what is the lifetime of a Deuterium lamp?

[Arodrigues1975]

That's new for me.... Optical age short then D2 Lamp... Where had you that equipment? Ã

[Arodrigues1975]

Let's talk a litle of D2 lamps..

Filament.... that burns when you switch on the lamp = if you switch on and off many times that can be damage. (I don't remember to change to many lamps with that malfuntion)

UV Energy - When the lamp is new, it will be around 120% till the firts 100 hours... after that we can take in this time 100% of energy.... Life time of that D2 will lower till X% of energy. How much that energy is good or not? Some D2 manufactors says: in 1000, 2000, 3000...hours that get 60%...50%... of energy...
Depends of the aplication were that lamp is instaled and the increase of noise you will get on the time. Surely hour after hour that energy will decrease and with that decrease you results will be noisy.

If you don't change the lamp in any case... sometimes with 4Mh or 5Mh.. you can damage the power supply and in that case you will pay not only one lamp but maybe the price of 3 ou 4 lamps.
Usualy I change D2 for sure after 1500h and before 2000h.

Hi

We made the experience that switching on/off the D2 Lamp often will have significant influence on the lifetime of the lamp....the less swithing the longer the lifetime....

But if your lamp reached the 1000hs its a good time to have one in stock....

Best regards
Chris

[danko]

Hi Arodrigues,
[quote]That's new for me.... Optical age short then D2 Lamp... Where had you that equipment? Ã

[Arodrigues1975]

To explain the error when start analyses after replace D2 lamp.

120% = to prevent lost of signal during analises after install a new lamp, you have to burn a feew hours to reach the 100% normal and stable energy.

UV Energy - When the lamp is new, it will be around 120%

120% of what? Of the whole energy? [/quote]

[HW Mueller]

Hm some of these statements would be quite important if they were correct. How about giving some references so one may be able to judge on the quality of the info.? Especially interesting would be to see what led to the statement that a bad lamp can ruin the power supply.
My UV and fluorescence detectors are now at least 15 years old, all of them have performance checks. I replaced the lamps if the performance checks, or unacceptable increase in noise, indicated a bad lamp. . . . not too much discussion here.

[unmgvar]

from my own experience i know also that switching off the lamp too often is worst then leaving it on.
i saw a lamp on an instrument that when out only after about 3000 hours. on average we saw that it had been switched on/off 3 times per day.
if you will use the instrument all week long then don't turn the lamp off, do make sure that you leave the flow runnig as it serves to cool the flow cell and will prevent distortion of the parts and dust being deposited on the lens

if your lab is dusty then after a few years dust will be deposited on the mirrors.
no need to replace them. dipp the lens while holding the base of the mirror in a soap solution with lab/hplc water at 40-50 degrees, stirr a bit. after a few moments take it out, wash with hplc grade water leave to dry or use hplc grade acetone. do not wype. put back in place

[dblux_]

...Especially interesting would be to see what led to the statement that a bad lamp can ruin the power supply.
...

When lamp is lit the distance between electrodes increases slightly, especially as an effect of electroerosion of arrow ended electrode.
This leads to higher voltages necessary to lit the lamp and thus higher load of power supply. All good designed and constructed with high quality elements PSU should withstand such increased load (n.b. after lamp is lit PSU load comes back to normal). However we can imagine situation when discrete elements (e.g. capacitors) won't survive prolonged and repeated higher loads in PSU.

Wojtek, Sz-n

[HW Mueller]

Didn´t think there was that much junk/junked out there.
The last instrument in which I ever saw such an optical part as a mirror was ~ a 30 year old Shoeffel (then Kratos, then...?). Even that "visible" part did not have discernible dust on it, but clearly a burn (see also Danko). The nice thing about this machine was that we didn´t see a lot of dirt in our samples. Needless to say: we used it for a few special applications, then retired it.

[Mattias]

I don't think there is one single UV detector out there that doesn't contain mirrors to transport the light beam.

I agree that it is not dust that is the problem - it is the light itself that degrades the surface of the mirrors (also stated by Danko). It seems to be instrument dependant how often you need to change the mirrors. We bought a few Dionex systems in 2003 - and we needed to change mirrors in all of them this year. The Waters 2487 detector doesn't last forever either. On the other hand, we have never opened any Waters 996 or any Agilent detectors so far.

[HW Mueller]

The reason I have not seen any mirrors in all but very old apparati is that the optical paths were well protected or even sealed, not because they were not there.

[unmgvar]

As you state HW good detector have their optic sealed and protected, but it is not done well in all detectors. this was a procedure that i learned in order to maintaine hitachi detectors. my region being very dusty, we needed to do this every 2-2.5 years per instrument. it does affect the energy of the lamp on the sample because it absorbs and disperses the light.
only changed a mirror once on a very old instrument thou.

[danko]

I wonder how you managed to perform reliable analysis in all that dust you keep mentioning.

only changed a mirror once on a very old instrument thou.

What did you see on the replaced mirror? Let me guess – a lot of dust

Best Regards

[unmgvar]

no we saw an oxidised surface under a lot of dust

[tom jupille]

Something like this ?



For what it's worth, I've also seen cell windows that had "solarized" beautiful shades of amethyst or tan after long exposure to UV.

[wanda50]

An aside:

We have found that not all labs monitor their lamp performance. When we write validation protocols, checking lamp energies is included in the protocol. Extreme, but...

Case in Point: One fellow was getting poor chromatography. I looked at it and said the problem was the lamp. He replied to me that the detector was just calibrated 10 months ago, so it couldn't be the lamp. Can't argue with that logic! After many days of this, he changed to another machine and everything was solved. He later admitted to me that the lamp on the first LC had over 2000 hours and failed the detector test.

At that point, I didn't have the heart to tell him about fogged mirrors in the detector. I am saving it for another day.