Search the CONFOCAL archive at
http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal

At the risk of stating the obvious, do check
carefully the settings of your power supply.
I had similar overheating and short life problems
when a (well-known) microscope company delivered
and installled a system set to the correct voltage
but the wrong frequency (60 Hz instead of 50).
This seems to make a very big difference.  (Whereas
220v instead of 240v on another system did very
little).

                                        Guy



> Hi Andrew,
>
> As you have noticed, the high temperatures are oxidizing the copper.
> that might affect ignition, but will certainly be an issue for stability
> over the long haul. It will be good to nickel plate the (cleaned, or
> new) copper fittings you have made. Osram also has a data sheet that
> specifies, I think, that the electrodes should not exceed 230C - 200C is
> safer. At this temperature I wouldn't expect too rapid - like what you
> have seen - oxidation so maybe the lamp is running too hot. The high
> contact resistance will also add to the heating, a nasty downward spiral.
>
> The Anode (+) terminal should be at the bottom (I say this since you
> have rewired and changed the connections - lamp housings are made
> originally to enforce this based on the size difference of terminals).
> The upper (-) electrode tends to get hottest and most upper "terminals"
> incorporate some heatsink fins to radiate heat. You can measure the
> upper terminal with a fine wire thermocouple - probably best to not
> connect to your thermocouple meter until after ignition ;-) to check the
> actual base temps. The voltage across the lamp can also be measured
> (<<POST IGNITION>>) to check the operation parameters - should be ~17V
> immediately post-ignition, rising to ~22V after warmup if the power
> supply is driving it correctly.
>
> Assuming that the oxide is cleaned off, a new lamp should ignite and run
> so you can make the measurements - shut off immediately if out of range
> so you don't toast the new lamp.
>
> Remember that the lamps are under high pressure when HOT, so never
> operate with the cover removed to make measurements. Bad UV also.
>
> Hope any of this helps.
>
> Dale Callaham
> The University of Massachusetts @ Amherst
>
> Andrew M. O'Grady wrote:
>> Search the CONFOCAL archive at
>> http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal
>>
>> Hello Everyone,
>>
>>   This is my first email to this list. I apologize that this is not
>> exactly related to microscopy, but this seems to be the largest
>> knowledge base about mercury vapor lamps that I have been able to
>> find on the internet.
>>   We are using an HBO 103W bulb for our Schlieren imaging system for
>> supersonic flow visualization. The old setup used an OSRAM 200W/2
>> bulb with a lamp housing and lens (J. Unertl Optical Co.) and a
>> power supply from Robert W. Gates & co.  The equipment must be at
>> least 30 years old, if not 50 years. Recently the power supply
>> stopped working, and we purchased a Nikon C-SHP1 power supply. We
>> hooked up the power supply to the old housing (making certain
>> modifications), and used an HBO103W bulb which worked fantastically
>> for about 30 hours, having been turned off and on maybe 10 times.
>> The last time I tried to light it, the "Lamp Ready" light flickered
>> for 1 or 2 seconds and then went off. Examining the lamp, it did not
>> explode, but the inside of the glass is covered with mercury, so
>> that the electrodes are not even visible. I did not even try to
>> relight the bulb, I just assumed it was ruined.
>>
>> I think that the bulb has probably failed from the modifications
>> that we made to the connections in the housing. The first
>> modification was that the mount for the bottom of the bulb had to
>> be made smaller because of the smaller diameter of the 103Wbulb.
>> This we did by using a small sleeve of copper inserted into
>> whatever metal material the existing mount was, possibly bronze.
>> After the bulb stopped working, the inside of the copper sleeve is
>> tinged pink, compared to the outside.
>>
>>  For the top bulb mount, the old style bulb we were using had a
>> screw on it, and was attached to a flexible wire protected by
>> ceramic beads. The new style does not have a screw, so we made
>> another sleeve out of copper with a screw to tighten. The flexible
>> wiring with ceramic is still used to account for thermal expansion.
>> After running, the top copper mount is now dark gray-colored
>> (probably oxide), but this can be scratched off.
>>
>> One other thing I noticed before the bulb would not light was that
>> there appeared to be some small bubbles inside of the bulb near the
>> bottom electrode after I had run it a few times.
>>
>> I appreciate people taking the time to help me on this, and I intend
>> on attempting to contact OSRAM regarding this matter, but this
>> seemed like a good place to look for assistance.
>>
>> Thanks,
>> Andrew O'Grady
>>
>


-- 
Associate Professor Guy Cox
Electron Microscope Unit,
University of Sydney,
NSW 2006, Australia

Phone:+61 2 9351 3176    Fax:+61 2 9351 7682
http://www.guycox.net