Hello,
Heating was a big mistake, now the oil is everywhere !



MasCom GmbH
Norderoog 1
28259 Bremen
Germany

Tel. 0049 (0)421 57297 13
Fax 0049 (0)421 571032

Homepage: www.mascom-ms.de
email: [log in to unmask]




-----Ursprüngliche Nachricht-----
Von: Geldern, Robert [mailto:[log in to unmask]]
Gesendet: Dienstag, 6. April 2004 11:07
An: [log in to unmask]
Betreff: [ISOGEOCHEM] Failure of fore vacuum pump and opening of the
source (Delta S)


Greetings,
one of our rotary vane fore vacuum pumps (Pfeiffer Duo 5) on our Finnigan
Delta S failed and it seems that oil vapor entered the vacuum tube
connection from the fore vacuum to the turbo pump (the large one under the
source). The backround scan showed lots of peaks. The complete vacuum tube
system (from all three fore vacuum pumps) was disconnected and cleaned
carefully. The large turbo pump under the source was dismounted and after
putting it on the side, some oil seeped out of the pump. We cleaned the pump
as good as possible, but I am not sure if this was enough. The fore vacuum
pump was replaced. By the same time the source was opened since a leak was
detected on the tube connecting the change over with the source. The gaskets
were replaced.

After re-connecting everything and the reintegration of the source we pumped
over the weekend an reached good vacuum (HV and FV, the FV took a while...).
We checked for leaks with Argon but did not find something. We heated the
source and the flight tube to 130 to 150 °C overnight (baking and a kind of
heating wire wrapped around the source and flight tube).

The backround scan (slow magnet scan) this morning still showed lots of
peaks which should not be there. We will check for a leak again, but I am
concerned that the contamination with oil vapour (maybe into the source/tube
itself?) is causing this.

Any suggestions on what we can do are highly appreciated.

Robert



--
Robert van Geldern
GGA-Institut (Leibniz Institute for Applied Geosciences)
Stilleweg 2, D-30655 Hannover (Germany)
fon: +49 (0)511 643-2716 (fax: -3665)
mailto:[log in to unmask]