Have been thinking a lot about how to make the vacuum chamber.
The material has to be aluminium or inox steel.
Buying stuff can be very expensive, so I looked at a scrap metal dump to see if I have any luck finding useful material, there was lots of useful stuff, have to get back there tomorrow, here are some pictures:
Have been thinking a lot about how to make the vacuum chamber.
Have been working on the HV supply for the PMT this weekend.
First, using a toroid from a broken computer power supply, I made the transformer, 4wdg primary, 160wdg secondary:
The wire on the primary side is plain copper wire, secondary wire is thin isolated wire, primary is about 80uH, secondary has a resistance of 47.3 ohms.
Next, the low voltage side of the power supply:
I used the SIR-470DP MOSFET from Vishay, it was the only mosfet I had which could handle the current, but it's a bit of an overkill (I'm practising for the 50kV).
I used 0.02 ohm for the sensing resistor (the on resistance of the SIR-470DP is 2.3 mohm, a bit of a waste to use 20 mohm series just for sensing).
Also not in the schematic is a 2200 uF, 10 mohm capacitor across the 12V, and a 4.7nF low impedance capacitor directly next to the LTC1619 also across the power.
For the high voltage diode I used a TVR30.
Everything seemed to work OK, but the voltage on the feedback pin of was just 1V (should be 1.24). Analysing the thing with the oscilloscope showed continuous switching at 300kHz, with high positive peaks at the primary side of the transformer, and simultaneous negative peaks at the secondary side. Just not enough power on the secondary side to generate -1000V.
Maybe we are saturating the core (I don't know anything about the magnetic properties of the core), or maybe the primary inductance is too high.
I will try with fewer primary windings, and just to be sure I will replace the 10nF/3kV ceramic capacitor on the high voltage side (got it out of a broken TV, so I don't know how good it is, maybe it leaks).
At least the switching is OK! Just a matter of transforming the transformer to make it work.
Based on the LTC1619 this PMT power supply provides a voltage of 1000V, you can download de schematic (for LTspice IV, free simulator from linear technology) here.
The output voltage can be adjusted by changing the value of R4.
The mosfet can be any low on resistance 40Vds+ mosfet which can handle about 3A peak.
This is the simulation of the output voltage (click on the image for a higher resolution):
The simulation shows no ripple!
Going to put the thing together this weekend.
Oopss.. error in the schematic, L2 should be 16mH, not 4! (200/5)^2 = 1600, simulation shows that it works much better with L2=16mH.
To drive my turbo pump (on it's way! found it on ebay) I need a 3 phase power source.
Started with a simple oscillator, this is the concept:
You have to click the image to see the schematic.
It's just 3 times a phase shift of 60 degrees + inverter (180 deg) gives me 120 degrees.
Simple isn't it?
Now I need to add amplitude controlled gain (two diodes, another fet with a cap on it's gate will control gain of one of the stages).
The gain per stage should be 4 for perfect oscillation (at 60 deg phase shift absolute damping of RC circuit is 2, there is another divide-by-two because I use two capacitors per stage).
Adding frequency control is easy, when I insert a switch in the feedback loop, the whole thing acts as a sample-and-hold circuit, so frequency can be controlled by a PWM signal of about 100kHz (max oscillator frequency is 1.2kHz).
Then I need to add an output driver which takes the 3 signals and converts it to a high current.
I need voltage to current, because a voltage source would drive very high currents to the turbo pump at low frequencies.
Update: I have a turbo pump controler now, so I don't need the 3-phase oscillator. But if anyone wants to use this design as a frequency controllable 3-phase sine oscillator, you can add 3 switches from a CD4066 to the circuit between the drains of the fet's and the capacitor to the next fet. When driving the 4066 switches with a PWM signal (e.g. 100kHz) all the fet stages behave as a sample-and-hold during the off part of the PWM signal. By adding a capacitor across every drain resistor you can filter the 100kHz part of the signal. (if your max 3-phase frequency is 1000Hz, put the -3db point of the drain RC filter on 2kHz and you won't see any 100kHz in the output). I've tried and it works, gives a great 3-phase sine form 10Hz to 1kHz. (I did add some gain control with an extra fet between source and ground for one of the oscillators' fets).
Got some second hand vacuum feedthrough material, a whole box full of it, here is a picture:
Got 4 with one pole, 2 with 4 poles and 3 with 8 poles. The seller told me they can handle 5kV and 150A.
The pins on the vacuum side are quite long, I wonder if I could just cut them, material might be alumel.
Does anyone recognize them, maybe even knows the manufacturer? And how do I connect something on the vacuum side?
The manufacturer is "Varian", and they can (at least?) handle 12kV.
Which is too low for fusion, but enough to start measurements on my "new magrid" idea.
About connecting something on the inside, there seems to be some kind of "vacuum side connector". The material is not alumel, it's inox.
Got a reply from Pfeiffer and Balzers, after a phonecall with one of their service engineers I can order the spare parts and the oil.
Great! Service engineer was very friendly and very interested, "why do you have a WKP500 at home"?????? :)
So I refered to fusor.net, I hope he still wants to sell me the spare parts after reading it....
Took a look around at what I had "in stock" for high-voltage generation.
Found a very old 110 to 8000 volt transformer from a Xerox machine:
It's only 0.8 VA, got it from a dump when I was 14 years old, remember that I connected the primary to 220 when I got it, 2cm sparks!
Can be a little bit lethal....
See what the rest of my search brought me, high voltage capacitors:
"The electronic goldmine", no idea why I bought them years ago, maybe just because the were so "special".
For a cascade I also need som high voltage diodes, found some TVR30 diodes in my collection:
Looking good, the 0.8VA transformer power is too low, but the rest is ok.
Got myself an XP1000 photomultiplier tube for only 5 euro's!
Marktplaats is just great for all those people who want to build a nuclear fusion reactor.
I took a look at "Avalanche photodiodes", but they are way more expensive than used PMT's, they also need a high voltage (not as high as the PMT's, but still a high voltage), and they are less sensitive.
Finding the datasheet of the XP1000 turned out to be a bit hard, even the original manufacturer (Philips) didn't have it anymore.
So if anyone out there has a complete datasheet of the XP1000 PMT, please leave a note on this blog!
Everything you ever wanted to know about photomultipliers but were afraid to ask is here.
I did find some info about the XP1000 in an online "radio museum", it's not much, but enough to get it working.
The moderator can be normal plastic (PE) and I got some scintillator foil from ebay. It's ZnS with Ag, came with documentation, got to read it 3 or 4 times to make some sense out of it.
I would prefer to use BC-720 as a scintillator from "Saint Gobin Crystals" since it would only detect the fast neutrons, it would be insensitive to gamma en I would not need the moderator.
But it's too expensive. If anyone has some fast-neutron specific scintillator for sale, please leave a note on this blog, I cannot find it on any surplus or second hand site.
Aren't tubes just great! Makes me think of Edison who first found out about "cathode rays", sure if people like Franklin, Edison or Tesla were still alive they would work on fusion now.
We need fusion, we cannot keep burning oil and we sure cannot produce radioactive material which will be dangerous for another 10000 years!
We are unable to plan something over more that 5 years, so what are we thinking when producing radioactive waste!
For now, the vacuum pumping and chamber comes first, but in time I would need to calibrate my neutron detector.
If there is anyone out there who has a source of fast neutrons of known intensity and wants to calibrate my neutron detector, please leave a note.
So I decided to build my own nuclear fusion reactor.
After reading fusor.net and everything I could find about the work of Bussard, my conclusion was that we needed a different type of magnetic field, a different kind of "magrid" where the magnetic fields work much more in favor of "star mode".
To test this, I started to find parts on ebay and markplaats.nl
Got a very good deal on a pfeiffer and balzers WKP500 pump, so for the vacuum I still need a rotary vane pump and a turbomolecular pump.
It was dirt-cheap, and it's broken, one of the "looking glasses" for the oil is broken, and I don't have 3-phase power to test the motor.
For the rest it looks ok, so I will replace the glass, get some new oil, and test the motor (maybe I should get a 3-phase diesel generator to power my reactor).
The WKP500 is heavy! 95 kilo's, had a hard time getting it out of my car on my own. I should get some wheels and iron bars to weld myself a portable "vacuum setup cart".
Here is a picture of the WKP500:
A bit dangerous on the workmate, if the workmate breaks! So it's on a pile of stones now.
Emailed Pfeiffer and Balzers about spare parts and oil.
With a rotary vane pump this will get me down to 0.1 Pa.
Read more about these wunderfull pumps here