Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://zebu.uoregon.edu/~uochep/talks/talks07/emi.pdf Äàòà èçìåíåíèÿ: Fri Jun 29 01:20:33 2007 Äàòà èíäåêñèðîâàíèÿ: Tue Oct 2 09:39:39 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: mercury program

Beam induced EMI and detector electronics
Nick Sinev, University of Oregon Gary Bower, SLAC Yasuhiro Sugimoto, KEK


Important Contributors
· · · · · · · · · · Ray Arnold, SLAC Karl Bane, SLAC SLAC Eric Colby, SLAC Joe Frisch, SLAC Doug McCormick, SLAC Marc Ross, SLAC Yasuhiro Sugimoto, KEK Mike Woods, SLAC SLAC Hitoshi Yamamoto, Tohoku University Japan-US Cooperative Program in High Energy ti Hi Physics by JSPS
Fermilab SiD Workshop 2

April 7, 2007


The problem problem
· Since the advent of short bunch length colliders in the 70s, the effect of beam induced EMI on proximate detector electronics has been a worry. · There have a been a few cases where it ay played ole may have payed a roe. · There have been some limited and inconclusive efforts at exploring the problem efforts at exploring problem · But it is a difficult situation to study.
April 7, 2007 Fermilab SiD Workshop 3


Beam induced EMI
· EMI = ElectroMagnetic Interfence in the form of RF radiation. · E&M theory predicts a moving bunch charge creates dynamic EM fields. · In the proximity of matter (eg, accelerator beam pipe elements), these fields are complex in structure.

April 7, 2007

Fermilab SiD Workshop

4


Calculating EMI
· In principle Maxwell's equations give a complete description of EMI for any situation. · In practice, both analytic calculations and even computer beam codes are very co ple compex and difficut to use to understand cult to use to de sta a realistic beam line situation. · They can however provide a crude crude understanding of phenomena.
April 7, 2007 Fermilab SiD Workshop 5


Measuring EMI EMI
· In principle it is possible to measure beam induced EMI. · In practice giving a complete characterization is challenging. · The RF frequency range of potential interest covers many orders of magnitude. · Individual antennas have limited range. · Scopes have limited resolution.
April 7, 2007 Fermilab SiD Workshop 6


The tools used at SLAC ESA used at SLAC
· Two RF antennas sensitive in the high MHz to very low GHz range. · An HP digital scope with resolution in the very low GHz range. · A few specially designed narrow band few pass antenna for 30GHz, 100GHz and 300GHz. · An electronics module from the SLD Vertex detector that may have suffered th ff from EMI during the run at SLC.
April 7, 2007 Fermilab SiD Workshop 7


Biconical antenna

April 7, 2007

Fermilab SiD Workshop

8


Yagi antenna

April 7, 2007

Fermilab SiD Workshop

9


RF horn antenna

April 7, 2007

Fermilab SiD Workshop

10


SLD VXD electronics board

April 7, 2007

Fermilab SiD Workshop

11


Beam line EMI emitting "gaps" EMI
· EM fields within the (conducting) beam pp pipe are contained by the small skin depth. · However, any dielectric gap can emit EM radiation out of the beam pipe. · Common "gaps" are camera windows, BPM feedthroughs, toroid gaps, etc.

April 7, 2007

Fermilab SiD Workshop

12


Ceramic gap

April 7, 2007

Fermilab SiD Workshop

13


· The SLAC ESA ILC program consists of approximately nine independent test beam experiments, including the EMI study. · There have been three runs of about 2.5 weeks each. · Beam time has been shared between the different projects. · Unlike the others the EMI project requires frequent accesses to rearrange resources. to rearrange · In the last run EMI had 13 accesses.
April 7, 2007 Fermilab SiD Workshop 14

The ESA beam line


Beam line

April 7, 2007

Fermilab SiD Workshop

15


Observed EMI in ESA in
· Antennas placed near (~1 m) gaps observed pulses of EMI in the high MHz range with strengths up to ~20 V/m. · The pulse shapes are very stable over widely varying beam conditions indicating they are determined by the geometry of the beam line elements. · The pulse amplitudes varied in proportion to the bunch charge but were independent th of the bunch length.
April 7, 2007 Fermilab SiD Workshop 16


Pulse shape (yagi, bicon) shape

April 7, 2007

Fermilab SiD Workshop

17


Data: E vs charge vs charge
· Plot shows linear relation as predicted. ti di

April 7, 2007

Fermilab SiD Workshop

18


VXD electronics results
· One vertex detector readout module from SLD was placed near the beam line. · A "ready" signal was monitored that had been an indicator of momentary board failure during SLC running. · When exposed to sufficient EMI the ready signal dropped.

April 7, 2007

Fermilab SiD Workshop

19


VXD phase lock loop drops lock loop

Top trace: VXD board phase-lock loop signal Other traces: the two EMI antennas. Ti Time offsets are due to cable length differences. bl th diff
April 7, 2007 Fermilab SiD Workshop 20


Was "airborne" EMI the culprit?
· Or was it image current coupling, ground faults, etc? · The rate of failure declined as VXD was rate of failure moved further away from the source and antennas confirmed the EMI amplitude the amplitude declined. · Placing just the VXD board inside an the aluminum foil shielded box stopped the failures. · Covering the gap also stopped failures.
April 7, 2007 Fermilab SiD Workshop 21


VXD failure rate vs EMI strength strength
· The VXD module phase lock loop lost lock Th VXD on about 85% of beam crossing when the module was exposed to ~20 V/m of EMI. · The VXD module lost lock about 5% when exposed to ~1 V/m of EMI. · That was what we thought we learned what from run in May of 2006. However, 2007 run has shown, that signal seen in that antennas is not the signal, causing failure
April 7, 2007 Fermilab SiD Workshop 22


Recent shielding experiments experiments
· A single layer of common 5mil aluminum foil was placed over the ceramic gap and clamped at both ends to provide an image current path. · The signal amplitude was reduced by >x10. · EMI from upstream sources limited the from resolution.
April 7, 2007 Fermilab SiD Workshop 23


Shielding and VXD
· The aluminum foil gap cover stopped VXD failures. · A 1 cm x 1 cm hole in the gap foil cover emitted enough EMI to cause about 50% to cause VXD failure rate at ~1m distance. (With no foil rate would be 100% at this distance.) · There was no failure with a 0.6 cm x 0.6 was with cm cm hole.
April 7, 2007 Fermilab SiD Workshop 24


Antennas signals with gap covered with foil

April 7, 2007

Fermilab SiD Workshop

25


Failure with 1 cm2 hole in foil cm in

April 7, 2007

Fermilab SiD Workshop

26


Hole results interpretations interpretations
· Did the larger hole induce failure because it emitted an overall stronger EMI signal? · Did the larger hole induce failure because it emitted a stronger signal at a lower wavelength? · If the latter is true then the hole dimensions suggests ~30GHz is a critical critical frequency for VXD.
April 7, 2007 Fermilab SiD Workshop 27


What have we learned? we learned?
· Results so far suggest that significant amounts of EMI can be common at short bunch length accelerators. · To observe all EMI signals, very high frequency antennas and registration system required. · Electronics operations can be disrupted by very high frequency signals (not seen with antenna/scope) ith · Such signals can be generated whenever small (~1 cm2) hole is present in the beam pipe. is present · There have been very few instances where EMI was suspected of being a problem for electronics over 30+ of problem electronics years. So EMI sensitivity is the feature of specific electronics design.
April 7, 2007 Fermilab SiD Workshop 28


Next steps
· Which of the two "hole" interpretations is correct? · Explore counter-intuitive observation that partial shielding increased EMI??? · Explore methods of shielding beam line emission sources for use at ILC. · Try to understand why VXD is failing to to avoid a similar problem at ILC.
April 7, 2007 Fermilab SiD Workshop 29