Transcript Diapositiva 1

DT GROUNDING & SHIELDING
Presented by A. BENVENUTI
M.PEGORARO
Grounding Workshop 24th Jan 08
Grounding & Shielding
Superlayers & Chambers
•
Analog FE Electronics is inside gas volume of each Superlayer (that is actually a shielding case).
Signal connections (data & control) go to minicrate onboard the chamber.
Power (5V & 2.5V) for Analog FE arrives to the Splitter Board whose copper heatsink is strongly
connected to Superlayers’ enclosures on FE side. The heatsink acts also as common point for the
independent returns of supplies.
•
Most of ReadOut & Trigger Electronics are housed in the Minicrate.
Power (5V & 3.3V) arrives independently from FE one and the common point for returns is the
Al case tightly connected to the Superlayers’s enclosures and to the heatsink of the splitter board.
•
On the other side of Superlayers there are HV boards inside gas volume with GND connection
provided by inner Al plates of the layers. 2 HV cables per Superlayer carry voltages and returns;
these latter are connected to inner Al plates and decoupled using resistors.
•
Low & High Voltages are obviously filtered at chamber input wrt common GND and with
distributed capacitance where loads are applied. All internal connections of the chamber are
done through shielded cables except for short ending lengths.
M.PEGORARO
Grounding Workshop 24th Jan 08
Inside Superlayer
shield of HV cable
Al Plates
HV ref
M.PEGORARO
Grounding Workshop 24th Jan 08
Input node
Sig ref
LV ref
Inside Chamber
to HV
Junction
Boxes
HONEYCOMB
Vcc & Vdd
NO RETURN
Splitter
Board
Heatsink
Copper
Braid to
Magnet
Iron
FE LV
Cable
Splitter
Board
MINI
CRATE
Supporting
Al frame
M.PEGORARO
Grounding Workshop 24th Jan 08
Grounding & Shielding
Cables & Braids
•
The enclosures of Superlayers are connected each other and to supporting frame on enclosures
with copper braids. Every chamber is a case electrically insulated from magnet iron with supply
returns & GND concentrated on the heatsink of the Splitter Board.
•
The Reference Ground (Splitter Board) of the Detector is also Safety Earth; this has no
consequence on LVDS signals from Frontend that end inside the Minicrate.
•
A short (600 mm) copper braid connects this point to magnet iron.
RPC detector is close, but insulated, to external Superlayers, and is connected to the same point.
•
All cables to/from chamber are shielded with shields connected on detector side. HV and
Minicrate LV cables have junctions inside metallic boxes located on iron where cable shields can
be decoupled.
•
Connection with DAQ and TRG acquisition on tower racks is done via LVDS AC coupled;
DCS communication is done through fibers or floating copper.
•
All of above cables go to Tower Racks at nearest floor (all floors are concerned, with X2 near
housing also Acquisition & Trigger crates)
M.PEGORARO
Grounding Workshop 24th Jan 08
Grounding & Shielding
Tower Racks
•
HV crates in X3 & X4 racks contain floating Voltage Regulators; Voltage Generators are in USC
racks and feed (and control) these crates through long cables. So reference and shielding must be
shifted from Earth of USC racks at the input to detector GND at the output and we have to take
care of common mode noise.
•
Trigger and Acquisition crates in X2 near receive Low Voltage Differential Signals with input
compliance of about 1V at high frequencies. Communication with USC is made through fiber optic
and they receive power supply from nearby Easy LV crate that feeds also chambers: they are
practically insulated.
•
EASY LV system is on X2 and X4 racks and includes AC/DC converters (MAO) and DC/DC
converters in cascade: the latters are insulated (with certain limits) from the case while the
formers have Safety Earth, and thus chassis, connected to Earth coming from USC S4F zone;
moreover their 48V output is also referred to case and so to USC.
•
As a result of not ideal isolation of each DC/DC module from case and rack and direct connection
of USC safety Earth to Tower Racks (via MAO’s case) there could be some interference to delicate
items like HV and LVDS common mode at the input of acquisition modules; also LV could be
affected. We have to consider that Tower Racks contain electronics from several detectors.
•
Also some tests have shown that shielding of cables is of the maximum importance to prevent noise
entering that way (the welding machine was an example confirmed by further tests)
M.PEGORARO
Grounding Workshop 24th Jan 08
Grounding & Shielding
Conclusions (?)
•
We have to prevent as much as possible problems that could arise when the whole detector,
including magnet and all power supplies, will be on.
•
One uncertainty is given by Safety Earth that comes from USC S4F to Tower Racks: we ask for
some kind of decoupling at high frequencies.
•
The other idea is to have a strong connection from Magnet Iron to Tower Racks at all floors; it
should be particularly effective for high frequencies. In this way, connecting to it modules in Tower
Racks, we could override the interferences due to the shift of GND reference from USC to detector.
The idea is to have a short copper braid with large surface to reduce radio frequency impedance
going from magnet iron to a copper panel placed on bottom of each Tower Rack.
Also the input reference of DAQ & Trigger modules can use this low impedance connection to iron.
•
Then we could use the strong ground from Magnet Iron to reinforce and test different options for
connection of cable shields and Low & High Voltage modules.
M.PEGORARO
Grounding Workshop 24th Jan 08