>>> 'gartek' problem from rnorman -- Mon Jan  9 15:43:08 1989 <<<

I talked with Mark Brown of Innotec and he is having a hard time
getting our PCB for the RF Network. When he gets it he will call
us for arangements; this should be latter this week or early next.

>>> 'gartek' problem from hebert -- Mon Jan  9 18:51:55 1989 <<<

RF power supply doesn't match very well. The LOAD pot cannot
be adjusted manually. 

EST REPAIR TIME: 1.5 hours to determine why LOAD pot doesn't
                 adjust.



Can someone call ERATRON and let them know that we're in a bind. 
We can't deposit 70 A of Al repeatably with the supply in this
state of disrepair. Thank you.

>>> 'gartek' problem from rnorman -- Tue Jan 10 09:01:42 1989 <<<

Problem: Rf power supply not matching well.

Solution: I connected the dummy load to the RF power supply and it was
         a perfect match. I then ignited an Ar plasma: 170 sccm @ Nb;
         the match was also perfect up to about 500 watts, which had
         less then about 3 watts reflected power. In short, I could
         not duplicate the problem. The gartek is up. Maybe Dave can
         show me the problem he had.

>>> 'gartek' problem from rnorman -- Tue Jan 10 11:53:44 1989 <<<

Problem: Manual RF tuning not working.

Cause: The coupler between the motor and the cap was loose.

Solution: The screws were tightened in the coulper and the RF
          network was tested. The Gartek is up.

>>> 'gartek' problem from hebert -- Tue Jan 17 19:27:43 1989 <<<

Can't ignite RF plasma. WHEN IS ERATRON COMING BACK TO FIX THE
MATCHING NETWORK? CAN THEY TELL US HOW TO GET 50 TO 150 W DC OUT
OF THE DC SUPPLY THAT WE USE TO SPUTTER NIOBIUM? THE DC SUPPLY
RANGE IS 300 TO 5KW. IF WE COULD LOWER THE POWER LEVEL, THEN WE
COULD USE THE DC SUPPLY TO SPUTTER ALUMINUM AS WELL AS NIOBIUM, AND
WE COULD GET RID OF THE RF SUPPLY.

>>> 'gartek' problem from hebert -- Thu Jan 19 19:09:58 1989 <<<

The M&W chiller that cools the substrate table and drives the
magnetrons is very noisy. The water pressure was zero PSI. I
turned the unit off.

>>> 'gartek' problem from rnorman -- Fri Jan 20 14:36:36 1989 <<<

Problem: M&W chiller very noisy.

Cause: The brass drive key between the motor and the procon water
       pump was stripped.

Solution: The key was replaced and the chiller is up.

>>> 'gartek' problem from vallath -- Fri Jan 20 15:11:58 1989 <<<

The rf power supply has a problem:

I used it to presputter with no problem.
The next time, the plasma wouldn't ignite, and the power supply
shut itself off within a couple of seconds.
The RF on/off LEDs were both off, and the green OFF LED
wouldn't come on even with a wait and even when the
over-current switch was flipped to reset.
There is water flowing.  Under normal conditions, the
RF off LED should be glowing.
Maybe a fuse has blown?
Vallath

>>> 'gartek' problem from mikeb -- Sat Jan 21 13:06:57 1989 <<<

1. problem:	RF (aluminum) plasma does not ignite, as reported by Vallath
		yesterday.

   cause:	There is a problem with the H2O interlock.  The RF plasma
		ignites normally if the interlock is defeated (jumpered),
		with the H2O running, of course.

2. problem:	I found the coupling pin again destroyed on the heater/chiller
		pump.  Given our good supply of pins, I went ahead and replaced
		it, lubricating with "Moly Grease" aerosol spray.

		There is no reason to think that this coupling pin will not
		fail again, and so I will not leave the M&W on overnight.

   possible cause for both problems:
		If the pump is dirty it might act as a heavy load on the motor
		and cause excess wear on the coupling pin.  It may also have
		reduced flow so that the flow interlock flakes out.

>>> 'gartek' problem from hebert -- Mon Jan 23 18:20:54 1989 <<<

The regulator in the DC power supply is bad. It was pulled by ERATRON
field service rep for repair. It will be back around 1PM on 1/24/89.
Until then, the Gartek is DOWN.
From phillip Wed Jan 25 13:21:48 1989
To: gartek
Subject: taurus control
Cc:  $pf/gartek-p 

 The gartek machine is connected to taurus this day & hour.


>>> 'gartek' problem from vallath -- Thu Jan 26 13:53:28 1989 <<<

The guy who came to fix the rf power supplies told us that
our grounding scheme contained a lot of loops, which radiated
rf which causes our instruments to fluctuate.  He also said that
it was not safe for the operator as it is, and has suggested an
alternate grounding scheme.  Could this be done?  Robert knows 
about it.
VN

>>> 'gartek' problem from hebert -- Tue Jan 31 15:17:07 1989 <<<

The sputter chamber cryo has been warmed-up and vented. It can be
pulled for rebuild at any time. The high vac valve to this cryo
is closed, and the chambers remain under high vac with the
XPS cryo.

>>> 'gartek' problem from hebert -- Tue Jan 31 16:36:19 1989 <<<

The high vac valve to the sput cham cryo leaks when closed.
For this reason, the cryo is not vented. Both cryos are at RT,
and the chamber is being pumped on by the turbo pump.

>>> 'gartek' problem from bob -- Wed Feb  1 15:41:18 1989 <<<

There is a muffing fan out in the DC supply. It is operational
however this need fixing.

>>> 'gartek' problem from hebert -- Wed Feb  1 17:35:21 1989 <<<

I pulled the sputter chamber cryo pump. I covered all open flanges
with aluminum foil. The pump is on the table in GL3. I started to
remove the N2 purge heater and didn't finish becaUSE I thought that
James might prefer to remove the entire heating assy at the flange
and I had no way to cover the flange after removing the assy.

The cryo had an unusual odor to it upon removal -- kind of like
beef broth. I don't know if this means anything. Maybe beef is real
food for real pumps (whatever that means).

Please make sure that all of the parts to install the new gas lines 
are in GL3 so that I don't have to look all over for them on 2/3
when I install them.

While we're at it, the sput cham high vac valve doesn't close all
the way and needs to be fixed.

I turned off the turbo and the backing pump since I don't have
a way to blank off the cryo flange (i.e. we can't keep the chamber
under vacuum).

I turned off the air and N2 to the gartek.

>>> 'gartek' problem from hebert -- Thu Feb  2 20:21:52 1989 <<<

The plumbing for the new gas lines has been installed. James Parrish
did a GREAT job designing and fabricating the new lines. The 
installation went smoothly. I anchored the gas bottles to the
wall in there new location. I still have to connect the air
cylinders and the electronics.

The sput cham high vac valve still needs to be repaired, and the
grounding problem needs to be resolved.


I left the old gas lines in the hall. I covered the ends with foil.
They are rather long and I thought the lab might be able to use them
elsewhere since they only had O2 and Ar flowing through them.

>>> 'gartek' problem from hebert -- Fri Feb  3 15:43:13 1989 <<<

Problem: Leaky high vac/throttle valve to sput cham cryo.

Solution: Removed old o-ring at the sealing flange. It was
          damaged along most of the length of the o-ring
          (scratches and severe blemishes). The o-ring groove,
          sealing surface were cleaned and a new o-ring was
          installed.

>>> 'gartek' problem from hebert -- Tue Feb  7 16:35:59 1989 <<<

I need a Type K TC (teflon coated) approx 6 feet long for 
measuring the temp at the substrate cooling table. I'd like
to have this tomorrow so that I can install it before pumping
out the chamber.

>>> 'gartek' problem from hebert -- Tue Feb  7 16:53:05 1989 <<<

TC: I asked for type K, but I think that this isn't the 
right choice for the temp range I need. The table is
typically at 4 deg C and we may want to go lower than
that in the future. The lowest temp we may look at is 77 deg K.
So I'd like a TC that can give me temp readings as low as possible
below 0 deg C. A max temp of 25 - 50 deg C is OK. I was leaned
towards a type K because I have a type K readout in the loadlock.
But if I have to, I can read thermoelectric voltage with a 
voltmeter and reference a chart to determine the actual substrate
temperature.

>>> 'gartek' problem from hebert -- Tue Feb  7 17:16:06 1989 <<<

Sorry to bother you again about the TC. I checked an OMEGA catalog
and TYPE K's look like they go down to -270C. So a Type K TC should
be OK.

Can you tell me what I need to get so that I can switch the DC
supply between the niobium and aluminum targets?

>>> 'gartek' problem from hebert -- Fri Feb 10 11:39:41 1989 <<<

I'm having a problem with a Parker valve that opens a gas line.
I need to consult with someone.

EST time: 15 minutes

>>> 'gartek' problem from vallath -- Mon Feb 13 11:22:42 1989 <<<

The rebuilt pump was installed and the system is almost
down to the usual base pressure (Now it is 7x10e-8, no presputter).

With an Ar flow of 300 sccm, the temp in the cryo rises about
1 deg. in 2 min.

WIth 400 sccm, temp rises 2 deg. in a little over 1 min.

The table 1 down sensor is dysfunctional.  It does not
light up the LED even when moved around.  Or the led is blown.
Vallath

>>> 'gartek' problem from hebert -- Mon Feb 13 13:09:29 1989 <<<

The He pressure in the sput cham cryo compressor is 265-270 PSI.
It is 270-275 PSI in the XPS cryo compressor. I think that He
needs to be added to both compressors. From what I remember the
correct pressure is 280-290 PSI.

>>> 'gartek' problem from hebert -- Mon Feb 13 17:48:34 1989 <<<

The CTI manual states that the He pressure should be 245-250PSI.
If this is correct, the He pressure in both compressors is too
high. Is this right? I thought that the pressure should be about
285. RNorman told me that he thinks it is supposed to be 290-295PSI.
Has the pressure requirement been changed by CTI, or is the manual
accurate? In view of the pump rebuild, what else can cause the pump to
warm up when pumping argon to 420sccm flow? 
DH

>>> 'gartek' problem from hebert -- Mon Feb 13 20:32:00 1989 <<<

Problem: Bad RF grounding.

Solution: I replaced all of the RF braided ground straps
with copper sheet metal ground straps. There is still RF
interference with the surrounding electronics. I'll continue
to work on this on 2/14/89.

>>> 'gartek' problem from vallath -- Tue Feb 14 11:23:22 1989 <<<

The molex connector to the table 1 sensor has a broken wire,
and some nearly broken wire.  That's why the sensor appeared 
to not work.
VN

>>> 'gartek' problem from rnorman -- Tue Feb 14 15:33:23 1989 <<<

Problem: The molex connector on table 1 had some broken wires.

Cause: Most likly due to pulling on the wire when disconnecting
       the molex connector.

Solution: Installed and soldered 4 new pins. The table now works.

>>> 'gartek' problem from hebert -- Tue Feb 14 21:51:40 1989 <<<

I turned off both cryo compressors so that they can be charged
with He. 

There is a water leak beneath the gartek. It is on a water line
from the M&W chiller.


EST REPAIR TIME:  HE charge for both compressors: 1 hr.
                  Water leak: 10 minutes.

>>> 'gartek' problem from hebert -- Wed Feb 15 07:41:08 1989 <<<

Problem: Sputter chamber cryo warms-up when pumping 300 sccm Ar
         (N2 MFC). It warms about 1 deg K in two minutes at this flow.
         Before the pump was rebuilt it warmed up about 2 deg K
         in two minutes at the same flow.

What else can cause the pump to warm-up?

The CTI manual states thar the He pressure should be 245 PSI.

The pressure stated in the manual is supposed to be a "static"
pressure (i.e. the pressure with the compressors OFF0.

The dynamic pressure (i.e. weith compressors ON) is 265 PSI. RNorman
has said in the past that the dynamic pressure should be 290 PSI
(REF GARTEK PROBLEM DD 8/12/88). Last year we had a base pressure
problem in the sput chamber that went away after the compressor
was recharged to 290 PSI (dynamic).


I was told that the static pressure would be lower than the dynamic pressure,
and that the static pressure would be low if the dynamic pressure was
low.

Last night I turned OFF the compressor. This morning the static pressure
is 250 PSI. It is within spec. However, in the past we operated at a
higher dynamic pressure. That's why I think the compressor needs He
recharge.

Solution: RECHARGE COMPRESSOR. 

Note: This worked in the past. However, if you don't think that this
approach is right I'd appreciate someone giving me some other idea
about what is causing this pump to warm-up. The XPS cryo doesn't warm-up
using the aforementioned Ar gas flow. The dynamic He pressure in the
XPS cryo compressor is about 10 - 15 PSI higher than in the sput cham cryo.
Both compressors have the same static pressure (i.e. 250 PSI). Since the
static pressure is the same in both compressors and the dynamic pressure
is different, can this indicate that the sput cham cryo compressor is
weak? A solution would be to swap compressors if this is the case. HELP!

>>> 'gartek' problem from rnorman -- Wed Feb 15 11:54:49 1989 <<<

Problem: Sputter cryo warms up when pumping 300 sccm Ar.

Cause: Possibly compressor pressure low, when on was around 265 psi and
       should be between 270-290 psi.

Solution: I recharged both cryos to about 260 psi static; adding 10 psi.
          The He lines were switched between the cryos at the cryo. If there
          is a compressor or line problem it will switch to the other cryo.
          The gartek is up.

>>> 'gartek' problem from hebert -- Thu Feb 16 11:59:00 1989 <<<

The sputter chamber cryo still warms-up to 16 deg K in 2 min
while pumping 300 sccm (N2 MFC) Ar. This is an improvement from before
the pump rebuild. Yesterday RNorman swapped compressors with the
XPS cryo by switching the He lines at the cryo pumps. He also added
about 10 PSI He to each compressor.

The only other thing that can be done is to switch the XPS cryo with
the one from the sputtering chamber. I not sure if I want to do that,
however. Is it possible that the warming that we're seeing is due to
inaccuracy in the hydrogen bulb thermometer?

Other gartek issues:

    1) I've installed new RF grounds, and replaced the 8U power cable
       with an R214 cable. We still have an RF leakage problem that
       effects the convectron gages. The result is that the load lock
       vents when the RF is on and the high vac valve to the turbo
       opens and closes sporadically.

    2) The problem listed in #1 above could be eliminated if I could
       get the DC supply connected to the DC power switch box and to 
       the Al target.

    3) The heat blanket to the cryo needs a power cord so that we can
       keep the case warm to prevent O2 from freezing out on the
       80 deg K array.

    4) There is a water leak between the M&W chiller and the chamber.

Whatever help I can get in finishing these things would be appreciated.
#3 and #4 are very minor -- I just haven't had a chance to get to these
myself. I don't like messing around with high voltage, so #2 is something
I need help with the most. Thank you.

>>> 'gartek' problem from hebert -- Thu Feb 16 13:28:25 1989 <<<

Problem: RF leakage effecting electronics, load lock vent, and
         high vac valve in load lock.

Solution: Mikeb found a bad RF cable between the RF power supply
          and the RF switch. The RF leakage is now at a minimum.

>>> 'gartek' problem from hebert -- Thu Feb 16 13:56:51 1989 <<<

Problem: Water leak between M&W chiller and chamber.

Solution: Found wet hose clamp at tygon to water shut-off valve
          at the substrate cooling table. Tightened all hose clamps.

>>> 'gartek' problem from hebert -- Thu Feb 16 16:13:21 1989 <<<

Problem: O2 and Ar mixed in common gas line when introducing
gas to load lock/process chamber.

Solution. I installed new gas lines to ensure that gases never
get mixed in a common line. The lines were pumped out to the
high 10E-8 Torr range on 2/10/89. They were then He leak checked.
No leaks were found. The Ar line was purged thru the load lock
at 1000 sccm for 30 minutes. The O2 line was purged through the
load lock at 200sccm for 30 minutes.

>>> 'gartek' problem from hebert -- Thu Feb 16 19:20:56 1989 <<<

Problem: Need heat blanket for sput chamber cryo to prevent
         O2 from freezing out on 80 deg K array.

Solution: I wired the blanket into the power wire to the
         N2 cryo regen purge switch. The blanket is on at
         all times. The hot N2 purge only works with the
         toggle switch ON.

>>> 'gartek' problem from vallath -- Sun Feb 19 22:35:04 1989 <<<

The sensors in table 1 were not working consistently- sometimes
they did not work.  The problem is not with the wires or connections
but with either the table not moving the full stroke or
with a weak magnet or faulty sensors.
During my last sputter deposition of Nb when making a trilayer
I noticed the shield in the dc chamber was shifted.  I got
my wafer out without any problem.
Afterwards, I ignited another plasma to see whether the
shield was indeed shifted, but was unable to see for sure,
since it looked ok.
The hi-vac open sensor is also not working.
Summary:
1. Intermittent sensor problem with table 1 position.
2. Process station 1 shield shifted?
3. Hi-vac open sensor in load-lock not working.

Vallath

>>> 'gartek' problem from vallath -- Tue Feb 21 14:58:45 1989 <<<

The shield in process station 1 was shifted due to rotated
alignment pin on table 1.
Opened up gartek and fixed shield.
system is now baking out.

>>> 'gartek' problem from mikeb -- Tue Feb 21 15:35:54 1989 <<<

problem:	Turbopump valve sensor lamp didn't come on when
		valve was open.

cause:		Microswitch needed adjustment.

fix:		Adjusted microswitch, resoldered wires.  OK now.

>>> 'gartek' problem from rnorman -- Fri Feb 24 08:53:45 1989 <<<

Problem: Turbo off.

Cause: Low water flow problem.

Solution: Restarted turbo. No problem with the cryos, 4x10-8.
          The gartek is up.

>>> 'gartek' problem from vallath -- Fri Feb 24 12:39:53 1989 <<<

RF water proteus faulty.
Connection to sensor from electronics bad.
Removed and replaced connectors.  It's now up.
VN

>>> 'gartek' problem from mikeb -- Sat Feb 25 18:53:22 1989 <<<


Shuttle crashed (I screwed up).  Shuttle fixed, system pumping out.

Vallath and I spent four hours fixing the shuttle.  There were two 
complications to the procedure: while testing the fixed shuttle we
noticed that a pin on the XPS chamber table interfered with table 
movement, and we had to remove the sideplate to fix it.  We replaced
the sideplate o-ring (number 369, from lab stock, 1 remaining).
Secondly, the RF connection to the aluminum target broke and needs 
repair.

>>> 'gartek' problem from vallath -- Sun Feb 26 14:44:57 1989 <<<

The gartek process chamber has a leak - pressure after about 
18 hrs of bakeout is 1.4 x 10e-6 (hot).
Pressure in the analysis chamber is 3.5 x 10 e-7.
VN

>>> 'gartek' problem from hebert -- Mon Feb 27 16:50:40 1989 <<<

Problem: High base pressure in both sput cham and XPS cham after
         straightening shuttle.

Solution: When the shuttle was bent over the weekend the wafer
          carrier that bent the shuttle was on Table #1. This
          can cause the Pb wire seal in the substrate cooling
          table (i.e. Table #1) to leak. About the highest vacuum
          possible when this happens is low 10E-6 Torr, as was
          reported earlier by mikeb/vallath.

          The Pb seal in the cooling table was replaced.

          The shuttle was straightened, again.
 
          The flange on the right side of the XPS chamber tends to
          leak whenever the flange is removed and not tightened
          evenly or when the o-ring is even slighly flattened. The
          o-ring groove on this flage is too deep and makes it necessary
          to replace the o-ring frequently. A solution would be to have
          about 0.040 inches removed from the o-ring groove side of the
          flange. 

          I removed the XPS right flange, cleaned the o-ring and o-ring
          groove, and tightened the flange screws evenly. The screws
          must be tightened as if one was tightening the bolts to a head
          gasket on an engine: tighten from the inside to the outside
          alternating from side to side and turning the screws only slightly
          with each screw turn. This ensures that they all seat the same.
          If the screws are not tightened evenly enough, the flange will
          not be flat. It will be bowed, and considering the problem
          with the depth of the o-ring groove, not  much bow can occur
          before there is a leak at the seal.

The system is baking out. We'll know in the morning whether or not the
leaks are gone.

>>> 'gartek' problem from hebert -- Tue Feb 28 12:44:46 1989 <<<

Problem: Table 1 sensor not working.

Solution: Magnet on table elevator shaft was rotated. I repositioned
          the magnet, and the sensor now works.


NOTICE: The Pb wire for the crush wire seal in the substrate cooling
table is in my locker in GL3. It's 0.020 inch dia. Gold wire of
the same diameter would also work in a pinch. The seal requires about
3 inches of wire.

DH

>>> 'gartek' problem from hebert -- Thu Mar  2 17:10:04 1989 <<<

Problem: Bent shuttle.

Solution: Straightened shuttle and checked all clearances (e.g. thru
doors and beneath and above all tables, including check with 100mm
wafer carrier on the shuttle).

I adjusted the Table 1 UP/DOWN sensors. The DOWN sensor was sensing
the table fully down when it was only half way down. This would permit
one to move the shuttle while the table is moving down. This alone
could cause the shuttle to bend. The DOWN sensor was adjusted so that it
doesn't come on until the table is all the way down.

In the future, and problem with the table UP/DOWN sensor should first
be tackled by adjusting the magnet position on the elevator shaft.
If this magnet rotates the sensors become less sensitive giving the
appearance that the sensor position needs to be adjusted.

Users should allow about 2 seconds between activating functions. For example,
after switching for table #1 to go down, do NOT switch SHUTTLE LEFT
until the table down sensor LED has been on for at least 2 seconds.

>>> 'gartek' problem from hebert -- Fri Mar  3 15:39:07 1989 <<<

The process chamber pressure rises to the mid 10E-5 Torr range 
when the load lock is vented. The o-rings at this door need to be
replaced.

After processing a trilayer, the process chamber pressure did not
recover below 1.7x10E-6 Torr. There may be a leak in the process
chamber.

EST REPAIR TIME (door seals): 5 hours.

>>> 'gartek' problem from hebert -- Fri Mar  3 16:10:28 1989 <<<

Problem: Process chamber pressure is elevated. It does not go
below 1.5x10E-6 Torr.

I have isolated this problem. It is with the process chamber cryo
pump. If both the process chamber and XPS chamber are pumpoed on 
with only the XPS cryo pump, the base pressure falls to 4.0x10E-8 Torr.
Upon opening the process chamber high vac valve (e.g. using both
cryo pumps to pump on both chambers with the right door open), the
base pressure increases to 1.5x10E-6 Torr. The process chamber cryo pump
is clearly warming up and dumping gas. This pump was just rebuilt a few
weeks ago. It is my understanding that it was NOT tested by the
company that rebuilt it because their compressor was bad. It needs to be
pulled and rebuilt correctly. Last week I did optical emission 
spectroscopy of the Ar plasma while depositing aluminum and found
very large helium peaks. CTI claims that it is rare but possible for
a cryo to have an internal He leak. The fact that the process chmber
cryo does not recover at all indicates that it is saturated. The pump
was regenerated yesterday, so the most likely gas to be saturating
the pump is He.


Gartek users should NOT use this pump while processing trilayers.
Use only the XPS pump by defeating the right door interlock. Leave
the high vac valve to the process chamber closed!

This creates a major problem for our group. The original rebuild
put the system down for 1 week. We need a solution to this problem
immediately.

>>> 'gartek' problem from hebert -- Mon Mar  6 11:33:00 1989 <<<

Problem: Bad RF grounds.

Solution: I rewired all RF grounds to a common earth ground using
          1 inch wide copper sheet metal. The ends of the copper
          strips were tinned with Pb-Sn solder to ensure contact
          integrity.


This problem can be removed from the joblist.
DH

>>> 'gartek' problem from hebert -- Mon Mar  6 17:53:30 1989 <<<

Problem: Bad process chamber cryo pump.



I warm-up the bad pump so that it can be pulled on 3/7/89 and
replaced with another CTI-8.

>>> 'gartek' problem from hebert -- Mon Mar  6 17:54:52 1989 <<<

Problem: Shifted Nb shield. Also, shuttle is either bent or
the 100 mm wafer carrier fell off of the table/shuttle.

DC plasma won't ignite as a result. I'll repair on 3/7 when
the chamber gets vented to replace the process chamber cryo pump.

>>> 'gartek' problem from hebert -- Tue Mar  7 17:03:19 1989 <<<

Problem: Sputter chamber cryo pump is bad.

This problem was reported on Friday, March 3. As of this date it has not
been put on the job list. Has this job been officially assigned? I
thought that the pump was going to be installed today. Will it be installed
tomorrow? I need to get this system up.

>>> 'gartek' problem from hebert -- Wed Mar  8 11:40:03 1989 <<<

Problem: There is a leak below the PC hi vac valve after installation
of the rebuilt cryo pump that was originally on IONA.

I have tightened the cryo flange bolts, checked other flanges attached
to the cryo, and cleaned the cryo check valve. 

The leak doesn't allow the cryo to be pumped out below 1.0x10E-2 Torr.

>>> 'gartek' problem from hebert -- Wed Mar  8 12:20:56 1989 <<<

Problem: Leak aat PC cryo after cryo installation.

I haven't located the leak. I decided to purge the pump with
N2 (heated), taking a chance that moisture in the pump is keeping
it from pumping out.

>>> 'gartek' problem from hebert -- Wed Mar  8 17:34:17 1989 <<<

Problem: Leak at PC cryo after cryo installation.

         The chamber wouldn't pump down below 10mTorr, unless
the PC hi vac valve was closed.

Solution: I capped off the N2 purge at the bottom of the cryo
pump to eliminate any source of a leak from the N2 regen purge
plumbing. The chamber pressure still wouldn't go below 10mTorr.

I dropped the cryo pump and cleaned the flanges, o-ring groove, and
the o-ring. After reinstalling the cryo the chamber pressure
immediately went down to 4.5 mTorr. 

After several N2 purges of the chamber by venting through the load
lock, the chamber pressure steadily improved, falling as low as
1.0 mTorr. It took about 20 minutes for the pressure to reduce to
1 mTorr.

I reinstalled the N2 purge hardware and cleaned the cryo check valve 
again. The chamber now pumps down to 1.0mTorr in about 5 minutes.

I suspect that there is outgassing within the cryo pump that is 
causing it to pump out so slowly. The outgassing may be due to
excessive moisture within the pump. Evidence of this is that the
pressure reduced to 0.1 mTorr within 5 minutes of turning on the
cryo compressor. In such a short period of time the cryo can only
trap water vapor.

>>> 'gartek' problem from hebert -- Thu Mar  9 12:07:51 1989 <<<

Problem: PC cryo warms-up when pumping Ar. 

         Pump S/N 8112869
         Drive S/N 1092

         Pump was previously installed on IONA, and was just
         rebuilt at Cryotek.

Test Condition: PC chamber hot from bake-out.
                Cryo temp measured from Lakeshore Digital readout.
                Three different argon flows through PC cryo and
                    record cryo T after 2 minutes of pumping Ar.
                Test pump by using differential pumping port to
                    ion mill on XPS chamber.

   Ar Flow   PC Pressure (mTorr)   T(K)i   T(K)f   Recovery Time (min)
   170 sccm        4.3              15      18            2
   300             6.3              15      20            2
   500             9.0              15      24            3

   170 (XPS) PC HI VAC CLOSED       16      17            1

Some of the warming of the pump may be due to the chamber being hot. 
Nevertheless, the last test (i.e. 170 sccm Ar (XPS)) indicates that
the pump does warm-up under relatively low gas flows. I have turned
off the heat to the chamber so that the test can be repeated on a cold
chamber later this afternoon.

>>> 'gartek' problem from hebert -- Thu Mar  9 16:21:40 1989 <<<

Problem: PC cryo warms when pumping Ar.

I retested the pump with a cool chamber. At 170sccm Ar flow,
the cryo T still increased by 3 degrees in 2 minutes (14-17 deg K).

>>> 'gartek' problem from hebert -- Fri Mar 10 17:26:21 1989 <<<

Problem: Shuttle bent again. The problem occurs when shuttling
right and then lowering Table #1. 

I'm baffeled by this. It keeps happening and I can't see why.
The shuttle has been straightened at least three times in the
past two weeks. Each time the clearances look good, the shuttle
moves smoothly, and there is no problem. That is until the
chamber is under vacuum. Maybe it's time to replace the shuttle
bearings.

>>> 'gartek' problem from hebert -- Tue Mar 14 18:28:38 1989 <<<

Problem: Shuttle bends frequently. Clearances OK at atm, but 
Table 1 snags the shuttle when the shuttle is in the right
position and Table 1 is moved.

Solution: 1) Replaced all shuttle bearings.
          2) Cleaned precision shaft that shuttle is guided by.
          3) Tightened loose magnet on internal drive that rubbed
             against the rear wall of the chamber.
          4) Adjusted the shuttle right mechanical stop.
          5) Straightened the shuttle.
          6) Adjusted the clearance between the external drive
             magnets and the chamber back plate.
          7) Reduced the speed at which the shuttle moves to
             minimize shuttle backlash whenever it is moved.
          8) Installed and leveled the straightened shuttle. 
             Checked all clearances with/without a wafer on the
             shuttle/tables.

The chamber is baking out at this time. I expect the system to be up
by late Wed/ early Thurs of this week.

>>> 'gartek' problem from hebert -- Thu Mar 16 16:13:15 1989 <<<

Problem: Shuttle bent again. As before, while using the 4in
wafer carrier. This time the shuttle bent when lowering Table1
onto the center shuttle fork. Previously the shuttle bent when
lowering Table1 onto the left shuttle fork.


THE GARTEK IS DOWN.

>>> 'gartek' problem from hebert -- Thu Mar 16 16:23:57 1989 <<<

Problem: Bent shuttle.

Can I get RNorman to wok on this? I've had this problem 4 times
in the past two weeks or so and can't eliminate it. I've done
everything that I can think of to solve the problem. Maybe he can
find sometime that I haven't been able to.rk

>>> 'gartek' problem from rnorman -- Fri Mar 17 13:33:10 1989 <<<

I opened the chamber and found table 1 in the up position and a 4" wafer
carrier under neath it laying on a bent fork of the shuttle. This looks like
the wafer carrier fell off and when the table came down it pushed the carrier
onto the shuttle and bent the fork. The question is, how did the wafer carrier
get off the table to underneath it? It looks like the crash was caused by
the wafer carrier falling off and getting in the way of the downward motion
of the table, the carrier was half on the fork and half on the bottom of the
chamber. 

The shuttle and all was still in alignment, except for one prong on the fork.
Dave is now working of the table. He thinks the table is rotating under vacuum
causing this problem. Another cause could be the shuttle slamming against the
stop and the carrier sliding or flipping off. 

>>> 'gartek' problem from hebert -- Fri Mar 17 15:48:18 1989 <<<

Problem: Bent Shuttle.

Solution: Found that the pneumatic cylinder to elevator shaft
connection at Table 1 was loose. I put on a lock washer and the
machine shop made a special wrench to tighten the connection.
The loose connection caused the table to rotate counter clockwise
and bend the shuttle, probably from the center of mass of the
cooling table assy not being coaxial with the elevator shaft.

I straightened the shuttle and replaced the Pb seal at the
cooling table. The seal was replaced as a precaution -- no
evidence of a leak was found at the seal.

The shuttle was cycled several times in vacuum with the 100mm
wafer carrier without incident.

Our cryo pump that was sent to CTI for rebuild was installed. It is
beingf regenerated. I will be in over the weekend to start the
bakeout if no leaks are present. DO NOT TURN ON THE PC CHAMBER CRYO
COMPRESSOR!!!!

>>> 'gartek' problem from hebert -- Sun Mar 19 08:52:27 1989 <<<

Problem: Leak at PC cryo check valve.

Solution: Removed check valve, cleaned, and reinstalled.

The system is baking-out until 3AM Monday 3/20/89. The chamber
roughed out to the low 10E-5 range with the turbo after the
check valve was cleaned. Before it was cleaned it wouldn't
pump out below 3 mTorr.

>>> 'gartek' problem from hebert -- Mon Mar 20 16:31:58 1989 <<<

The RF doesn't ignite. I've checked all the cables, eliminated
the problem being from the load/tune box and RF switch. All cables,
except the power to the load lock, are good.

The load lock sputter etch power cable needs to be repaired.

A dummy load was placed at the Al target (i.e. disconnect RF cable
from RF switch to the Al target, and connect the cable to the
50 ohm dummy load), and the load match network tuned at 500 W.
That is to say that with the dummy load in place the RF supply seems
to work. However, nothing appears to be wrong with the cable
between the RF switch and the Al target. There is no internal
short in the cable and the target internal resistance is about 650K
ohms (measured RF power in to ground, the resistance is due to the
glycol in the sputter head).

Est Repair Time: 2 hours.

PS: Can we get the DC supply connected to the Al target? I'm tired
of having to deal with RF problems.

>>> 'gartek' problem from hebert -- Mon Mar 20 16:38:44 1989 <<<

The PC cryo that was rebuilt by CTI seems to be working fine.
I did a throughput test today. Using 300 sccm argon (N2 MFC)
for 2 minutes, there was no change in the cryo temperature
(read from H2 bulb). Before the pump was rebuilt, the pump was
observed to warm 2-3 degrees K in two minutes at 300 sccm Ar.

>>> 'gartek' problem from hebert -- Mon Mar 20 16:41:43 1989 <<<

Problem: Need DC power to the Al target.


Was a switch ordered to do this? If so, when will it be in?
Can I get the target switching box from the SFS installed on
the gartek in the interim?

>>> 'gartek' problem from hebert -- Tue Mar 21 11:22:01 1989 <<<

Problem: Intermitent RF ignition.

Solution: RF ignitor was intermitently making contact to ground.
Plasm wouldn't ignite below 20 mTorr. Rf ignitor wire was repositioned.

>>> 'gartek' problem from hebert -- Wed Mar 22 12:25:43 1989 <<<

Problem: Bent shuttle when lowering Table 1 thru center shuttle
         fork.


Solution: Removed substrate cooling table. Inspected the bellows.
No bellows damage was apparent to explain table rotation. The rotation
is not due to excess bellows flexure due to torsion.

The Table 1 pneumatic cylinder is not designed to prevent the 
cylinder shaft from rotating when the shaft moves up or down. I 
replaced all 5 o-rings in the cylinder assy expecting that this might
reduce the amount of shaft rotation. It seems to have helped a little
in reducing rotation, but more so in the smoothness at which the
shaft moves up or down.

The old substrate table (uncooled) was reinstalled. It seems to be
working OK. It has a rectangular elevator shaft where the cooling
table has a round elevator shaft. The old table has an elevator
shaft guide with a square hole in it, apparently to reduce the
amount of table rotation when actuating the air cylinder.

An alignment feature may have to be added to the cooling table before
it can be reinstalled.

The base pressure this morning was in the mid 10E-7 Torr range without
baking out the PC chamber. The chamber is baking out at this time to
see if the base pressure will come down.

The old table is in without any of the pins that lift the ground
shield or locate a wafer. I didn't have the allen wrench necessary
to remove the pins from the cooling table last night. Making new
pins wasn't an option because the wire for this wasn't available
either. The Table 1 stroke is adjusted to bring the clamped was
wafer into the plane of the ground shield. The dep rate should be
verified for this set-up. I don't think that the pins to locate
the substrate are really necessary as long as we are usin a clamped
substrate. I think that they are primarily intended to keep
unclamped 100mm wafers from sliding off of the table during transfer
to/from the shuttle. I prefer not to install these pins until
the system has been run with the old table to see if the shuttle
bending problem has finally been resolved.

>>> 'gartek' problem from rnorman -- Mon Mar 27 13:55:00 1989 <<<




Problem: Intermittant RF Cable to etch station.

Solution: The end connector was changed and the cable was ohmed out and
          there was no short. The gartek up for process tests.

>>> 'gartek' problem from hebert -- Tue Apr 11 16:40:28 1989 <<<

Problem: We need to be able to DC and RF sputter Al.

Can we get the switch installed ASAP? We need the connection
to be made in a way that allows us to switch between DC and RF
power by swapping a cable. 

>>> 'gartek' problem from hebert -- Fri Apr 14 12:50:13 1989 <<<

Problem: High reflected power when sputter etching in the load lock.
         Reduced Al dep rate at 100 W RF (i.e. 20% reduction in 
         dep rate compared to normal rate at 100W).

The reflected power at 30 sccm Ar flow through the load lock is 
normally less than 10W. Today the reflected power reads 25W with
a forward power of 100W. The reflected power goes to zero if the
Ar flow rate is increased to 130 sccm. Reducing the flow by about
10 sccm causes the reflected power to increase to 10 W. Below about
110 sccm the reflected power is 25W. The reflected power continues
to read 25 W even with no gas flow. The reflected power measurement
is obviously incorrect.

>>> 'gartek' problem from hebert -- Mon Apr 17 13:18:37 1989 <<<

Problem: High reflected power during sputter etch in the load lock.

I checked the RF cables with a ohm meter: all cables look OK. I borrowed
an RF cable from the v775 and used it between the RF switch and the
table in the load lock. The reflected power went down from 25 W to
about 15 W (100 W forward power) when the cables were changed.

I dismantled the load lock sputter etch upper electrode. The ground
shield was coated with copper and niobium, and the metal was peeling.
The teflon insulators were coated with metal particles. I cleaned
everything and reassembled. The reflected power was still high after
doing this.

>>> 'gartek' problem from hebert -- Mon Apr 17 20:12:09 1989 <<<

Problem: Load lock B/P = 8.5 x10E-7 Torr. Normal is 2 x10E-7 Torr.
 Pressure has been elevated since reassembly or electrode after
 cleaning this morning.

Solution: Bake-out load lock overnight to see if B/P recovers.

>>> 'gartek' problem from hebert -- Tue Apr 18 17:50:58 1989 <<<

Problem: Bad RF matching in load lock/ difficult to ignite plasma.

Today I put a 50 ohm dummy load at various points in the circuit.
The new RF cable between the RF switch and the PWR IN connector
checked out OK using the dummy load (i.e. the RF supply matched
perfectly to the PWR IN connector using the new cable). Therefore,
the problem isn't with the new cable.

Attaching the RF cable to GROUND to the dummy load created the same
problem -- no matching. I switched the GROUND cable with another
cable that is known to be good, and the problem didn't go away.

The matching problem is within the load lock itself. I checked for
shorting at the table, and this tested to be an open circuit (test
between the table top and bottom, the pieces being separated by a
sheet of alumina.

When using the dummy load at the PWR IN, there was 0 reflected power
at 100W forward pwr. The match was at tune=39 and load=6.

I will continue to look into this on 4/19.

>>> 'gartek' problem from hebert -- Wed Apr 19 19:18:13 1989 <<<

Problem: RF ignition problem.

Previous work reported that the problem was isolated in the load
lock itself.

I used the High Pot to test breakdown voltage of key components in the
RF sputter etch circuit.

   Component                                Breakdown

1) Top RF feedthrough (isolated)           4 kV
2) Lower RF feedthrough (PWR IN)
   (isolated)                              6 kV
3) Top of table to solder joint on
   lower RF feedthrough braid              1.3 kV

   Resoldered braid to ground on
   lower RF feedthrough. Improved
   breakdown.                              3.3 kV

4) Top of table to bottom of table         
   with fastening screws in place          2300uA at zero volts

   Top of table to bottom of table
   WITHOUT fastening screws in place       2300uA at zero volts

5) Fastening screw to teflon washer        3.3 kV

6) Alumina Table insulator (flat sheet
   that isolates the top of the table
   from the bottom of the table)           100uA at 23kV

The problem is with the power IN line
to the table. It is insulated with broken
lengths on alumina tubing. Minimizing the
spacing between the pieces of tubing gives a 
breakdown of 3.3 kV. Bending the cable in any way
causes the powered wire to short to the braid that
covers the alumina tubing, resulting in the high
leakage current at zero volts, as reported above.

This explains why the plasma ignites at high pressure and not at
low pressure as it normally does. The plasma ignites only when the
plasma impedence is lower than the resistance to ground through
the cable insulation.

I need to replace the cable insulator with teflon so that it 
can be routed to the table without shorting. Replacing the alumina
insulators (broken pieces about 1 - 1.5 inches long) with molded
alumina cable insulators would probably also work. Can I get some
help locating the right materials to finish the job?

>>> 'gartek' problem from hebert -- Fri Apr 21 10:03:08 1989 <<<

Problem: I need to be able to sputter Al with either RF or DC
         power.

The DC power to the Al target will be 500W max, and the switch
needs to be able to carry 2 amperes.

I want to be able to switch a cable to either the RF matching network
or the DC switch so that I can select the power source for depositing
Al. This means that the cable is fixed on one end to the Al target.

I reviewed this set-up with Phil on 4/20. He needs the switch from
RNorman so that he can get started on this job.

>>> 'gartek' problem from hebert -- Mon Apr 24 16:36:17 1989 <<<

Problem: RF wouldn't ignight in the load lock below 100 mTorr Ar
         (120 sccm flow rate).

Solution: There were several pints in the circuit that had very low
breakdown voltage. The RF "IN" feedthrough intermitently tested good
with the HI POT TESTER. This feedthrough eventually broke and a new one
was ordered on 4/21/89 from MDC (4-6 week delivery was quoted).

Since we have no plans to sputter deposit in the load lock, I took
the RF out feedthrough and used it in place of the bad feedthrough.
The center connector to the grounded electrode was shorted to the
load lock cover, and RF ground strap was installed to ground the
cover to the sputtering chamber.

The cable in the chamber was completely redone using new ceramic
pieces and teflon tubing. The break down voltage of the cable is
greater than 25kV. 

The weakness in the sputter etch design appears to be the way that table
isolation is achieved. The power cable tended to leak to the lower part\
of the table which is at ground. The upper part of the table is separated
from the lower part by a piece of ceramic sheet that is about 0.075 inches
thick and about 0.5 inches wide. I was able to get the powered cable to the
top of the table and maintain 25kV breakdown until I bent the wire so that
it could be secured to the top of the table with a screw. Because of this,
I drilled out the hole that the powered cable passes through and placed
"fishbone" type insulators around the powered cable. 25 kV breakdown was
maintained while partially bending the cable, but bending caused the
insulator to chip. Whenever the insulator chipped the breakdown was at
near zero volts. I drilled the hole in the upper part of the table 1/32 in
larger in diameter and used a piece of ceramic rod through the table.
With this configuration was was able to bend the wire and secure it to the
top of the table and maintain a breakdown voltage of about 1.5kV.

The sputter etch in the load lock is UP.

>>> 'gartek' problem from hebert -- Thu Apr 27 10:25:47 1989 <<<

Problem: Blown circuit breaker effecting MFC controller.
         Both cryo pump compressors OFF, causing cryo's to dump.

I don't know why this happened. I'm regenerating the cryo pumps
at this time.

>>> 'gartek' problem from rnorman -- Mon May  1 08:19:01 1989 <<<

The H2O REC will be going down at 8:30; though there will be city H2O to
back us up. So, I closed the L/L gate valve, and shut down the turbo. I
also bypassed the water interlock for the cryos, if there is not enough
back up water to cool the compressors a thermo-switch will turn them off.
If this happens it will not hurt the compressors, but just shut off the cryos.
If all goes well your cryos will stay running and the system under high vac.

>>> 'gartek' problem from rnorman -- Mon May  1 09:57:34 1989 <<<

Problem: REC water going down; turned off turbo and bypassed cryo interlock.

Solution: After REC water came back up I restarted the turbo and took out
          interlock bypass. The cryos are fine and the system is up.

>>> 'gartek' problem from parrish -- Tue May 23 08:43:41 1989 <<<

oil level low in the turbo-pump, need filling.

>>> 'gartek' problem from hebert -- Tue May 23 10:41:20 1989 <<<

Problem: Low turbo oil level.

Parrish may add oil at any time.

>>> 'gartek' problem from parrish -- Tue May 23 11:50:40 1989 <<<

problem:    oil in turbo-pump low

cause:      marks on fill line not clear.

solition:   added oil and remarked oil lines

>>> 'gartek' problem from rnorman -- Tue Jun 13 16:27:43 1989 <<<

Problem: The DC supply would not turn on.

Cause: The Target #1 timmer was bad; no output signal to
       DC supply to turn on.  The timmer acts like an interlock
       for the DC supply; when the timmer turns on an internal
       relay closes which completes the interlock for the DC supply;
       like an on/off switch.

Solution: I switched the timmer with the RF igniter timmer. The 
          gartek was tested and is up.  I will order two new timmers;
          one for the Rf igniter and a spare. 

>>> 'gartek' problem from hebert -- Tue Jun 20 17:37:42 1989 <<<

Problem: Base pressure is about 1.7 x 10E-7 Torr after cryo
         dumped over weekend from power failure in the lab. 

Solution: I'm warming the cryo pumps tonight and I'll clean the
          cryo check valves in the morning.

>>> 'gartek' problem from hebert -- Wed Jun 21 18:37:50 1989 <<<


Elevated base pressure (1.5 x 10E-7 Torr). 

I cleaned and reinstalled the check valves to both cryo pumps.

The base pressure did not go below the above BP after cleaning
the check valves.

The system is on timed bakeout. The heat tapes will turn off at
3 AM on 6/22. Let's see if baking the chamber reduces the base pressure.

>>> 'gartek' problem from hebert -- Thu Jun 22 11:21:10 1989 <<<

Problem: Elevated base pressure after power failure last weekend.

Solution: Cleaned cryo check valves and baked out the chamber
overnight. The base pressure this morning after presputtering
both targets is 6 x 10E-8 Torr.

The chamber will be baked again tonight.

The gartek is UP.

>>> 'gartek' problem from hebert -- Fri Jul  7 15:28:23 1989 <<<

Problem: Shorted RF feedthrough prevented load lock sputter etching.

Solution: Ireplaced the feedthrough. The RF sputter etch now works.

>>> 'gartek' problem from vallath -- Mon Jul 10 23:00:05 1989 <<<


Water interlock for the rf power supply does not work. I saw that
the proteus was open.  I think it may be the connection to the
flow sensor (two crimp connectors) since I think I remember 
that we had a problem there once.  I used the system by defeating the
interlock.
Vallath

>>> 'gartek' problem from rnorman -- Wed Jul 12 11:11:07 1989 <<<

Problem: The water interlock for the RF power supply doesn't work.

Cause: The Proteus H2O interlock switch was bad.

Solution: I replaced the Proteus; was tested and works.  I also
          installed the RF igniter timer. NOTE: The timer is
          a 99.9 second timer, not 999. The gartek is up.

>>> 'gartek' problem from rnorman -- Tue Jul 18 14:43:09 1989 <<<

The RF timer mounting bracket arrived and was installed on the
gartek.  The RF timer is now mounted properly and is working.

>>> 'gartek' problem from mikeb -- Tue Jul 18 17:47:28 1989 <<<

The magnet used for the table2 down sensor came loose and fell on the
floor.  It should be properly replaced before moving table2.  Gartek
is up for operation using table 1 and load lock only.

>>> 'gartek' problem from mikeb -- Sat Jul 22 12:00:36 1989 <<<

problem:	Table2 position sensors not working.

cause:		Magnet fell off.

solution:	Glued magnet in place using RTV silicone adhesive.

Also, the rf water interlock is now wired to the table1 proteus, which is 
currently serving both targets.  The connection is made with jumper wires 
and prevents the operator from igniting an rf plasma without water flow.

>>> 'gartek' problem from hebert -- Thu Aug 17 10:29:58 1989 <<<

Problem: Turbo oil needs to be changed. The best time to do this 
would be Monday or Tuesday of next week. I prefer that the oil be 
changed in the morning before 10AM.

>>> 'gartek' problem from parrish -- Tue Sep  5 09:13:34 1989 <<<

oil has been changed in the Turbo-punp, Turbo-backing pump, and main
chamber roughing-pump.

>>> 'gartek' problem from hebert -- Thu Nov 16 11:02:25 1989 <<<

The crt display to the XPS PHI computer won't turn on.
It worked before the power shutdown 10 days ago, but
after powering the computer back up after the shutdown,
the crt didn't work.
From hoagland Mon Nov 27 16:54:47 1989
Received: by argon (5.57/1.16)
	id AA04467; Mon, 27 Nov 89 16:54:44 PST
Date: Mon, 27 Nov 89 16:54:44 PST
From: hoagland (Steve  Hoagland)
Message-Id: <8911280054.AA04467@argon>
To: hebert
Subject: Perkin-Elmer monitor
Cc: bob, hoagland, voros
Status: RO

I removed the power supply from your defunct monitor and found
at least 1 blown transistor.  The monitor is an OEM unit made by a
company called Barco for Perkin Elmer.  PK's field engineers may have
applicable schematics we could copy, but field service was not available
today and will contact us when they have more information.  I attempted
to find a replacement for the transistor, a Motorola part.  None of our
vendors could locate it.  I put in a request with Motorola for information
on the transistor, but they could find nothing on their current database(old
part) and will call back if they have any luck checking further.  Greater
efforts will be made tomorrow to locate Barco and see if they can provide
parts or information for this equipment.





>>> 'gartek' problem from hoagland -- Fri Dec  8 09:02:35 1989 <<<

After a weeks worth of discussions with Perkin-Elmer and Concurrent
about who assembled this monitor, who has documentation, and who
can release it, it was determined we have no access to schematics
or further technical information.  Based on tests made on a 
scorched power supply within the CRT, a bad diode and transistor
were located and replaced.  The power supply then tested OK, however
when installed and operated within the monitor it shutdown after
a couple of seconds later, most likely due to some other failure
in another section of electronics.  Dave wants this unit up ASAP
so he plans to rent another unit from Perkin-Elmer, then use that
shipping container to send the broken one back to Minnesota for
service.

>>> 'gartek' problem from hebert -- Fri Dec 15 10:08:44 1989 <<<

The cryo pumps were regenerated today.

The gartek is up.

>>> 'gartek' problem from hebert -- Fri Dec 22 12:26:32 1989 <<<

I installed the fixture in the load lock that would have
permit us to sputter deposit. The plasma ignited and after about
30 seconds to 1 minute at 300 W DC, the target got hot enough to
cause the teflon insulators to char and damage a weld on the lid.
The leak in the lid has been fixed by Ben in the Machine Shop (by 
welding). The load lock pumps out OK now. The load lock was cleaned
and the substrate table is still out. The Gazrtk is down until the
first week of Jan '90 when I will reinstall the substrate table.

The sputter dep project will not continue as a result of this incident.
I will suggest to Prof. Van Duzer that we get targets for the TOPGUN
that will permit us to contuct experiments in that system.

>>> 'gartek' problem from hebert -- Wed Jan  3 08:37:20 1990 <<<

Problems: 1) Ar MFC (0-1000 sccm) leaks Ar at 500 sccm when the
             controller is programmed for 0.0 sccm flow.
          2) The target cooling water needs algae removed and
             the coolant replaced.

Solution: 1) The Ar MFC that looked bad on 1/2/90 is OK today.
             The MFC must have had a dust particle preventing
             it from fully closing. I will NOT need the UNIT MFC
             afterall to replace the Vac Gen MFC.

           
Problem #2 still needs attention.

>>> 'gartek' problem from hebert -- Thu Jan  4 14:36:28 1990 <<<

Problem: Load loack sputter etch station not working.

Solution: I cleaned and rewired the power cables to the
powered substrate table. I improved the isolation between the
top powered section of the table and the grounded bottom section
of the table by replacing the stainless screws and teflon
insulators with nylon screws and teflon insulators. 

The top ground shield had been modified during the load lock
sputter deposition project that failed in late December '89.

The plasma not ignites at pressures as low as 10 mTorr. It used to
ignite only about 20 mTorr. The RF pot setting for 100 W is now
1.60. Previously it was 2.40.  

I checked the etch rate of thermal oxide (SiO2) on a clamped 100 mm
wafer. Previously the thickness varied by 9 to 10 percent after sputter
etching a substrate. Now the thickness variation is less than 1.0 percent
after sputter etching a 100 mm wafer.

THe sputter etch station is up.

>>> 'gartek' problem from rnorman -- Mon Jan 22 08:40:36 1990 <<<

The cryo's were regened after the power shutdown, but due to a misunderstanding
were regened again.  At this time the cryo's are under regeneration,
Dave is taking care of it.

>>> 'gartek' problem from bob -- Tue Jan 23 10:47:17 1990 <<<

The cryopumps on the gartek have been regenerated and the gartek is
up for use.

>>> 'gartek' problem from evan -- Wed Feb 28 14:31:32 1990 <<<

The taurus box does not seem to be activated when gartek enabled.  Is this
actually hooked up to anything?  

>>> 'gartek' problem from hebert -- Mon Mar  5 16:06:32 1990 <<<


Problem: Quartz lamp in load lock caused RF cable to melt. Teflon
deposited on LL surfaces. The heat control was set to 80 deg C.

Solution: I dismanteled the load lock and I'm cleaning all parts
and surfaces. I should complete this in the morning.

>>> 'gartek' problem from hebert -- Wed Mar  7 09:21:26 1990 <<<

Problem: Quartz lamp in load lock overheated RF cable to powered
substrate table and caused teflon insulation to melt. The teflon
coated the inside of the load lock.

Solution: I took apart the load lock and cleaned all surfaces.
I ran on O2 plasma ans Ar plasma afterwards to to further clean the
chamber.

The RF sputter etch is now UP. I sputter etched a 100mm wafer (clamped)
using 100 W RF (2.0 RF pot setting); 50 sccm Ar (7 mTorr); for 300 seconds.

              T        C       B       L       R
Init XOX   1399 A    1400   1399    1396    1395
Init Nf    1.459     1.458  1.459   1.456   1.459

Final XOX  1295 A    1287   1292    1296    1286
Final Nf   1.459     1.459  1.459   1.460   1.459

etch rate: 21.4 A/min

Since Nf is the same for SiO2 before and after sputter etch cycle,
the load lock seems to be clean enough that no teflon residue is present
to deposit on the substrate.

>>> 'gartek' problem from hebert -- Wed Mar  7 09:31:11 1990 <<<

Problem: Unstable DC power supply. When pot is set at 4.68, the power
varies from 2.9 to 3.2 kW. Normally the power is 2.9 to 3.0 kW.
There is also a dead band in the pot, where changing the pot setting
has little or no effect of output power. THe problem exists with both
the Al and Nb target, and when directly linking the power supply to each
target (by-passing the DC target switch).

There is probably a bad SCR in the supply. Peak power is only 4.5 kW.
It should be 5 kW.

>>> 'gartek' problem from rnorman -- Thu Mar  8 11:18:19 1990 <<<

Prower fluctuating from 2.9 - 3.0 KW is now fluctuating from 2.9 - 3.2 KW.
Dead band in knob and max power is now 4.5Kw instead of 5.0Kw.

See comment for details.  Most of these problem could be caused by plasma
fluctuations or Process changes, such as pressure.  I need to talk to
Hebert before continuing.

>>> 'gartek' problem from hebert -- Fri Mar  9 15:35:14 1990 <<<

The load lock mech pump oil filtration system has a torn poly
tube that caused oil to leak onto the floor in CY1. I turned off
the mech pump and put towels around the oil on the floor.

>>> 'gartek' problem from bob -- Fri Mar  9 15:59:16 1990 <<<

TThen oiline on the filtration unit if the gartek broke at the 
connection to the oil filtration pump. It has been repaired and
the oil cleaned up. The system is up.

>>> 'gartek' problem from hebert -- Fri Mar 23 08:01:40 1990 <<<

Power glitch caused turbo to shut dpown and load lock to vent. Leaky
seal at door between load lock and process chamber caused pc cryo
to dump. The cryo pumps rae being regenerated.

>>> 'gartek' problem from hebert -- Mon Mar 26 15:44:42 1990 <<<

Problem: PC base pressure is fluctuating between low 10E-6 Torr
and low 10E-7 Torr. This morning the base pressure was in the
mid 10E-8 Torr range. At 11 AM the base pressure started fluctuating.

The XPS chamber is at about 8 x 10E-8 Torr. When the door separating
the PC and the XPS chambers was opened, the pressure was in the 
low to mid 10E-7 Torr range. 

The He pressure in the compressor is 260 to 270 PSI.

I'm not sure at this time what is causingf the base pressure in the
PC to fluctuate. Both cryo pumps were regenerated last Friday after
a power glitch caused the pumps to warm to 30 deg K. The load lock
to PC door leaked and caused the PC cryo to warm.

>>> 'gartek' problem from hebert -- Tue Mar 27 08:15:06 1990 <<<

Problem: Base pressure in PC fluctuating after sputtering. The 
base pressure recovers after several hours.

Solution: RNorman recharged the cryo compressors to the XPS and
PC chamber. The base pressure in the PC fell from 5.6 x 10E-8 Torr
to 4.3 x 10E-8 Torr after recharging the PC cryo compressor. The
PC cryo compressor was about 30 psi low.

The Gartek cryos will now be regenerated.

The Gartek will be up either this afternoon or in the morning, depending
on how fast the base pressure gets into the 10E-8 Torr range after
regen.

>>> 'gartek' problem from hebert -- Mon Apr  2 21:28:57 1990 <<<

The base pressure in the PC is not recovering after sputtering.
Last week I had a problem with the base pressure fluctuating
between the 10 E-6 Torr range and the 10E-7 Torr range after sputtering.
Normally it recovers to the mid 10E-8 Torr range within a minute or
two of sputtering any film. 

Last week He was added to the compressors. I ran a test afterwards to
check the critical temperature of our niobium film. For the test, I
pumped 150 sccm Ar with the PC cryo for 10 minutes, then I deposited
3000 A Nb (90 s dep t), and the base pressure fell into the low
10E-7 Torr range right away after sputtering. The Tc of the film was
9.2 degrees Kelvin (perfect).

Tonight the base pressure in the PC stays in the low 10E-6 Torr range after
sputtering. The PC pressure falls into the 10E-8 Torr range if I pump
on the PC and XPS chamber with only the XPS cryo, and the pressure
in both chambers rises into the 10E-6 Torr range with both pumps
pumping on both chambers.

Initially the base pressure in the PC was 6.5 x 10E-8 Torr. I am warming
both cryo pumps. I'll swap the compressors in the morning and clean the
cryo check valves. I'll start up the compressors and see if the problem
goes away. If not, the PC cryo may have to be rebuilt. 

>>> 'gartek' problem from hebert -- Tue Apr  3 11:22:48 1990 <<<

OES of Al dep using the same scan conditions as last year show
an increase in N2 and He peak heights since the pump was rebuilt.
This suggests that there is a He leak within the cryo pump that 
is causing the arrays to saturate, and N2 to dump.

I will pull the cryo pump off of the PC this afternoon. 

>>> 'gartek' problem from hebert -- Tue Apr  3 15:53:25 1990 <<<

The process chamber cryo pump (CTI S/N:16J47899) was replaced with
a new CTI-8 (S/N 9J47650). The old CTI-8 was rebuilt in March '89.
It was determined by OES that the old pump was dumping O2, N2, and
COx (x=1 or 2). The He peak height was also greater than that observed
last year after the rebuild. 

The N2 purge was not installed onto the new cryo pump. The new cryo has
a digital cryo temp readout.

I'd like to thank James Parrish for his assistance in the installation of the
new pump!

I'm roughing out the cryo's over night to see base pressure before the 
cryo compressors are turned on. 

>>> 'gartek' problem from hebert -- Wed Apr  4 16:06:05 1990 <<<

The DC sputter controller won't power up. I tried resetting the
circuit breakers and turning the key, but was not successful
at getting it to turn ON.

>>> 'gartek' problem from rnorman -- Thu Apr  5 08:01:55 1990 <<<

Problem: DC supply control panel not turning on, no LED display.

Cause: Fuse F5, 24 volt supply, a 2 amp fuse was blown.

Solution: The DC power supply was opened up.  The AC power from the 
          main braker was present.  Below the main braker is 5 fuses
          the far right one is F5 for the 24 volt supply, it was replaced.
          The gartek is up.

>>> 'gartek' problem from hebert -- Thu Apr  5 11:05:47 1990 <<<

1) Power supply (DC): RNorman replaced fuse in the DC supply, and
   DC controller came up.
2) I installed the heater blanket on the new PC cryo pump.
3) The base pressure this morning before presputterinfg the targets
   and before degas of IG was 2.5 x 10E-7 Torr. The heat tapes overloaded
   the power strip, so I don't know how long the chamber was baked.
   The base pressure is normal for a chamber that has been baked out for
   less than 24 hour after cleaning. 
4) The PC chamber cyo temp remains at 8 deg K after pumping on the chamber
   for almnost 24 hours. This pretty much eliminates any concerns about
   chamber leaks after the chamber was disassembled for cleaning.
5) I tested the new cryo pump to see how much it warmed up when pumping
   Ar. Pumping 240 sccm Ar for 20 minutes, the cryo temp did not exceed
   10 degrees K. Raising the flow rate to 640 sccm, the cryo temp did NOT
   exceed 16 degrees K, after pumping at this flow rate for about 2 minutes.
   The Ar flow rates have been corrected (i.e. mulitiply N2 cal MFC reading
   by 1.6).
6) I am baking the chamber again to help reduce the base pressure.

>>> 'gartek' problem from hebert -- Fri Apr  6 07:38:20 1990 <<<

The cryo compressors have low He pressure. This is from swapping
compressors before the new cryo was installed, and then compounded
by the installation of the new cryo pump. The He pressure is about
260 PSI in the compressors. 

I'd like to have these recharged with He.

>>> 'gartek' fix from hebert -- Fri Apr  6 08:04:09 1990 <<<

RNorman recharged both cryo compressors with He.

The base pressure this morning in the process chamber was 1.0 x 10E-7 Torr
before presputtering the targets or degassing the IG. This is good,
considering that the heat tapes turned off at 4 AM and that it takes
about 6 hours for the chamber to cool down. After presputtering the
targets and degassing the IG the PC base pressure is 5 x 10 E-8 Torr.
Normally it is about 2 x 10E-8 Torr after doing this. I expect the
system to UP by 10 AM this morning.

>>> 'gartek' problem from parrish -- Mon Apr 16 08:40:02 1990 <<<

oil line busted on the filtration unit,system is down

>>> 'gartek' fix from parrish -- Wed Apr 18 10:22:14 1990 <<<

problem:    oil on the floor

cause:      oil line broke

solution:   cleaned up oil,refilled pump(without oil filtration unit)and
            started pump.

>>> 'gartek' problem from hebert -- Mon May 14 12:15:27 1990 <<<

The mech pump that backs up the load lock turbo is noisy. It sounds
like the fan to the electric motor is loose.
