[BTeam] mechanisms for bleed-through to D?

Thu Sep 23 10:21:59 EDT 2021

Hi folks,

Here is a slide I have shown at several GlueX collaboration meetings.
It illustrates the issue as I understand it with Hall D bleedthrough
when a high-current hall (e.g. Hall C on the slide) is also running
on 5th pass and Hall D/C lasers are at 249.5 MHz.

Any extinction bleedthrough from C beam to the C' bucket in the
picture will also end up going to Hall D since the phase of the
748.5 MHz 5th pass separator is the same for the C' and D buckets.

My impression was that it is difficult to get nA-level extinction
from the C laser in the C' bucket when C is running at high current.
Hence there is usually some bleedthrough. Note that this bleedthrough
goes through a different slit than the slit being used by the Hall D
beam, so closing the Hall D slit doesn't eliminate this bleedthrough.

The Hall D timing measurements of bleedthrough should be able to
discriminate whether this mechanism is plausible.

                                                 -Todd

======================================================================
Todd Satogata
Director, Center for Advanced Studies of Accelerators
Jefferson Lab
12000 Jefferson Avenue
Newport News, VA 23606
Cell: (631) 807-0674
======================================================================

On 9/23/21 10:05 AM, Michael Tiefenback wrote:
> Gentle folk:
> 
> Some of my comments about potentially avoiding bleed-through from the C beam into Hall D have been predicated on Yan Wang's recent assertion that the D bleed-through is not an injector-sourced issue.  I have not understood his line of reasoning, but that may be simply my shortcoming.  His assertion, if I interpret it corrrectly, is that the bleedthrough to Hall D is the result of scattering from the C-directed beam pulses into the D line.  Hall D's timing results seem to run counter to this.
> 
> The RF separator system ought to filter beam temporally to Hall D.  In principle, nothing should enter the D line from any hall for which beam is extracted at lower than 5th pass.  The electrons from the relevant chopper slit should be collected by the 499 MHz lower-pass separator(s).  In our present case, the B beam is extracted at 2nd pass, and should not make it into arc 4A. ("Should not" is an operative and important qualifier.)  Hall C beam is configured to pass through to the AE extraction 
> line along with Hall D beam, to be separated by the 750 MHz horizontal deflection.  Now, D is using a different chopper slit (slit A) than C.  Any laser afterglow from C illuminating the photocathode at the time of the D laser pulse would arrive at Hall D at the same time as D's main beam pulse.  This appears not to be the case.
> 
> The bucket sequence from the injector's perspective, as I recall, is B-A-C with respect to the chopper slits.  With D using the A slit, and using lower case to represent the pulses which would be extracted in the AE line to Hall D, this might be represented by
> 
> B   d   C   b   D   c   [B   d   C   b   D   c] ....
> 
> at a 1499 MHz sequence rate.  The 4E separator pulls the B and b pulses off into Hall B.  The 750 MHz chopper in the AE line should receive the D, d, C, and c buckets and drive the d and c pulses into Hall D's line, directing the D and C pulses into the Hall C line.  The D pulse in this labeling is ideally empty, as is the c pulse.
> 
> Today's 8 am meeting referenced Hall D's timing data posted in
> https://logbooks.jlab.org/entry/3910800 <https://logbooks.jlab.org/entry/3910800>.
> This diagram seems to indicate that the bleedthrough experienced by Hall D comes from the "c" pulse timing passed through the C chopper slit.  This seems be counter to Yan's "not from the injector" comment, and appears to indicate that the Hall C laser has a DC component which is passed through the chopper system to Hall D as the (supposedly empty) "c" pulse.  This is not readily eliminated from the system as presently installed.
> 
> We have experienced beam halo directed into Hall B from scattering on the (narrow) apertures of the vertical (499 MHz) separators used to split beams to A/B/C halls, and it is such scattered beam that I hoped to find ways to suppress for Hall D.  What I see in the data presented to date indicates that Hall D's bleedthrough beam is indeed due to the Hall C afterglow (DC light) generating beam in the nominally empty "c" alternate bucket.  That bucket is intrinsically passed through to Hall D.
> 
> HOWEVER, such beam can be eliminated in the future with enough budget to provide a 499 MHz deflecting cavity at a judiciously chosen location in ArcA (arc 10).  Such a cavity would function in the same way as the 499 MHz separators in the lower passes, passing the d pulse and diverting the (quite low-current) a and c pulses into some collimator "dump."  This has the prospect of generating a nasty radiological issue, but perhaps might be considered.
> 
> Hall D's measurement mentioned above ought in this analysis show quite similar bleedthrough to D with the C slit closed, D laser off, and the A slit open.  The DC level from the C laser should still illuminate the A slit and pass beam at 2 nanosecond intervals into Hall D at their "normal" beam arrival time.  Detection of the c pulse within the accelerator could in principle be done with the SLMs, if they are operating at previously observed sensitivity.  We have seen light from Hall B beam in 
> the Arc 1 SLM at less than 1/2 nA (perhaps ~100 pA) current a very long time ago.  The ArcA SLM should be able to see this with the aperture slit and other attenuators removed, as should the BS SLM.  It is not clear how useful this would be, considering the above mechanism for bleedthrough into D.
> 
> If I have missed something important, and the above content is somehow misguided, I apologize for the diversion.
> 
> Michael Tiefenback
> 
> 
> _______________________________________________
> BTeam mailing list
> BTeam at jlab.org
> https://mailman.jlab.org/mailman/listinfo/bteam
> 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Screen Shot 2021-09-23 at 10.15.45 AM.png
Type: image/png
Size: 1693148 bytes
Desc: not available
URL: <https://mailman.jlab.org/pipermail/bteam/attachments/20210923/1f8de736/attachment-0001.png>