From shott at stanford.edu  Wed Jan  7 18:59:57 2009
From: shott at stanford.edu (John Shott)
Date: Wed, 07 Jan 2009 18:59:57 -0800
Subject: Drytek2 update .... limited success.
Message-ID: <49656C2D.5020006@stanford.edu>

Drytek2 Fans:

I know that the drytek2 has been ailing for some time.  I can report 
limited success in terms of addressing the low, drifting RF power.  For 
some time people have seen a maximum power of from 300-400 watts .... 
and that power tends to drift over time.

While we have not been able to restore the full 1000 watts that this 
supply is supposed to be able to generate, we have at least been able to 
find enough minor problems that it will now generate and control powers 
of 500W.  I just ran a 30 minute oxygen plasma and it was able to 
generate and control 500W of forward power during the entire time.  
While it doesn't have great margin above 500 watts .... maybe only 525 
or 550 after it heats up .... I think that is both better than it was 
and likely usable for a greater fraction of the drytek community.

While we didn't find any one major problem .... and still haven't really 
found out what is preventing it from generating full output power .... 
we did find and fix several minor things that were contributing to the 
limited power problem including aging electrolytic capacitors on the +12 
and +40 VDC supplies, an undersized (wattage-wise) emitter degeneration 
resistor in the RF oscillator, and a couple of marginal components in 
the gain control stage.

After consulting with Jim McVittie, I also made an initial attempt to 
address the RF tuning problem .... where we don't every get a great null 
in the  reflected power and only approach tune at minimum coil and 
maximum capacitance.  I'm sorry to report that my initial efforts this 
afternoon didn't result in any significant improvement in that 
situation.  As a result, I'll try to spend more time with Jim to see 
where either my understanding or my approach failed to produce results.

While waiting to better resolve the tuning issues, however, I believe 
that we have at least improved the RF power situation for those of you 
needing to run at 500 watts.  As a result, I believe that many of you 
may cautiously begin to use this tool again and I'll hope that Nancy can 
do some etch rate runs tomorrow at 500 watts RF power to double check 
etch rates for standard recipes and conditions.

We'll hope to have resolved the remaining tuning issues in the near future.

I personally apologize to all of your for the extended delays in getting 
this tool fully operational and take full responsibility for those delays.

Please report your experiences or problems when using this tool.

Thank you for your continued support,

John



From shott at stanford.edu  Wed Jan 21 16:31:14 2009
From: shott at stanford.edu (John D Shott)
Date: Wed, 21 Jan 2009 16:31:14 -0800 (PST)
Subject: Improvements to RF tuning ....
In-Reply-To: <242297072.1699421232584210317.JavaMail.root@zm07.stanford.edu>
Message-ID: <1687909719.1699571232584274547.JavaMail.root@zm07.stanford.edu>

Drytek2 community:

Today, based on some recommendations from Jim McVittie, Team Drytek made some changes to the RF matching network.  Following these changes, we have now been able to see a nice minimum in the reflected RF power.  With a forward power of 425 watts, we were able to consistently tune to a reflected power of 20 watts or less .... and not be at the extremes of either the coil or the capacitor.

For Process 1 running the oxygen plasma recipe, I found optimum tuning at a coil setting of 1.60 (out of 10) and at a coil setting of about 2.2 (also out of 10).  Previously, "near tune" was occurring at a coil setting of zero and a capacitor setting of about 10.

It would be a help to others if people would record the coil and cap settings for different recipes (and when run in process 1, 2 and 3).

I am hopeful that this will at least result in better tuning performance on this tool .... and will now turn my attention back to the problem of consistently achieving higher power settings.

Please let us know if you encounter any serious tuning problems and we will do our best to address them.

Happy processing,

John