Sunday, April 12, 2020

Diffusion Cloud Chamber (Part 2)

After the cooling element was constructed and installed we did a quick test run to prove the concept and to make sure that we can achieve our goal of -36°C.

The IR thermometer I used is not at all accurate at very low temperatures! I had a black aluminum tape over the metal for this measurement in order to reduce the reflectivity and improve the emissivity of IR and it still showed about 10°C error. The actual temperature, established later with a thermocouple was -37°C - still exceeding our design goal.

Wiring of the front panel. We used the rest of the 2RU aluminum blank for the front panel. All Cloud Chamber control switches, LEDs and the cold plate thermometer were mounted on the front panel. The back side was left open for the water feedlines, power connectors and air cooling.
The Cloud Chamber is powered with two individual Power Supplies.
The main Power Supply is a 450W PC ATX power supply. This type of power supply serves the purpose well as it provides enough power and variety of voltages for all of the necessary circuits - the heater element (+12V), Ion Scrubber HV supply (+3.3V), Light elements (+5V), Front Panel Thermometer (+5V), First stage TEC (+3.3V), Second Stage TEC (+6V (or 12V/2)) and 5th Stage cooling fans (+12V). Same power supply also powers the Liquid Cooling Plant (+12V and +5V).
The ATX power supply plugs in directly, in the female connector of the main wire harness and it is switched on/off by the main power switch on the front panel.
The second power supply (Power B) is 15V/32A (Alinco DM-330MVT) and it is strictly dedicated to Stage 3 TECs. The voltage in Stage 3 should never exceed 15V! (maximum rated voltage for TEC1-12709 module is 15.1V). All wiring for Stage 3 is done 12AWG stranded wire due to the high current demand of  the dual TEC1-12709 modules.

To facilitate the wiring of multiple circuits, three 5-position terminal blocks were mounted. One block is all GND terminals, another block is dedicated to the TEC modules and the third block is used for all other circuits - HV PS, Lights, Heater element, etc. The ATX PS main wire harness was connected on one side of the terminal blocks. We used PC PSU Extender Cable for the main power harness - one side comes with a female ATX connector - the male connector on the other end of the extender cable is simply removed and all wires are terminated with spade terminals.

The top side of the Cloud Chamber under construction. Visible is the Cooling Unit, attached with countersunk brass screws to the plywood. 4 x Black Delrin 5" x 1/2" rods are attached with stainless steel machine screws (1/4"-20 x 1") to create the "support pillars" for the Positive mesh of the High Voltage Ion Scrubber and the Alcohol Evaporator Unit / Heater sitting on top..
Delrin (Polyacetal) is a good choice as it is resistant to the hostile environment of super-cold super-saturated alcohol vapors and it is easy to work with. Other possible choices are Ceramic or Teflon. Plexiglass/Acrylic will become brittle under these conditions so I don't recommend it.
The Delrin rods I got were solid so I had to drill and tap all rods on both sides for 1/4-20 thread. One rod was drilled end-to-end to act as a conduit for the heater wires. I drilled a small hole on the side about 3/4" from one of the ends. This hole was drilled at 45 degrees to aid with the cable routing.
The rods are placed very close to the inside wall and corners of the Cloud Chamber Cover (outlined with pencil marks on the top surface) while making sure there is enough clearance to slide freely the cover.

The fine aluminum mesh of the Positive terminal of the Ion Scrubber is stretched over a brass tube frame, using a very thin silver-plated wire. The frame is constructed out of 4 brass tubes, slightly larger in diameter than solid 12AWG wire. Each corner of the frame is a solid copper 12AWG wire loop formed in a loop around the Delrin rods with ends at 90 degrees. The ends of the loop are inserted in the brass tube segments and then soldered. This arrangement is allowing the frame / mesh to be moved up or down, sliding over the corner posts. We wanted to be able to adjust the height of the mesh in order to control the intensity of the electric field above the Active zone. A number of rubber O-rings were placed over the Delrin rods below and above the corner loops of Ion Scrubber mesh to act as "movable stoppers" suspending the mesh at an adjustable height.
Another brass frame made of flat brass stock supports the tin-plated steel mesh (not installed yet in this picture) of the Alcohol Evaporator Unit (AEU). This frame is fixed on the very top of the Delrin rods. Each corner is a copper washer to which we soldered the brass frame elements and the frame is then secured on top of the support pillars with stainless steel screws. The AEU support steel mesh is soldered on the bottom of this brass frame.

This is the completed support structure with the Ion Scrubber HV mesh, the Alcohol Evaporator Unit (the heater element is visible on the very top). The black and green O-ring stoppers and another green O-ring which acts as an "insulator rib" to lengthen the electrical distance between the high-voltage mesh and the Alcohol Evaporator.
(!) Since the Cloud Chamber is filled with a mixture of air and alcohol vapors during operation, an extra effort was made to prevent any remote possibility of arcing! The temperature in the top portion of the volume where the HV mesh is located is above the alcohol's flash-point temperature. Last thing one wants is a fireball! The HV operates at 4.6kV and the Air breakdown voltage is 3kV/mm, so a minimum of 3 mm distance (~2x the breakdown distance) to GND is absolutely mandatory! Any High-Voltage Arcing inside the Cloud Chamber can ignite the alcohol vapors and can also destroy the main Power Supply. This is one of the reason why the heater supply wires were conducted on the inside of the Delrin tube and not just wrapped around it when passing the HV mesh.

The almost finished control panel with all SPST (Power and Power B) switches, DPDT switches, LED indicators and the Industrial Grade Thermocouple Thermometer for the Cold Plate. A quick test showed a good temperature drop. The digital thermometer's display mode button was installed later in the build process.

Visible, on the right-hand side is the High-Voltage Ion Scrubber power supply. This was taken out of an Electric Fly Swatter /Zapper. The circuit is made of a DC pulse converter, high-voltage step-up transformer and a voltage multiplier. At 3.3V input voltage I measured output of 4.6 kV.
CERN's recommendation for the Ion scrubber is minimum of 100 V/cm and 4.6 kV is more than enough to satisfy it.
The HV supply is housed in a poly-carbonate plastic box (from CCI .22 Cal Ammunition) to provide additional insulation and safety. A special, high-voltage, Teflon insulated, corona-free wire (rated at 15kV) is used to connect the positive output to the top-side mesh. The negative pole is connected to both, the aluminum plate and to the copper cold plate of the cooling element.
The HV supply has a capacitor on the output and since this is a school project, I would imagine that no kid would like to be shocked by 4.6kV. I designed a dedicated Discharge circuit to prevent any unpleasant accidents. Half (one pole) of the Ion Scrubber DPDT switch is used to control power to the HV PS. The other pole of the same switch (on the opposite throw), when the HV PS is in OFF position is used to short the output of the HV PS thru a 15 kOhm / 3W metal-oxide resistor to the negative side. The middle position of the switch is OFF, the UP position turns power ON to the HV PS and the DOWN position is DISCHARGE - it shorts the HV output capacitor (In this position, the power to the HV PS is already turned off).
The safety procedure requires the Cloud Chamber to be opened and manipulated with the switch in Discharge position. This circuit can be avoided if enough time is afforded after HV is switched off, for the voltage to drop down due to HV capacitor and wire leakage - I measured this to be approx. 30 sec. Using a Discharge circuit just takes the guessing out and guarantees that the mesh is at 0V.

1 comment:

Willy Tarreau said...

Great work, thanks Andrey for sharing this, too bad I didn't see it before building mine :-)