Spark-protected high-rate...P.Fonte CERN 1998 a1 Spark-protected high-rate parallel geometry gas chambers P.Fonte Laboratório de Instrumentação e Física Experimental de Partículas (LIP) e Instituto Superior de Engenharia de Coimbra (ISEC) Coimbra, Portugal CERN 1998

Spark-protected high-rate...P.Fonte CERN 1998 a2 Summary Brief survey of the basic parallel plate detector types, physics and operation modes. Hybrid wire mesh/resistive plate detector, made with a medium resistivity plate for improved rate capability and spark protection. Rate-gain limitations on parallel geometry chambers and other gaseous detectors. Thin gap parallel mesh chambers.

Spark-protected high-rate...P.Fonte CERN 1998 a3 Basic parallel plate detector types metal plate amplification gap metal plate Parallel Plate Chamber (PPC) (J.S.Townsend, 1900) resistive plate amplification gap resistive plate Resistive Plate Chamber (RPC) (Pestov, 1978) wire mesh amplification gap wire mesh Parallel Plate Avalanche Chamber (PPAC) (G.Charpak and F.Sauli, 1978) Hybrid types amplification gap wire mesh cathode resistive plate anode A subject of this talk amplification gap wire mesh cathode metal plate anode Another subject of this talk metal plate amplification gap resistive plate Not very popular etc.....

Spark-protected high-rate...P.Fonte CERN 1998 a4 PPC operation modes anode cathode ionizing particle high-voltage pulse spark Spark Chamber (optical or electric read out) anode cathode ionizing particle narrow high-voltage pulse streamers Streamer Chamber (optical read out) anode cathode ionizing particle constant voltage Proportional mode (electric readout) avalanche* * eventually there will be also a few sparks

Spark-protected high-rate...P.Fonte CERN 1998 a5 RPC operation modes ionizing particle streamer Streamer mode (development of a spark is avoided by the current limitation provided by the resistive electrodes) higher constant voltage resistive anode resistive cathode resistive anode resistive cathode ionizing particle Limited proportional mode (sub-exponential gain due space-charge effect in single avalanches) avalanche constant voltage PPAC operation modes Proportional mode only*, but a large number of configurations can be formed. A drift gap provides good energy resolution and efficiency for MIPs drift gap pre-amplification gap ionizing particle Photon transfer gap amplification gap transfer gap read out electrode gate electrode Example of a multistep PPAC * plus a few occasional sparks

Spark-protected high-rate...P.Fonte CERN 1998 a6 Basic physics Electron (fast) signalIon (slow) signal Charge in slow signal N i = N t (1-1/ln(G)) in practice about 90% of Nt Charge in slow signal N i = N t (1-1/ln(G)) in practice about 90% of Nt Charge in fast signal N e = N t /ln(G) in practice 7% to 14% of N t 50ns 10 to 30  s Gain: G(x)=exp(  x);  = First Townsend coefficient Total collected charge: N t = N 0· G(x) cathode anode Primary electron created at distance x from the anode Avalanche multiplication x amplifying gap Primary electron cloud x Ionization track Ionizing particle N 0 = number of primary electrons Photon Total collected charge: N t = N 0 /  d · G(d) d The electrons closer to the cathode generate most of the final charge  localization of the entry point If N 0 =50, then the equivalent number of concentrated primary electrons will be only  50  0.07 = 3.5

Spark-protected high-rate...P.Fonte CERN 1998 a7 From avalanche to streamer Slow breakdown modeFast breakdown mode Slow breakdown depends on photon feedback to the cathode: “generations” mechanism. Fast breakdown is a local process: “streamer” mechanism. When the amount of hydrocarbons (quencher) is increased there is always a transition from the slow to the fast breakdown mode. No matter the gain or the gas composition there is a hard limit on the total avalanche charge of a few times 10 8 electrons.  Mixtures with TEA Mixtures with CH 4 and C 2 H 6

Spark-protected high-rate...P.Fonte CERN 1998 a8 Streamer theory The streamer process can be separated in three stages - proportional avalanche stage - avalanche-streamer transition stage - streamer development stage - spark No SQS mode

Spark-protected high-rate...P.Fonte CERN 1998 a9 Proportional avalanche stage Avalanche-streamer transition stage

Spark-protected high-rate...P.Fonte CERN 1998 a10 Streamer development stage

Spark-protected high-rate...P.Fonte CERN 1998 a11 From streamer to spark Avalanche/streamer Glow formation Diffuse glow Filamentary glow Spark (S.C.Haydon, 8th Int. Conf. on Phen. in Ion. Gases, Vienna, 1967)

Spark-protected high-rate...P.Fonte CERN 1998 a12 Motivation for an hybrid PPAC/RPC detector PPAC High counting rate (up to 10 5 /mm 2 ) Violent sparks Versatile (drift, multistep, etc..) No dark noise Proportional mode only Almost not used PPAC + RPC Fast signal (tens of ns) Large areas Breakdown via streamers at a few times 10 8 electrons/avalanche Good position resolution (100  m) in 2 dimensions (prop.mode). Good timing (less than 1 ns  ). RPC Low counting rate (up to a few times 10/mm 2 ) Mild sparks / “indestructible” Less versatile (MIPs only) Dark noise Streamer mode or proportional mode Widely used Hi-rate RPC High counting rate Mild sparks / “indestructible” Versatile (MIPs, X-rays) No dark noise Proportional mode Good timing Widely used (wish list) PPACRPC 18 orders of magnitude in electrode resistivity Explore the electrode resistivity parameter High-rate PPAC/RPC

Spark-protected high-rate...P.Fonte CERN 1998 a13 How fast can a PPAC count and at what gain? Pulse-height doesn’t depend on rate Maximum counting rate is determined by the appearance of sparks For low gains the maximum counting rate Rmax and the gain G=Q/(eN 0 ) are related by RmaxQ/D=C were C is a constant with a value around 100 nA/mm or electrons/(s mm). There is a linear dependence on the beam diameter and not on the beam cross-section! High-rate PPAC/RPC

Spark-protected high-rate...P.Fonte CERN 1998 a14 A new material is needed if  =  cm  10 Hz/mm 2 then  = 10 7  cm  10 5 Hz/mm 2 Our solution: epoxy+ink black soft rubber (not staining)  = to 2  10 7  cm dielectric strength > 16 kV/mm Controllable resistivity. Ohmic behavior Mechanically inconvenient. Too soft and the surface resistivity is strongly affected by dryness. A much more convenient material, based on ABS plastic, is now being investigated in collaboration with a specialized group. Bulk resistivity High-rate PPAC/RPC

Spark-protected high-rate...P.Fonte CERN 1998 a15 Setup 15 mm 3.5 mm Drift Amplification 3  300 pF To scope Resistive plate over a metal base Wire meshes Sharp focus X-ray gen. 40 mm Collimator To current amp Gas: Ar + 10 to 20% C 2 H % of methanol V.P. Since C gap « C plate « C readout the voltage change across C gap is mainly determined by the signal charge stored in C plate 3  C gap C plate C readout Signal current R plate Equivalent electrical circuit Mechanical arrangement High-rate PPAC/RPC

Spark-protected high-rate...P.Fonte CERN 1998 a16 Low-rate behavior  = 3  10 8  cm L = 1.5 mm High-rate PPAC/RPC

Spark-protected high-rate...P.Fonte CERN 1998 a17 Counting rate capabilities Ohmic model Reasonable agreement, considering that the model doesn’t take into account beam-edge effects, material non-linearities, etc... High-rate PPAC/RPC

Spark-protected high-rate...P.Fonte CERN 1998 a18 Streamer charge 50 mV/div 250 ns/div 20 mV/div 10  s/div 50 mV/div 250 ns/div 50 mV/div 250 ns/div 50 mV/div 10  s/div 50 mV/div 10  s/div Streamer current on 3  (not all pictures on same gas)  = 4   cm L = 0.25 mm (melamine) Q meas.  33 nC Q stored  40 nC/cm 2  = 3  10 8  cm L = 1 mm Q meas.  28 nC Q stored  7 nC/cm 2  = 4  10 7  cm L = 1.5 mm Q meas.  21 nC Q stored  11 nC/cm 2 There is some contribution from conduction current across the plate, but it is comparable to the discharge of the plate-equivalent capacitor. High-rate PPAC/RPC

Spark-protected high-rate...P.Fonte CERN 1998 a19 Summary of high-rate PPAC/RPC The detector can operate in proportional mode up to the intrinsic counting rate limits of metallic PPACs (about 10 5 Hz/mm 2 with gain above 10 4 ). Materials with suitable mechanical properties are needed with resistivity about 10 7  cm. The streamer has a charge typically of the order of 10 nC (relatively independent of the substrate resistivity) and poses no threat to the integrity of the detector. Beam area seems to have only a minor effect on rate capability. High-rate PPAC/RPC

Spark-protected high-rate...P.Fonte CERN 1998 a s Detector current (on 1 M  ) Slow current increase just before breakdown Cyclic breakdown, with frequency dependent on detector current. At higher gain there is continuous sparking. This kind of continuous sparking is totally absent in low-rate sparking. 1  s The breakdown pulse is preceded by many individual spurious avalanches at a growing rate and amplitude We call this the “cathode excitation” effect (cannot be photon feedback). May be improved by choice of cathode materials, geometry or gases. Why there is rate-induced breakdown in PPACs? Rate-induced breakdown Aftercurrent

Spark-protected high-rate...P.Fonte CERN 1998 a21 What about other detectors? Rate-induced breakdown Data presented by V.Peskov at the Wien WCC98 It seems that there is a general tendency for a reduction in the maximum gain as the counting rate increases. Unknown physical origin.

Spark-protected high-rate...P.Fonte CERN 1998 a22 How can we improve gain x rate (at least in PPACs)? Rate-induced breakdown If we succeed this will clarify the physical nature of the rate-induced breakdown process. The result may be useful also for other detectors. We measured the maximum current in a 3 mm gap PPAC for the following combinations of parameters Anode or cathode copper plate or mesh 150  m or 50  m mesh wires Check if the effect depends on the ion density over the cathode surface Oxidized or clean copper High (ethane) or low (methanol) ionization potential ions in gas Presence or absence of a drift gap The current ranges from 100 to 200 nA (3 mm 2 beam) independently of any of these parameters but… it jumps to 1  A if the gap width is reduced to 0.6 mm!

Spark-protected high-rate...P.Fonte CERN 1998 a23 Setup Single gap 10 mm 0.6 mm Drift Thin amp. gap 300 pF Slow current amp Metal plate Wire meshes Sharp focus X-ray gen. 100 mm Collimator (2 mm diam.=3.1 mm 2 ) Fast current amp 5 M  -HV Pre-amplified thin gap 10 mm 0.6 mm Drift Thin amp. gap 300 pF Slow current amp Metal plate Wire meshes Sharp focus X-ray gen. 100 mm Collimator (2 mm diam.=3.1 mm 2 ) Fast current amp 5 M  -HV 2 mm Transfer 2 mm Thick preamp. gap Electronics Slow current amp. (DC coupled): sensitivity 1 or 10 V/  A averaging time 5 ms Fast current amp. (AC coupled): sensitivity 0.1 V/  A averaging time 200 ns Thin gap PPAC

Spark-protected high-rate...P.Fonte CERN 1998 a24 Gain calibration Ion pulse shape The gap current was measured by the slow current amp., calibrated against a Keithley 414s picoammeter. 200 nA 1  s Q ion  1.6 pC Q electrons  Q ion /(ln(G)-1)  Q ion /10  160 fC v electrons  5 cm/  s Pulse width electrons  gap/v electrons  12 ns Pulse heigth electrons  (Q/ Pw) electrons  13  A G  Q ion / (e N 0 )   Gap field 30 kV/cm Thin gap PPAC

Spark-protected high-rate...P.Fonte CERN 1998 a25 Rate-gain capabilities Mesh pitch = 500  m Gap thickness = 600  m Thin gap PPAC * V.Peskov, Wien WCC, 1998

Spark-protected high-rate...P.Fonte CERN 1998 a26 From the point of view of current Thin gap PPAC

Spark-protected high-rate...P.Fonte CERN 1998 a27 Discharges 1  A 20 ms Q discharge  20 nC but in the readout capacitor there were Q C  1.5 kV  300 pF = 450 nC For some (yet unknown) reason in thin gaps full sparks don’t develop and the discharge is self-limited! 0.5  A 500 ms Rate-induced spark in a thick gap chamber using the same amplifier as the previous one Thin gap PPAC

Spark-protected high-rate...P.Fonte CERN 1998 a28 Detector “B” Detector “C” Detector “D” (preamp) Energy resolution The mesh pitch is comparable to the gap width causing the field in the gap to be non uniform May be improved by a finer mesh 5.9 keV X-rays 5.9 keV X-rays 5.9 keV X-rays Thin gap PPAC

Spark-protected high-rate...P.Fonte CERN 1998 a29 Position resolution From the literature on thick-gap PPACs and RPCs: about 100  m. Quadratic contributions to the distribution beam width:100  m beam divergence: 20  m electronics noise: 22  m detector: 48  m Our own measurements using a 9 mm gap RPC and a collimated X-ray beam Timing accuracy From the literature on PPCs and RPCs: better than 1 ns . (Arefiev et al.) Thin gap PPAC

Spark-protected high-rate...P.Fonte CERN 1998 a30 Thin gap parallel mesh chamber summary Negative points  Mesh pitch comparable with the gap width Bad energy resolution (50%) when compared with the best resolution of PPACs (14%). Intrinsic granularity May be solved (or not) by using a finer mesh Probably affects negatively the timing accuracy  Detector physics not fully understood (but quite a lot is know about parallel geometry chambers)  Discharges are not totally avoided. Thin gap PPAC

Spark-protected high-rate...P.Fonte CERN 1998 a31 Thin gap PPAC summary Positive points  Large current capability (essentially gain  rate)  Made with standard stainless steel wire mesh Mechanically robust No melting Can be strongly stretched to achieve good parallelism Free of defects (spikes, etc..)  Large maximum (low-rate) gain 10 7 Hz/mm gain Hz/mm gain 10 4  Very mild discharges + strong electrodes  virtually indestructible  “Macro”-technology  cheap and easy to build  Free of dielectrics  no  charging-up effects  Good timing and position resolution expected (from experience with thick gap PPCs and RPCs) Thin gap PPAC