Diapositiva 1

1
The role of the Resistive Plate response
function in bringing an RPC to a stationary
situation
M.A. Fernandez-Morales, J.A. Garzón, D.
González-Díaz
2
RD in high rate / large area tRPCs.
lines obtained with the functional dependence
also used in
D. Gonzalez-Diaz et al., Nuclear Physics B
(Proc. Suppl.) 158(2006)111
3
RD in high rate / large area tRPCs (DC model
scaling).
.
DC model
(CBM goal)
4
The DC model works reasonably well when the RPC
reached a stationary situation. What happens
before?
5
Transient behaviour in RPCs (I).
(Asymmetric 8 mm RPC)
E. Cerron Zeballos et al., NIM A
367(1995)388, also I. Crotty et al. NIM A
337(1994)370
6
Transient behaviour in RPCs (II).
0.66 mm-4 gap RPC
P. Colrain, E. Polycarpo et al., NIM 456(2000)62
7
Transient behaviour in RPCs (III).
0.3 mm-4 gap RPC
D. Gonzalez-Diaz et al., Nuclear Physics B (Proc.
Suppl.) 158(2006)111
8
Transient behaviour in RPCs (IV)
M. M. Fraga et al., NIM 419(1998)485
transient time depends on rate!
9
What did we do?
10
Experimental setup
RPC (3.5x3.5 cm2) (standard timing mixture)
X-ray tube
mylar window
Cu electrode
Cu absorbers
glass (1.8mm)
gap (0.3 mm)
goniometer (0.1 deg precission)
11
transients of the current in the circuit
12
Part I. Characterization of the electrical
properties of glass
13
Measurement of the glass electric properties
guard ring
The response function of glass f(t) can be
expressed as
Fisica de dielectricos, J.M. Albella. J.M.
Martinez, Barcelona 1984
14
Some basic relations (I)

• being
• eS the permittivity of the material when it is
  'relaxed' (t-gt8), or 'static'.
• e8 the permittivity of the material at high
  frequency (short times), before relaxation takes
  place.

Most materials can be described as RC circuits
with C-gtC(?)
R
ImC(?)
ReC(?)
15
Some basic relations(II)
Some ways of describing the relaxation process in
soda-lime/window glass
M. M. Fraga et al., NIM 419(1998)485
1.
(simple RC circuit)
2.
3.
Distribution of relaxation times
dN/dtsech(1/2t)dt
Taylor, H.E., J. Soc. Glass. Tech. 41(1957)350.

• Experimental procedure in order to rule out
  electrode-related effects we followed H.E. Taylor
  systematic approach, and so different
  experimental configurations were tested
• Different electrode materials (Au, Ag, Cu).
• Different electrode technologies (tape, polymer,
  acetate, evaporation).
• Different electrode contact (welded, by presion).
• With and without guard ring.

Only the consistent data is shown in the
following !
16
Dependence of the response function on voltage (I)
17
Dependence of the response function on voltage
(II)
T27 oC
For voltages up to 1 kV (higher voltage drops
have little practical implication in the tRPC
response) we conclude that the window glass of
the kind used here is highly ohmic, and its
response function is fairly independent on
voltage.
18
Part II. Characterization of the RPC behaviour
at high rates.
19
Experimental setup
RPC (3.5x3.5 cm2) (standard timing mixture)
X-ray tube
mylar window
Cu electrode
Cu absorbers
glass (1.8mm)
gap (0.3 mm)
goniometer (0.1 deg precision)
20
X-ray tube characteristics
energy determined with TeCd detector
Energy spectrum for HV 50 kV (maximum)
ltEgt12 keV
The X-ray response can be considered as
instantaneous at the time scale we are interested
in determined through the current response of a
diode operated in reverse bias voltage.
21
Experimental setup. Orientation respect to the
X-rays direction of incidence (I)
Parameters Pglass, Pgas
AilluminatedA
AilluminatedA/2
AilluminatedA
22
Experimental setup. Orientation respect to the
X-rays direction of incidence (II)
working point
region B
region A
23
MC description. Modeling the RPC behaviour

• -gt At low rates, the dependence of the average
  total charge with the applied field in the
  space-charge regime is well described by a
  straight line

G. Carboni et al. NIM A 498(2003)135 2 mm RPC
G. Aielli et al. NIM A 508(2001)6 2 mm RPC
D. González-Díaz et al. NIM A 555(2005)72 0. 3
mm RPC
At high rates, and once the DC/stationary
situation is reached, the total charge is
affected by the field drop caused in the glass
itself
R resistance F flux Hz/cm2 A
illuminated RPC area r F A rate Hz
24
MC description. Modeling the electric response (I)
The area of influence of an avalanche is
approximated by Acell
M. Abbrescia NIM A 533(2004)7
In the MC, the situation is described by analogy
with a random shot noise generator in parallel
with the gap capacitance.
D. Gonzalez-Diaz et al., Nuclear Physics B (Proc.
Suppl.) 158(2006)111
The current measurable in the external circuit is
given by
Fisica de dielectricos, J.M. Albella, J.M.
Martinez, Barcelona 1984
where the first term gives the current induced
instantaneously by each new avalanche and the
second is the current induced due to the
instantaneous voltage across the glass.
25
MC description. Modeling the electric response
(II)
C8 denotes the RPC capacitance at high
frequencies (?1-109 GHz), readily accessible
from the C8 of glass with an LCR analyzer (here
er 7-8) and CS is the RPC capacitance at low
frequencies, which is more difficult to access
and is the one actually responsible for the
relaxation process. Going from a glass response
of the type f(t)exp(-t/t)ß to the RPC response
function has sizeable technical difficulties,
since non-analytical Fourier Transforms are
involved. The following method will be used
instead the response function is assumed to be a
simple exponential with tRCs, and Cs is left as
a free parameter
The first term is a bit complicated to compute
numerically but, as far as average values are
concerned, it can be replaced by
26
Consistency criteria of the MC
Under the assumption of linearity between total
charge and voltage, the following relations are
fulfilled in the DC limit
(therefore, in the DC limit i ?V/R)
27
Constraining aF and Vth parameters in the DC limit
28
Results from the MC
?
?
?
?
29
MC and data after fitting (only CS is a free
parameter)
30
MC and data after fitting (only CS is a free
parameter). Zoom
31
Expected behaviour of the total charge from MC
32
Transients in current and charge. MC and data
stabilization time needed for charge is totally
different when starting and stopping illumination
33
Time scale of transients
teq,I(?V) equilibration time time at 65 of
the DC value of I (?V)
stabilization of the current in the external
circuit does not mean stabilization of the RPC!
34
Conclusions

• A MC devised to describe transient effects in
  RPCs based on a previous code (D. Gonzalez-Diaz,
  Nucl. Phys. B(Proc. Supl.), 158(2006)111) has
  been introduced showing a reasonable description
  of data.
• A simple RC circuit was assumed, because of the
  much more simplified situation, but it was
  illustrated how a detailed comparison would
  require indeed to introduce the response function
  through the generalized R(?)C(?) circuit.
• It has been shown that Cu tape electrodes and
  both Au, Ag evaporated electrodes provide a
  consistent determination on the glass response
  function, being highly independent on the voltage
  up to 1kV. Ag acetate and polymer provided
  discrepancies of a factor 2, and visible
  instabilities in the polymer case.
• It was shown how the transient times for current
  stabilization are different from the transient
  times for charge stabilization (driven by the
  stabilization of the voltage across the glass).
• In particular it is shown how, naturally, the
  stabilization time (both for current and charge)
  depends mainly on the primary rate whenever
  irradiation starts, and depends mainly on the
  relaxation time whenever it stops.

35
Appendix
36
Loosely relevant parameters
The shape of the charge distribution does not
seem to have any influence (chequed with a
Poissonian distribution). The influence of the
Acell is also small (here Acell 0.3 mm2) as
long as it does not result in a single shot of
magnitude comparable to the maximum voltage drop
?VV-Vth
This is also consistent with the simple
derivation of the equilibration time from
D. Gonzalez-Diaz et al., Nuclear Physics B (Proc.
Suppl.) 158(2006)111
37
Determination of the flux over the RPC
extrapolation to high fluxes
FRPC observed flux with RPC
Fdosimiter observed flux determined with a
comercial dosimiter.
saturation due to the voltage drop in the glass
Fdosimiter Hz/cm2
checked that the proportionality between tube
current and the primary flux exists
IX-ray tube nA
38
Fit to the absorption curves