Page 14 of 15

Re: Mark3 operation 3e-4 torr / quad ion source @ -40kV, 1mA

Posted: Wed Mar 29, 2017 12:01 am
by Andrew Seltzman
The fusor has been tested with the new molecular drag pump and dry diaphragm pump. The MDP was mounted to the high vacuum solenoid valve on the fusor and was running off of it's built in AC power supply, while the diaphragm pump was running off of the fusor's 24Vdc supply. Vacuum performance was generally good. The solenoid valve between the fusor and MDP was opened at ~100s while the fusor was at ~100mTorr, pump down into the high 1e-4 range occurred with generally good performance. Deuterium pressure was held at ~8mtorr and the fusor generated ~2.3E6n/s at 40kV and 17.6mA, also generally about the same as running with the turbo station.
MDP mounted on fusor
My last deuterium tank had run out of pressure, having only ~90psi when I got it, it lasted about a year of intermittent use. New deuterium tank has ~1750psi in it and should last a significant time. The tank is now mounted inside the fusor frame(more compact and better protected) and all VCR fittings have been upgraded to copper o-rings rather then the nylon washers I was using for testing.
The fusor's going to be hooked back up to the old turbo station until I can get the new one with the MDP built though.

Re: Mark3 operation and upgrades

Posted: Sun Jul 08, 2018 9:50 pm
by Andrew Seltzman
A recent upgrade to the mark 3 has replaced the 6 way conflat cube with an 8 way kimball physics vacuum hexagon. The hexagon is considerably smaller and more compact, and has 2 more ports then the cube. The hexagon is connected to the core with a brazed ceramic isolator with copper conflat o-rings, which replaced the PEEK plastic isolator with viton o-ring. In addition the hexagon mounts an MKS 901P with integrated display, x-ray shielded 1.33CF viewport, deuterium inlet valve, Pfeiffer EVI 005 vacuum valve, an electron collector probe on a BNC feedthrough, and a germanium viewport for IR measurements. This upgrade replaced 1 plastic insulator, 2x 2.75CF viton o-rings, 1x KF40 Viton o-ring, and 1x KF16 o-ring with all metal parts, is about 4kg lighter then the previous unit, and significantly more compact, which will allow a diaphragm pump / turbo pump system to be integrated into the frame. The MKS 901P is the RS485 version which will allow several vacuum gauges and the turbopump to talk on the same bus.

The Pfeiffer EVI 005 valve was the one tested in this post

The deuterium inlet valve was welded with this procedure:

CAD designs on the hexagon are attached below
Pictures of the completed hexagon
Comparison to the old vacuum cube:
Vacuum tests of the new system are very encouraging, the lower conductance of the smaller valve is offset by the removal of the viton and plastic; the system now archives considerably better vacuum performance, and will max out the gauge at below 1e-5 torr.
First pump down with bakeout starting at ~700s and pressure rise with valves closed.
In addition a new digital control system is in the process of being designed, which will replace the older analog feedback controllers with an arduino controller and provide a touchscreen display for monitoring and control

Re: Mark3 operation 3e-4 torr / quad ion source @ -40kV, 1mA-9mA

Posted: Mon Jul 09, 2018 6:15 pm
by Richard Hull
Fabulous upgrade Andrew!! One of the better systems out there and a fusioneer that has been around a while. Thanks for the upgrade. I am envious of the valved off 1000+ second hold to about 50 microns. I am at 100 microns after about 360 seconds, but I have operated like this for about 9 years now.

Again, nice stuff! ( as an old friend used to comment to me).

Richard Hull

Re: Mark3 operation 3e-4 torr / quad ion source @ -40kV, 1mA-9mA

Posted: Mon Sep 03, 2018 9:58 pm
by Andrew Seltzman
Latest upgrade progress:

Diaphragm pump, rough vacuum gauge, and turbo pump to be installed on fusor frame.
CAD design of mounting position:
Vacuum gauges and turbo pump will share an rs485 bus for communication. The USB to RS485 dongle used for turbo pump testing is the following: ... UTF8&psc=1

Pump actual speed, set speed, and controller current, voltage, and power.
Matlab control program for pfeiffer turbo pump:
(2.27 KiB) Downloaded 130 times

Re: Mark3 operation 3e-4 torr / quad ion source @ -40kV, 1mA-9mA

Posted: Wed Oct 17, 2018 4:38 am
by Andrew Seltzman
More upgrades for the fusor:

Upgrades for mounting internal vacuum pumps, custom backing and mounting bracket for the TPD011 turbo pump
Front and back of bracket, replaces the NW16 backing connector and provides 4x 10-32 tapped threads on a 1.75"x97mm pattern
Turbo pump and diaphragm backing pump
Mounting plate to hold vacuum pumps
Locations of vacuum pump within fusor frame
Thermal camera mounting system on vacuum hexagon to observe grid temperature
Right angle bending mirror
Optics path

Re: Mark3 operation 3e-4 torr / quad ion source @ -40kV, 1mA-9mA

Posted: Sun Nov 11, 2018 2:48 am
by Andrew Seltzman
The new CNC'd mounting plate for the internal vacuum pumps have arrived and the turbo and diaphragm pump are installed within the fusor frame.

Diaphragm pump, turbo pump and MKS-901P gauge for roughing line pressure measurement are mounted on the aluminum plate.
Plate mounts on top of fusor
Vacuum line connects turbo pump to foreline solenoid valve
Pumps powered up and backing vacuum gauge added to the RS485 bus
Since both gauges are now on the RS485 bus, backing vacuum and high vacuum can be measured simultaneously, pump down was measured with the turbo at standby speed, at full speed the pumping rate increases. On standby, ultimate vacuum is below 1e-5 torr.
Testing of the thermal camera was also successful.

Re: Mark3 operation 3e-4 torr / quad ion source @ -40kV, 1mA-9mA

Posted: Sun Nov 11, 2018 11:11 pm
by Richard Hull
Wow! You are already in the show-piece class and I sense a move into museum grade exposition. This is how it is to be done if one is "long-term serious" about making a world class fusor system and sticking with it. Great work Andrew! Turbo standby mode is more than adequate to do fusion........

Richard Hull

Re: Mark3 operation 3e-4 torr / quad ion source @ -40kV, 1mA-9mA

Posted: Sun Nov 18, 2018 12:34 pm
by Dennis P Brown
Very impressive system. When money is no object, one can create a very professional system!

Your thermal camera might need optical filtering to reduce the intensity since that filament will be in the thousand centigrade range. So you should keep that in mind to provide room for any such pre-lens filtering. Also, noise might be an issue from the EMF.

Re: Mark3 operation 3e-4 torr / quad ion source @ -40kV, 1mA-9mA

Posted: Sun Nov 18, 2018 10:27 pm
by Andrew Seltzman
The thermal camera is viewing the grid during operation, my grid is liquid cooled so it will remain at fairly low temperatures as long as there is coolant flow. There is a little EMI noise on the video, though I have not yet cleaned up the wiring and the video is being sent over twisted pair wire.

Re: Mark3 operation 3e-4 torr / quad ion source @ -40kV, 1mA-9mA

Posted: Sat Feb 02, 2019 5:09 am
by Andrew Seltzman
I have recently added an x-ray camera to one of the fusor's viewports to image the locations electrons are striking the shell through bremsstrahlung intensity distributions.

The x-ray sensor is a Hamamatsu S10814 CsI crystal needle array coated on top of a fiber optic plate collimator/shield coupled to a 1700x1200 CCD detector, controlled by a C9266 USB readout box. This is the industrial equivalent of the modern digital dental x-ray sensors. The sensor has 20x20um pixels, 42x30mm size, 20Lp/mm resolution, and probably on the order of 60 LSB/mR sensitivity. ... d1124e.pdf
The sensor gets mounted on a thorlabs optical assembly holding a 1.5mm molybdenum pinhole looking through a Pyrex glass viewport which has negligible attenuation for the more energetic x-rays.
The first x-ray images from the camera are shown below
Remember that my fusor has magnetic deflector yokes on 4 of the nipples, shown circled in red.
SAM_4425a.jpg ... shield.htm

These magnetic deflectors were installed to prevent the electron beam emitted through the grid openings from striking and damaging the optical viewport. Instead of striking the viewport, the beam is deflected into the wall of the nipple. In this case, x-rays from electrons hitting both the near and far nipple walls are visible, along with the x-ray attenuation shadow due to the grid insulator and molybdenum field control cap. The far electron beam is striking the top of the nipple.
Rotating the far deflector field 90 degrees moves the electron beam strike point from the top to the side of the nipple
Rotating the far deflector field 180 degrees moves the electron beam strike point from the top to the bottom of the nipple
Removing the far deflector from the nipple causes the electron beam to illuminate an area of the far nipple behind the grid
In this case the grid wires and insulator are more clearly visible