A spontaneous discharge or a significantly higher dark count rate indicates a problem such as a glow discharge or the presence
of charged particles/photons triggering the MCP. If this rate is excessive the MCP can be damaged. In such an event turn off
the high voltage and verify your setup again. Note that it may occur that an excessive load of particles on the MCP detector is
not detected by verifying the MCP signal because the MCP stack reaches current saturation before producing individual pulses
of detectable signal height (i.e. above noise level). Such MCP saturation can usually be recognized from non-linear bias/current
characteristics. Therefore we recommend using high voltage supplies with current reading, as available from
RoentDek.
Note, that some high voltage supplies may also increase the noise level, especially at high bias.
It is recommended to initially operate the detector in the so-called “ion mode” as used for detecting slow (and light) positive
ions or fast (neutral) particles and photons, having MCP back at zero voltage to ground and the anode bias on few hundred
Volts positive potential. At such potential discharge events on the “rear part” of the detector are virtually excluded. Only the
MCP front side is at high negative potential and the risk for discharge (malfunction) from preparation mistakes is lowest.
Remaining hazards are too-close metal parts in front of the MCP front face or at the cable contacts: nearby cable contacts
should either be at least 3mm away or biased with the same potential for avoiding “sparks” (spontaneous discharges) and glow
discharges which can occur from pointy parts already at relative potential well below the critical limit for spontaneous discharge.
Typical voltage settings are
Reference wires
(respectively timing anode)
Collecting (Signal) wires
Table 2.3: Typical detector voltage settings (chevron sets of 60:1 MCP). The (delay-line anode) “holder” potential
needs to be carefully adjusted later to achieve optimal linear imaging performance at outer detection diameters
While increasing MCP front bias in ion mode with negative polarity in steps of 100V every few minutes the MCP current should
be recorded. If you use a high voltage supply from
RoentDek turn on the “kill” option so that the voltage is turned off in
case of unexpected current peaking (i.e. a discharge, see also the respective manual). It is also recommended to reduce the
maximum current limit to the lowest setting just above the expected default current, usually <0.3mA (other high voltage supplies
may have similar safety features which should be engaged). It is recommended disconnecting the signal cables leading to
amplifiers or other sensitive equipment (like the oscilloscope) until you have reached about 70%- 80% of the default MCP bias
(as recommended for normal operation, see below). Otherwise there is a risk of damaging follow-up electronics in case of an
unexpected discharge event.
Make sure that the MCP back side remains on or near ground potential while increasing the potential on MCP front. If you use
high voltage power supplies from
RoentDek to bias MCP back, switch off this channel (like most high voltage supplies it
may otherwise be “drawn away”, see below). Or connect a
RoentDek HVT or SHV-G (ground plug) to the MCP back high
voltage input on the decoupler instead.
As you increase the MCP front voltage calculate the MCP stack resistance from the current reading for each voltage step.
The MCP stack resistance should stay constant as the voltage is increased.
Typical MCP stack resistance values are between 10MΩ and several 100MΩ. You may have received info on the default MCP
resistance to compare it with your reading
*
. Deviation from strict MCP resistance constancy can arise from temperature effects:
For a 1°K increase in temperature the resistance drops by about 1%. Note, that the MCP stack can be heated up from the strip
current, i.e. the resistance at high current (voltage) may be lower than at low current (voltage). In-line decoupler resistors or an
(undetected) minor “voltage pull-up” of the high voltage supply for MCP back may also lead to small deviations between
measured and real MCP resistance. If you use a
RoentDek HVZ module at this point there will apparently be a strong non-
linear response below 300V MCP bias. In this case please refer to the respective manual.
*
Note that specific resistance values given by the MCP producers may not be accurate, systematic deviations by a factor of 2
can occur. However, MCPs specified to be of same resistance will always form a matched stack.
MCP Delay Line Detector Manual (11.0.1304.1) Page 31 of 83
To Next Page
Loading...