 1955

[ppppenguin]

i hadn't realised that filament lamps emitted any UV at all.

The question is: Is there any easy way to find out how much UV (and whether it's harmful UVC or other types) is coming from a MAR? I can't imagine UVC will make it through the glass of the bulb. Much UVB will also be absorbed but a lot of UVA will probably get through.
https://hps.org/publicinformation/ate/q12082.html

Watching for fluorescence doesn't tell you the wavelength and would be difficult because of the amount of visible light from both the MAR and lamp load.

[DrStrangelove]

Apparently there's enough UV emitted by a tungsten halogen linear lamp to expose PCB etch resist, this being the method proposed by one site.

Does anyone else remember the big UV lamp thingie at the health centre a lifetime ago? (i.e. late 50s early 60s).

I can remember the smell of proper ozone and wearing the eye protection.

Dunno what it was supposed to do to me other than generate vitamin D.

Oh. There's jolly then:

https://www.medscape.com/viewarticle/875073?form=fpf

The one I remember was a mercury vapour lamp shown further on.

Oh well.

[Kalee20]

i hadn't realised that filament lamps emitted any UV at all.

I hadn't either! Though the hotter the surface, the more radiation is emitted, and the higher the generally broad-band the frequencies are.

What's surprising is that the UV content is enough to be significant.

Gas discharges are, of course, a different mechanism.

[ppppenguin]

Gas discharges are, of course, a different mechanism.

And arcs are different again. A lot of UV. https://www.allaboutvision.com/condition...e-welding/

[DrStrangelove]

Gas discharges are, of course, a different mechanism.

And arcs are different again. A lot of UV. https://www.allaboutvision.com/condition...e-welding/

Indeed. As I recall from a UV burn from mig welding back in the day: from the junction between gauntlet and overalls, ditto between the face shield and overalls/shirt.

Stings a bit.

[PerdioPal]

Halogen bulbs were (and still are in surviving units) used for UV cure glues in industry and dentists. They have a filter to attenuate the visible so it isn't like using a searchlight! The new ones are LED based.

[Mike Watterson]

Tanning salons using florescent tubes cause skin cancer. The cheapest UV tubes for resist exposure (screen printing in late 1970s) seem to be just the regular florescent tubes without a phosphor, but the one in my old insectacutor seems filtered. Our current death-to-flying-bugs seems to use violet LEDs in tubes. That makes the yellow phosphor on the filament stick style LED lamps glow.

Also some washing powders and laser / copier paper have "brighteners" and fresh whites or copier paper can be very bright white with UV or violet light*.

The RGB model as used by displays and cameras works mostly, but doesn't quite match the human response, so visible violet LEDs (beyond blue but not actually UV) can be looking much bluer via camera and screen.

However many sensors have poor IR and UV rejection so white balance on tungsten or bright sunlight can be off. I have UV filters for all my DSLR lenses and they make a difference.

[* It's of course nonsense that a rainbow has seven colours. Newton deliberately picked 7]

Actual magenta / purple is light without the green, not a spectrum colour at all. But visible far violet spectrum causes some reaction to the red sensing in our eyes, so for some people the monochromatic violet just past "blue" looks purple like magenta, i.e. white light with green blocked or mixed red and green light. So a really good camera feeding a regular RGB display shouldn't look blue with near UV "violet" light. It should look closer to violet by using a blueish magenta.

[ppppenguin]

Most cameras don't show the violet light from a MAR very well. The colour of the discharge varies a little between MARs. Mine is definitely on the blue side, others are much more violet.

In general, different combinations of wavelengths can give the same visual appearance. This is known as metamerism. It's important in any industry where consistent colour appearance matters.
https://en.wikipedia.org/wiki/Metamerism_(color)

I was looking for some acrylic sheet to replace the metal side and front panels. I found a piece that was just a fraction larger than the side panel. So close that it's going to be difficult to trim off the odd millimetre or two. This is Sod getting his revenge for "any wire, pipe etc cut to size will be slightly too small". I can't saw off that little. Any ideas? Maybe clamp the metal piece to the acrylic and use a power sander?