[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[CCB-930] RF Power Accounting Issue with CCR#20

Dear Colleagues,

CCB has received a communication on Nov.16 from Brian Chase concerning 
the topic above. Although all of you must have received a summary
from Stefan Simrock and a write-up from Sergei Nagaitsev earlier,
since not all of you may have received Brian's email, I am  attaching it 
below. I hope this is OK with him.

Brian has produced an excel sheet in which he tried to list up the
power losses to consider in each of our systems. While there may be some
differences of opinions among us all, I would highly appreciate
his attempt at putting many issues to consider into a very nice perspective.

Now, I have picked up Brian's excel sheet and created my version
where I tried to compare his claim with what I think we heard from
Chris Adolphsen during the CCB hearing at Valencia on Nov.9.

I am attaching my excel sheet, too.

The first page shows the calculations for 9-8-9.

Several notes.

1. Brian proposes to consider 5% de-rating factor for the klystron
   performance during its lifetime, and has done his accounting as such.

   My objection is that when we say 10MW, we had better mean to mean
   that a tube is to produce 10MW for ITS ENTIRE LIFE. In other words,
   "10MW lifetime" is the lifetime + performance to consider for klystrons.

   (I think CCB had better clarify this "assumption" in its recommendation
   report, whatever CCB ends up saying)

   Therefore, I produced my column labeled "Toge (20061120)", and it
   it the de-rating factor is 0%.

2. Brian is considering 3% loss due to the circulator, besides the
   waveguide, and he also considers modulator ripples, and power loss
   due to cavity gradient variation in the area of HLRF. His net
   HLRF power loss amounts to 18%, as opposed to Chris's 5% or Toge's
   13%, whose only difference from Brian is that Toge does not consider 
   klystron de-rating 5% to be a valid entry to consider here.

3. Brian moves on to account for LLRF loss factors, or loss factors
   and tuning factors put together, to be more exact. By taking a
   linear sum of these factors he ends up with 12.3%, while Chris
   claimed that approximately 10% is available.

   Brian's individual accounting for the LLRF factors seems reasonable
   to me. However, I am not entirely sure if we should take a linear sum 
   or a square sum for these LLRF elements, so I have inserted rows 
   indicating both. The truth would be somewhere in between, I guess.

4. The net result is that if one uses up all the "headrooms", Chris
   claims we can still reach 33.5MV/m. However, in Brian's accounting
   he can reach only 28.2MV/m... or the other way of saying is that
   Brian's result is short of the power headroom by 1.36MW. If we
   adopt a square-sum of LLRF loss factors, Brian reaches 30MV/m
   or the power headroom is -0.90MW. Similar calculations are shown
   for Toge's case where klystron de-rating is not assumed.

5. In summary, if we adopt Brian's ways of accounting for the HLRF
   and LLRF factors, although there are some variations of numbers 
   depending on exactly what you include how, we are barely able to ensure 
   operation at 31.5MV/m,
The second page shows a case with 8-8-8. The differences of this page
with respect to the first page are

- Somewhat bigger loss factor for the WG, and
- 24 cavs instead of 26 cavities.

It shows, even in case of 8-8-8, if we follow Brian's accounting,
we are barely ensuring 31.5MV/m.  

If that is the case, what we are observing is that -

    Before even talking about 8-8-8 vs 9-8-9, the CCB will
    have to ask the ML AG leaders to redo an adequate power accounting
    and come up with a version of the ML unit design that safely ensures
    operation at 31.5MV/m "on average" first.


Thank you for your attention and comments, remarks, words of wisdom,
whatever constructive on this matter.


- Nobu Toge (KEK, Accelerator Lab)
  email: toge@xxxxxxxxxxxx
  voice: +81-29-864-5224
  fax:   +81-29-864-3182

Attachment: CCR20-RF-Toge20061120A.xls
Description: MS-Excel spreadsheet

--- Begin Message ---
To address the complexity of  CCR#20, the LLRF team of Sefan Simrock,
Shinichiro Michizono and Brian Chase have prepared a spreadsheet to create a common format for discussion. The spreadsheet gives place holders for parameters
that impact the power delivered and the power required by the cavities.
The values that we have chosen are our best 50% estimates. The other two sheets explore the effect of detuning and multi-gradient response to beam and no beam conditions.

Best Regards,
Brian Chase

Attachment: ILCPower_Mic.xls
Description: Binary data

--- End Message ---