amtetcher late night problems .....

[John Shott]

SNF Lab Members:

As a number of your know only too well, the amtetcher often fails to 
pump down late at night .... typically close to midnight and later.

While we do not yet have a fix for this problem, we now have some 
reasonably convincing data and a hypothesis as to what is triggering 
this problem.

I apologize in advance for the length of this email .... but if you've 
been plagued by these problems, you may wish to read on.

As you are likely aware, the energy costs of a cleanroom are very high 
and are dominated by the energy required to heat, cool, and move all of 
the air that makes a cleanroom clean.  At normal fan speed there are 7 
to 10 air changes per minute in there.  To save energy, the fan speed 
has been reduced starting at 11 p.m. and returning to full speed at 6 
a.m.  When did that start?  I don't know but it has been going on for at 
least several months.  Note: Facilities Operations (FacOps) runs and 
manages the HVAC system in SNF as well as virtually all of the other 
building systems.

While one would not expect this reduction in fan speed to result in 
changes in temperature or humidity, we now have data in hand that shows 
that the temperature near amtetcher drops dramatically by 2-3 degrees C 
between 11 p.m. and midnight and stays pretty low overnight until the 
fan speed turns back up at 6 a.m.  It also appears as if the relative 
humidity in that aisle increases by 7-8 percentage points between 11 
p.m. and midnight.  I have posted these plots outside the cleanroom 
entrance .... or, if you are curious, send me email and I'll send you 
the Excel file.

Our hypothesis is that significantly lower temperatures and somewhat 
higher humidity cause the bell jar and the hexode to absorb more water 
vapor when the bell jar is up and that additional moisture is just 
enough to keep the system from pumping down below 0.5 mTorr which is 
close to "as low as it goes" for that pumping package.

How can we fix this?  At this point, we don't yet have a detailed plan 
.... we need to look at a few things.  Those include:

1. Working with FacOps to understand why the temperature varies so 
dramatically and work to eliminate the dramatic temperature drop.  Note: 
simply getting the in-lab temperature to better track the temperature 
set point may not be what we want because the actual set point is very 
close to the low overnight temperature that causes problems.  While 
simply keeping the fan speed at the daytime setting may seem attractive, 
that is certainly not the "green" approach in this case.

2. See if we can either increase the temperature or the heat capacity of 
the hot loop (that is the liquid flow that heats the bell jar all the 
time and the hexode when the bell jar is open) to better resist moisture 
absorption.

3. Check to make sure that the pumping system is working optimally.

4. Make sure that the ionization gauge is properly calibrated ... even 
being off by 0.1-0.5 mTorr could greatly exacerbate this problem.

We will keep you posted as we work to resolve this issue, but we wanted 
to let you know that we believe that we have finally identified what is 
triggering this strange "time of day" failure.

Let me know if you have any questions,

John