MCP Delay Line (Version 6.2.90.5)
Page 65 of 80
In rare cases it happened that after shipping either the connection cable or the
DLATR6(8)
box underwent damage and the
reason for a discharge can actually be found in the cable or the
DLATR6(8)
box. In order to check this, remove the cables
from the feedthrough and repeat the procedure. No discharge should occur now. Otherwise contact
RoentDek
.
Whenever you make changes to the detector hardware or the cable connections you should repeat this procedure.
5.3
Initial Start-up Procedure
If you apply voltage across the MCP stack for the first time this must be done very slowly and in a very controlled way, also if
you increase the voltage further after operating formerly at lower voltage. Please also refer to the separate manual for MCP
handling.
Even after you have verified that there was no discharge when applying voltages to the detector, you could not yet verify a
possible discharge problem between “MCP front” and “MCP back” contact. Usually the risk for that is very small. Although
there is a minimum chance for amplifier damage by such a connection problem you should from now on operate with
amplifiers active to verify the MCP stack during the startup procedure.
If you want to apply voltages in the
electron detection mode
, connect the MCP front output (#1) of the
FT12(16)-TP
plug to the
ATR19/NDLATR6(8)
channel 1 or 2 and terminate the MCP back output (#2) with an adjustable lemo terminator. In case
of the
DLATR6(8)
use the A1 output to verify the signal from the MCP stack (front side). The switch for channel 2 on the
DLATR6(8)
front panel
should be set to “poti”. This is the safest way for the electronics in the event of a discharge. It is
recommended, however, to start the MCP operation for the first time in
ion detection mode
.
If you want to apply voltages in the
ion detection mode
, connect the MCP back output (#2) of the
FT12(16)-TP
plug to the
ATR19/NDLATR6(8)
channel 1 or 2 and terminate the MCP front output (#1) with an adjustable lemo terminator. In case
of the
DLATR6(8)
use the A2 output to verify the signal from the MCP stack (back side). The switch for channel 1 on the
DLATR6(8)
front panel should be set to “poti”. This is the safest way for the electronics in the event of a discharge.
The
ion detection mode
will generally be easier for the following start-up procedure as in this case the anode potential can be kept
constant and only one potential (“MCP front”) must slowly be increased, when putting operation voltage to the MCP stack.
Therefore this mode is strongly recommended for the following procedure, even if you plan to run the detector in
electron
detection mode
later.
1.
Turn off all ion gauges and any sources emitting a high rate of charged particles (ion pumps?) or high energetic
photons (UV and higher) in the vicinity of the detector. All detector potentials are on ground potential or close to
that.
2.
Set the potentials for the anode (holder, signal wire, reference wire) to the voltages for
ion detection mode
as
recommended in the “operation” sub-section.
Now the voltage across the MCP stack must be increased. For
ion detection mode
you only need to increase “MCP front”,
leaving “MCP back” at ground potential. For
electron detection mode
you should leave “MCP front” at or close to ground
potential while you increase “MCP back” and consequently also the anode potential to proper voltages between the anode
and “MCP back”. In this case also verify that the “MCP front” or “MCP back” potential, whichever you expect to stay near
ground potential really does so while you increase the potential on the other side of the MCP stack (see section 2.4.6). In the
following only the relative voltage across the MCP stack will be referred to and the other potentials must be adjusted in
electron detection mode
to maintain proper operation condition. Please refer to section 2.4.6 about these relative voltage limits.
3.
Increase the voltage across the MCP stack (bias) to 1000V. From now on verify the signals from the MCP on
Outputs A1 or A2 (monitor). Any real MCP signals (i.e. some “dark” counts) will not appear before 1500V MCP
bias. In the following you will increase the MCP bias in steps of 100V about every 10 minutes. If you see other
signals on the oscilloscope than electronic noise, or if you experience a discharge then reduce the bias to zero and
start again. In case of discharge verify that the amplifier still gives the “normal” noise level, otherwise it could have
been damaged (contact
RoentDek
). If the discharge problem persists or if a high rate of dark counts appears you
have some connection problem inside or the MCP stack is damaged. Check if you can rule out the presence of any
real particle sources (note, that a sharp tip like on a damaged mesh can emit electrons in the presence of a strong
field). If you can not find the reason for the malfunction please contact
RoentDek
.
4.
At a certain MCP bias (around 2000V) you should see a few dark counts (usually less than 50/s for the
DLD40
and
< 200/s for the DLD80). MCP signals from particle counts or dark counts at the An (monitor) outputs are fast
negative signals of a few ns rise time. They have amplitudes between noise level and about 400mV (on 50
Ω
input).
Before you reach 2500V MCP bias you should stop the procedure.
Содержание DLD120
Страница 6: ...Page 6 of 80 MCP Delay Line Version 6 2 90 5 ...
Страница 32: ...Page 32 of 80 MCP Delay Line Version 6 2 90 5 ...
Страница 51: ...MCP Delay Line Version 6 2 90 5 Page 51 of 80 Figure 4 4 BIASET2 Crate with four HV1 4 ...
Страница 68: ...Page 68 of 80 MCP Delay Line Version 6 2 90 5 ...
Страница 78: ...Page 78 of 80 MCP Delay Line Version 6 2 90 5 ...