Transcript For LHCb the induction gap is only 1mm

Chamber parameters that we can modify and
that affect the rising time
Intrinsic Resolution:
Number of ionisation clusters produced in the drift gap: Poisson Distribution
Probability distribution of the distance x of the ionisation cluster NEAREST to the first GEM :
With:
P (x )=n ・e( - nx )
Space Probability Distribution of the first GEM NEAREST Cluster
s(x)=1 /n
where n is the average number of ionisation clusters per unit length
Time Probability Distribution of the first GEM NEAREST Cluster
s(t)= 1 /( n ・v
drif t
)
Single cluster/electron Signal Formation:
Each electron produce a current pulse for a time
t = dinduction gap/ vdrift
Parameter that we can change:
vdrift Induction Field - Gas
n
Gas
Drift Velocity vs Electric Field in Ar/CO2 70/30
magboltz
In the actual configuration we have
1M+0.55M+1M+0.5M+1M+0.45M+1M
D~2.4kV/cm
T1~3.6kV/cm
T2~3.6kV/cm
Ar-CO2 (70-30)
I~3.6kV/cm
If we apply -4kV, it means that we will have 4kV/5.5M~727mA
In the induction region we will have 727mA x 1M=727V across 2mm, i.e. ~3.6kV/cm
In the plot we can find that at ~3.6kV/cm we have ~7cm/ms.
If we decided to use 1.5M, it means that we will have ~5.5kV/cm
With this field the drift velocity will be ~8cm/ms.
Changing only the induction field:
effects on the signal induced
Each electrons cloud produced by a cluster will produce a current pulse for a time
t = dinduction gap/ vdrift
The current induced will be approximately
I ~~ (N(ionization)x Gem GAIN x e) x (vdrift/ dinduction gap)
This mean that we will have a factor 1.15 (in the current induced by each clusters), if
we change the drift velocity from 7cm/ms to 8cm/ms
Changing the gas
Two effects
Changing the gas: Effects on the GAIN
Changing the gas: effects on the signal induced
Each electrons cloud produced by a cluster will produce a current pulse for a time
t = dinduction gap/ vdrift
The current induced will be approximately
I ~ (N(ionization)x Gem GAIN x e) x (vdrift/ dinduction gap)
This mean that, if we change the drift velocity adding CF4 to the gas mixture, we
will have a factor 1.5/1.3 (in intensity of the signal induced by each clusters).
One Example: LHCb GEMs
A factor of ~1.5/2 is possible respect with Ar/CO2
For LHCb the induction gap is only 1mm
Efficiency in 20ns time window
H.V. Board Resistors Modification (in Ar/CO2).
If we want to maintain the same fields in the other gaps and the same DV across the GEM(*)
We will need:
DV(ind) = 5kV/cm x 0.2cm = 1kV
R(ind) =
1kV / 728mA
~ 1.38MW [728mA in order to have DVG1+DVG2+DVG3 = 1090V as now]
Resistors Std value:
field[kV/cm]
Power [W]
DV[V] respect with 1M
@ 1mA(**)
@ 0.728mA
1.2M
~4.4
1.2
~150V
1.5M
~5.5
1.5
~370V
(*) if we change the gas we need to do also other consideration about the gain
(**) this is the maximum current from the module (to be extremely conservative).
H.V. Board Resistors Modification (in Ar/CO2).
Power Supply Module: -5kV, 1mA (maximum possible power 5W)
HV Divider : (let me say) 200k(filter)+5.5M(actual version)+DR(Induction Gap)
If I consider in total 6.2M (i.e. with 1.5M on induction), the max current available will be ~800mA
It means that respect with 728uA we could eventually increase the voltage of ~400V in total and 100V
on GEM.
In terms of gain this is ~~ a factor 5-10.
If I consider in total 5.9M (i.e. with 1.2M on induction), the max current available will be ~845mA
It means that respect with 728uA we could eventually increase the voltage of ~700V in total and ~175V
on GEM.
In terms of gain this is ~~ a factor 30-60.
TripleGEM timing properties from H8 test Beam:
MSPL Considerations…
H.V. on TripleGEM ~ -4.1kV
The needed MSPL with the actual configuration
The first MSPL that guarantees that all the signals are collected for a time period
bigger than 1 clk cycle is 4CLK . We need “for a time period bigger than 1 clk
cycle” because we can change the latency in steps of 1clk.
Threshold Scan: Acquisition Mode (40k triggers/point)
MSPL:2clk and 4clk
No differences. OK
Timing Studies:
•
•
•
•
•
MSPL tests revealed the possibility of using more clk cycles without degrading the
level of the noise.
To increase the drift velocity we can replace the induction gap resistor in the H.V.
divider. A good value could be 1.2/1.5M.
In order to have a bigger improvement of the rising time we could use Ar/CO2/CF4
that increase not only the drift velocity but also the number of primary clusters in
the drift gap (more effective than changing only the induction field).(We have to
be careful because using CF4, the gain of each GEM foil will be reduced.)
In order to have margin in the increase of the gain, taking into account the HV
power supply output range, a value of 1.2/1.3M for the induction gap resistors could
be a good compromise between the increase of the drift velocity and the possibility
to increase the voltage across the GEM foil if we will change the gas.
For the system set-up in H8:
– relief available
– we will check the additive connectors needed and we will have to install them.
– we will have to clean the rotameter in order to have all the gas lines needed to
flux 10 chambers.
– we will have to check the functionality of the mass flow-meter for the mixing
unit .