Microchannel Plate Q&A
(asked by our customers, answered by our technical support)

Q: What is the difference between your two phosphor screen mcp's. I see that one has a glass plate and the other has a fiber plate, but is there an advantage to one over the other?

A: It depends on image readout method. In case of phosphor screen on a glass plate (GPS model) image is registered with a CCD camera or visually (for example, in case of beam alignment application). When using phosphor screen on the fiber plate (IFP model) it can be transferred directly to CCD using fiber tape without additional optics.

Q: Ultimately, I would like to have a MCP with phosphor output to look at electron spatial distributions, but I will also need access to the time resolved signal. Can I pull the time resolved signal off of the phosphor screen just as if it were a metal anode? If so, are there any cautions, recommended ways of doing this, and what time resolution can I expect?

A: Phosphor screen is coated with a thin metal layer (typically Al), which is important in order to avoid charge effects and obtain uniform electric field between MCP-Out and phosphor. This metal layer act as a regular metal anode and provide a time resolved signal. Luminescent signal directly from the phosphor screen can also provide time-resolved signal, which is limited by the phosphor respond time. We offer P47 phosphor that have decay time about 80ns, which is much shorter that decay time for the most popular phosphor P20. New, faster phosphors are under development.

Q: How do you recommend making connections to your mcp's? The spot welder I have access to is rather large. How big are the connection tabs, they look quite small in the pictures? What material are they?

A: MCP housing metal is KOVAR. It’s chosen because it had thermal expansion coefficient close to glass and ceramics. Our assemblies usually supplied with stainless still wires spot-welded to the tabs.

Q: Could you tell the advantage of having a fibre optic plate against a phosphor screen, and how does one obtain an image from the plate? Can you use a fibre optic bundle to relay the image to another flange?

A: Image is registered with a CCD camera or visually (for example, in case of beam alignment application). When using phosphor screen on the fiber plate (IFP model) it can be transferred to CCD using fiber taper or relayed to another flange using optic bundle.

Q: (question received from reseller) This customer has received this detector, but it's not clear to know how to connect/supply the electrical operation condition. This customer thinks it should be 4 electrical connector, but there is 3 connector.

A: The reason we have 3 connectors (instead of 4) is the fact that we use "matched" MCP pair. That means MCPs have the same resistance and are in direct contact with each other. Matched pair needs only one voltage to be applied between MCP-In of the first MCP and MCP-Out of the second MCP. In the attached file "MCP wiring" we show one example of MCP detector wiring. Please note that only voltage difference between MCP-In, MCP-Out and Anode is important for detector operation. An absolute values of all this voltages depends on your customer set-up, type of particles to be registered (electrons, ions, photons etc.). For example MCP-In may be positive, negative or zero (ground). U (MCP-Out) = U (MCP-In) +2000V, U (Anode) = U (MCP-In) + 2200V. The values of impedance and capacitor depend on customer registration electronics. Typical values are 50Ohm and 1000pF.

Q: During unpacking, are there some warning procedures?

A: Detector should be opened carefully and placed  in the vacuum chamber as soon as possible.

Q: From your mail I understood the following: you can purchase the MCP-GPS34 assembly complete but demounted. We are interested in that  solution but I like to be sure not to receive just the two MCP's and the phosphor screen. I know very well that critical points are the correct  distance between MCP's and from screen and MCP for that reason I like to buy a tested assembly like yours.

A: An assembly consists of two main elements: MCP holder and luminescent screen. You can see 3D images of both parts in attached files. These two elements are fitted to each other in a way to provide optimum distance between Phosphor screen surface and MCP-out surface. This distance can also be further adjusted (if required) by placing a foil ring (30-100micron thick) on MCP holder before placing MCP. These assembly is designed and tested to obtain maximum spatial resolution in night-vision applications - and of course it will also work with maximum spatial resolution in other applications.

Q: Concerning the spatial resolution the value you quoted it is comprehensive >of the phosphor response ? Which kind of phosphor it is used (P11, P20 ...)?

A: Standard phosphor used is P20. Any other can be applied at customer request. Spatial resolution is limited not by the phosphor type, but the phosphor particles size. If particles are small enough, then the limiting factor is MCP channel size and pitch as well as distance/voltage between MCP and phosphor screen.

Q: We did not use the flange mounted detector received from you yet. The reason was we did not have experiments in ultra-high-vacuum, just in  vacuum of 10e-7 torr.

A: Of course it will work at this vacuum too. If you do not use it, please try to keep it in vacuum or in a dry atmosphere since the MCP glass react with a moisture.

Q: The detector worked pretty good with clusters but due to some voltage jumping (over 3kV) from power supply we lost the MCPs.

A: You should use a current limiting resistor to avoid serious damage to MCPs in case of discharge.

Q: I need you to confirm that we can also use the mcp for positive ion detection, i.e. have a large negative voltage on the front plate while the back side is on -50 to -200V and the anode on ground.

A: Yes, you can do that. However, maximum voltage difference applied between MCP-In and MCP-Out for 2 MCP assembly should not exceed 2200V (1100V per plate). For small ions 2keV is enough energy for efficient electron emission from MCP surface. However, if you plan to detect heavy ions, for example biomolecules, you need additional acceleration voltage. As you probably know most of MALDI-TOF systems accelerate ions to 20keV and higher.

Q: It is not clear to me if I have to directly supply the voltage to the anode, i.e. if I have to connect the fin of the anode to a power-supply, and extract the output signal from a T-wiring connection?

A: When applying high voltages to MCP detector electrodes including anode care should be taken to avoid unwanted discharges that can destroy MCPs. Discharges can develop in low vacuum conditions as well as in high vacuum along detector isolating surfaces if they are not clean enough or due to deposition of different materials from ion sources, pumps etc. The main approach to protect detector is to place safety resistors that will limit maximum average current in corresponding circuits. Typical resistance value of 100Kohm will limit maximum average current to 20mA at high voltages around 2kV. Signal connector should be wired to anode through a High Voltage rated (3kV) capacitor. Typical capacitance is about 1000pF.

Q: Incidentally I am confused by you web description that seems to imply that a time resolution of less than 1ns can be obtained using a simple planar metal anode. I thought one had to be careful to match the anode to the (50 Ohm) output signal cable, usually using some sort of matching cone? and Is it necessary to make an impedance adaptation of the anode? I am extracting the signals with an UHV-compatible 50Ohm coaxial cable and sending them to an amplifier and to a discriminator (pulse-mode operation).

A: For typical anode capacitance of 4pF and R=50 Ohm RC=200ps. If you are looking for time resolution in subnanosecond (0.2ns) range you should think about impedance matching (adaptation) of the anode. Creating specially shaped anodes usually does it. In most applications time resolution is limited by registration electronics and is in 1-10ns range. In this applications signal is integrated by electronics with a typical time constant much larger than RC.

Q: I read that the resistance of each MCP in my assembly is 2.0*10^8 Ohm. Does this mean that the overall resistance MCPin-MCPout in chevron assembly is 4.0*10^8Ohm?

A: Yes, an overall resistance R of the MCP assembly in chevron configuration is a sum (R1+R2) of MCP-In resistance R1 and MCP-Out resistance (R2).

Q: What is the value of the MCP-Anode capacity?

A: The capacitance is affected by three factors: the area of the plates, the distance between the plates, the dielectric constant of the material between the plates. We can estimate anode capacitance in two different ways: as single metal electrode or as a capacitor with two parallel plates (anode and MCP surface). For a single metal electrode in vacuum an estimated capacitance C is equal C=D/(9*10^11) F=1.1*D pF, where D is the size of the electrode in cm. For electrode size D=3.5cm C is about 4pF. The capacitance of a capacitor with two parallel plates in vacuum can be estimated using the formula: C = 9A/d, where C is capacitance in pF (picofarads), A is the area of one plate in m2, and d is the distance between plates in m. This estimation also gives a value about 4pF.

Q: I read that the MCP-in is supposed to be operated at ground voltage. Is it possible to apply it a slightly positive voltage, let's say +50 Volts, to induce electrons to hit the MCP?

A: In general, MCP assemblies can be operated with any electrode (MCP-in, MCP-out or anode) at a ground potential. When detecting electrons it's also possible to apply positive voltage to MCP-In electrode, keeping voltages between MCP-In MCP-Out and MCP-Out and Anode as recommended:
Between MCP-in and MCP-out: Set this voltage according to the required gain, 700 -1000V per MCP typical, 1100 V maximum, MCP out at positive polarity.
Between MCP-out and single anode: This is normally set at about 100 - 200 V.

On the other hand, one should take into account the following: Positive potential on the MCP-In electrode can result in higher noise signal from residual electrons in the vacuum chamber. Background electrons can be generated by ion pumps, external laser and UV sources, electrical discharge near high voltage electrodes etc. Adding small positive voltage (around +50V) may be not enough for efficient detection of electrons. It was experimentally shown that for better detection efficiency incoming electrons should be accelerated to about 700V.

Q: I read that the voltage MCPout-Anode is +200 Volts. Do you mean that when the MCPout is biased to +2000 Volts, the voltage induce on the anode is typically +200 Volts?

A: Recommended voltages are as follows: MCP-In = 0 (ground), MCP-Out = +2000V, Anode = +2200V For efficient detection of incoming electrons the following voltages can be used: MCP-In = +700V, MCP-Out = +2700V, Anode = +2900V

Q: Which kind of connection is set on the output anode to collect the signal? BNC, SMA, SHV coaxial?

A: Detector electrodes including anode have fins that can be connected to any standard connector.

Q: Can the whole stuff be baked? Can be operated in UHV (10^-10 Torr) without contaminating vacuum?

A: Whole MCP assembly can be baked up to 350°C. It's UHV compatible.

Q: Does MCP detector present insulator parts in the front, which can be charged by electron impacts?

A: No.

Q: Does the detector ship as an assembly?

A: We ship detector as assembly. We also can ship detector premounted on standard vacuum flanges.

Q: Do you have a typically wiring diagram for detecting positive and negative ions in a ToF? What external resistors and capacitors are recommended?

A: Typical wiring is described in the brochure. When MCP input is at a ground potential, typical wiring diagram is the same for detection positive or negative ions in ToF systems. External resistors and capacitors should match input parameters of preamplifier. Most commonly used are 50 Ohm and 1000 pF (rated up to 3kV).

Q: You don't have an electrode between the MCPs, which means that you do not ensure that you have the same potential drop over both MCPs. Is the resistance of the MCP plates so well defined that this is not a problem?

A: You are right, there is no separate electrode between MCPs in our standard assemblies. We use matched MCPs. Matched means that they have same resistance within 10% (we choose usually even less difference). We also make sure that the nominal operation voltage of two MCPs is about the same.

Q: We were thinking to remove the 3 fins that are supposed to be used for biasing the MCP and substitute them with some connectors we made. In order to attach these new connectors we were thinking to spot-weld them to the little metallic parts where the bias fins are actually connected. Anyway, this metallic part are directly connected to the MCP, so they get an electric discharge when I make the spot-weld. My question is if the spot-welding operation can somehow damage the MCP.

A: When making a spot-welding, place fins directly on the ground electrode (copper plate). In this case the current will go directly through this part without circulating around electrode. That will minimize any effects of welding current on MCP. It’s also necessary to protect MCP surface from any hot particles generated by sparks. Just place any protective screen between welding sport and MCP to prevent direct exposure of the MCP to any possible discharge erosion products.

Q: I am confused by your web description which says the MCP has diameter 24.2 or 24.8mm, thickness 0.46mm, and effective area 18mm. Presumable you mean effective diameter. Is it that the clamping ring is about 3.5mm wide, or have you allowed for other losses at the edge?

A: In fact actual effective diameter is larger than 18mm, but MCPs with diameter of 24.8mm are traditionally rated by MCP manufacturers as 18mm active diameter plates.

Q: I have looked through your web page but I was unable to spot a microchannel plate assembly associated with a metal anode that was matched to, say, a 50 Ohm transmission line. I am interested in an overall timing resolution of <100 picoseconds; 50 ps would be better. I am not interested in a position sensitive readout.

A: We do not offer assemblies with 50-100ps resolution at this time.

Q: I noticed that the bias angle is different from the plates we are using now. What are the consequences of this difference? I would guess that this would result in a loss in gain. (We use two plates in a Chevron configuration.)

A: As you probably know, the main application of MCPs is in the night vision devices. In this application typical currents are much higher than in ion-detection applications. Large bias angle (about 12-13 grad) is used to avoid an effect of ion feedback on photocathode. Ion feedback is a flow of ions spattered by electron avalanche from MCP channel wall, in the direction opposite to electron avalanche, towards MCP entrance and then onto photocathode surface. Ion feedback is negligible in ion detection applications and has no any serious effect, as there is no photocathode at all. In general, gain depends on the bias angle, but usually this dependence is not important and can be easily compensated by applied voltage. In fact gain variation in different MCP manufacturing runs is larger than gain variation due to the different bias angle. The figure in the attached file shows gain dependence for MCPs that are shipped to you. The gain of 1000 is at 760V. The gain depends exponentially on the applied voltage. Typical voltage increase necessary for 10 times increase in gain for MCP-33-10E is about 180-220V.

Q: 1st contact (front of first plate): -2,000 V > 2nd contact (back of second plate): 0 V > 3rd contact (phosphor screen): +4,500 V

A: those voltages are fine.

Q: I was also wondering about putting resistors in series with the high voltage power supplies that I use for the MCP. About 5 years ago the last  time I use a Chevron MCP I put resistors in series in order to limit the current going through the plates and to the screen. I burned out an assembly when I did not use these safety resistors. What current can my MCP handle and hence what values for safety resistors do you recommend?

A: When applying high voltages to MCP detector electrodes including anode care should be taken to avoid unwanted discharges that can destroy MCPs. Discharges can develop in low vacuum conditions as well as in high vacuum along detector isolating surfaces if they are not clear enough or due to deposition of different materials from ion sources, pumps etc. The main approach to protect detector is to place resistors that will limit maximum average current in corresponding circuits. Typical resistance value of 100Kohm will limit maximum average current to 20mA at high voltages around 2kV. Signal connector should be wired to anode through a High Voltage rated (3kV) capacitor. Typical capacitance is about 1000pF.

Q: We have a few spare MCPs in our lab. However some of them look a little bit suspicious. There are some small spots on the surface which could appear due to the long storage (what do you think about such a possibility?) or they could even be use before by somebody and just damaged by clusters or ions.

A: MCPs should be stored in a vacuum or in a dry atmosphere. When using silica gel (or similar absorber) to keep MCPs in a closed container it’s important to check humidity regularly. Silica gel absorb moisture over time and became inefficient. Even more, due to the temperature fluctuations it may be even dangerous to keep MCPs in silica gel containers for a long time. Imagine that silica gel have saturated during cold winter months. If it’s temperature increase by few degrees due to the warmer room temperature in summer – it will desorb a lot of water vapor back into the closed container creating critical atmosphere for MCP storage.

Del Mar Ventures 4119 Twilight Ridge San Diego, CA 92130 tel (858) 755-6727 fax (858) 755-6771 v/efax (858) 630-237 v/efax: (509) 752-0123 delmar@sciner.com http://www.sciner.com/ http://www.femtosecondsystems.com/