Archive for September, 2011
Recently, I have been working on a time to digital converter (TDC) core for Xilinx Spartan-6 FPGAs. I was contracted by CERN to develop it, and everything is available under the LGPL license from their open hardware repository.
The reasons for using the FPGA-based solution compared to the ASICs available on the market are flexibility (e.g. channel count), low latency (which enables applications such as fine delay generators that they cannot do with the current chips), low cost, more sourcing possibilities, and increased openness.
I still need to test the core on real hardware and precisely characterize its parameters, but you can already download the source and read the documentation. TDCs are useful even outside the context of large particle accelerators, for example in LIDARs and laser rangefinders.
Yesterday, I had the pleasure of visiting an incredibly cool one-man vacuum tube laboratory called PWL. The name stands for “Prywatna Wytwórnia Lamp” in Polish, which translates to “Private Tube Manufacturer” – a pun on PWLR (“National Tube Manufacturer”), the first Polish tube manufacturer after WWII.
First – it is awesome. The amount of ingenuity that Aleksander Zawada, who runs PWL, has had to keep his hobby affordable is remarkable. I particularly liked the microwave oven transformer based spot welding machine, and the hacked turntable used to spin glass while melting it (more on this later). Before the laboratory moved to its current premises in the Institute of Vacuum Technology, Aleksander started everything in his own private apartment.
Aleksander was kind enough to show me the full making of a triode. But mind you – he does not stop at triodes. The place is full of contraptions that he built: canal ray tubes, a RGB magic eye, several Crookes tubes, multiple (still untested) attempts at making Geiger tubes, and many other incredibly amazing devices.
He starts the triode by assembling the grid. To do this, he takes a piece of nickel wire, and soldered a small spiral of molybdenum wire on it – one turn and one solder at a time. He uses molybdenum because of its low emission of free electrons when heated (which causes unwanted grid current in tubes) and its high melting point. Soldering is done with a spot welding machine, which passes high current through the parts to be soldered (nickel and molybdenum wires). The current is so high that the metals heat and melt locally and form a small solder spot. How does one obtain such a high current? Aleksander simply took the transformer of a microwave oven, removed the high voltage secondary, and wound instead a few turns of a thick aluminum bar whose ends are connected to the copper electrodes of the welding machine. The solder current can be controlled by a triac-based dimmer connected in series with the transformer’s primary.
Aleksander then prepares the tube’s anode, which is simply a small piece of metallic sheet bent into a cylinder. He then solders everything together into the lamp base, which consists in a piece of glass with four electric wires going through it in a vacuum-tight fashion. This part was pre-made to save time, but he showed me a few ones that he built. The assembly included the tungsten cathode – yes, the spot welder can melt tungsten – and the barium-based flashed getter reservoir. The directly heated tungsten cathode is the simplest one can make, but Aleksander is also able to make cathodes coated with substances such as yttrium oxide for better electron emission capacity.
Before the electrodes are ready, Aleksander cleans them with a series of ultrasonic baths filled with three different solvents. The purpose of this treatments is to remove fingerprints and other organic deposits, which would otherwise evaporate and ruin the vacuum.
Now is time to prepare the lamp’s bulb. First and foremost – imprint a PWL logo into it. He uses a stamp with a special formulation of paint which, when heated, melts into the glass and marks it permanently.
As a cheap source of bulbs, Aleksander uses glass enclosures he easily obtained from light bulb factories – they had plenty of surplus stock when incandescent lights were removed from sales in Europe.
Now, let’s assemble the electrodes and the bulb together. He does that by melting the bulb into the base, by rotating it and heating the bottom with a propane torch. Of course, the spinner is also made of junk materials – this time, a turntable that was originally intended to play vinyls.
After the tube has been heated so much, the cool-down must be very slow or the glass would break. Aleksander has a small electric oven for this purpose, into which he places the tube to ensure a safe return to room temperature.
Ok, now the tube is cooling down, and the vacuum pumps are getting ready. In the meantime, why not have a look at those phosphor-coated screens in this corner of PWL?
Aleksander supplies some to a nuclear research facility about 30km of Warsaw, where they are used as X-ray fluorescent screens. He makes them using two methods. The first one, the simplest, consists in roughening the screen’s surface using glass powder scratched into it, in a process similar to using sandpaper. He then applies the phosphor powder, which gets into the microscopic grooves of the scratched screen surface and adheres to the glass. Note that the phosphor often does not actually contain phosphorus, but other chemicals such as zinc sulfide and europium, depending on the color and other properties.
The other method is based on sedimentation. Aleksander prepares a suspension of the phosphor powder in water, filters it to keep only the smallest grains, and adds a mixture of potassium silicate and strontium nitrate which helps with bonding the phosphor to the glass. The powder slowly sinks to the bottom and attaches itself to the glass.
Beep! Beep! The vacuum system calls for our attention. The diffusion pump has reached its working temperature and is now ready to evacuate our triode. Aleksander obtained the pump from the garbage of a research institute, and attached to it a primary mechanical pump in the common two-stage configuration for high vacuum systems.
The lamp base has a glass tube going through it which is used to drain the air from the tube. With a small hand-held propane torch, Aleksander welds it to the business end of the vacuum system, and before opening the valve that would empty the tube, uses a high voltage power supply connected to the tube’s electrodes to establish a series of sparks into the glass bulb. Those are used to verify that the tube is actually being emptied of its air. After he opens the valve, the sparks gradually turn into a pale blue glow that fills almost the entire tube, before disappearing almost completely. It’s working! But we are not there yet. Electrodes can hold a significant quantity of gases that would slowly expand into the tube, damaging the vacuum and drastically shortening the tube’s life. To get those gases out, he heats the tube’s parts glowing hot while the vacuum pump keeps running.
The most straightforward electrode to degas is the cathode. He simply runs an electrical current though it, until it becomes glowing white. On the vacuum gauge, the pressure rises as the extreme heat releases the gases trapped in the tungsten filament.
Then comes the grid. Since we have a working cathode emitter, why not bombard it with electrons to increase its temperature? With the cathode still glowing, he applies several hundred volts to the grid. Again, the pressure rises temporarily, until the vacuum pumps permanently rid the tube of those unwanted trapped gases.
The anode is a little more complicated. Using electron bombardment is more difficult, because the cathode as a limited emission capability and the is much bigger than the grid and therefore difficult to heat. But Aleksander has built a portable induction heater, which he uses to make the anode red-hot in a few seconds.
He is now ready to remove the tube from the vacuum system. This is simpler than you might think – with the propane torch, he heats the draining glass tube, which melts and nicely seals the tube.
One last step is required before the triode is ready: flashing the getter. In no time, the induction heater makes the little ring containing the getter materials glowing red-hot. Chemical reactions inside it produce pure barium, which evaporates and condensates in a silvery deposit on the tube’s walls. Barium is a very reactive element, which captures substances such as residual air in the tube and helps prolonging the triode’s life by making the vacuum better.
All is needed now is to solder a socket to the base of the triode, and use it to make (for example) a regenerative radio receiver!
I would like to thank Aleksander for this awesome visit. Keep up the good stuff!
The Milkymist One is an open hardware live video synthesizer. Connect a camera and a videoprojector, press the power button, and seconds later, everything you film becomes live psychedelic effects of color and light. Point the camera at a dancer on stage, at people attending your party, at toys, use UV-glow paint… there are no limits to creativity! If no camera setup is available, the Milkymist One can produce purely generative effects which react to the ambient sound, making it an ideal option for music bands, clubs and party organizers who want a turnkey solution for simple visual effects. VJ’s will appreciate the possibility to design new visual effects using the Flickernoise Patching (FNP) language, which allows you to create unique and personal shows without requiring extensive experience with computer programming. And if you do program computers, you will certainly like the fact that the complete Milkymist One design is open source.
Arduino is an open source electronics prototyping platform based on flexible, easy-to-use hardware and software. It’s intended for artists, designers, hobbyists, and anyone interested in creating interactive objects or environments. It became very popular and people have developed an immense variety of sensory interfaces for the Arduino in the form of “shields”, along with software making them easy to use.
Join us in Berlin to enjoy the best of these two worlds. Bring your Arduino (and its shields), laptop (for programming the Arduino) and Milkymist One video synthesizer, and we will teach you how you can make the Milkymist One and the Arduino talk together to create interactive video installations. If you do not have those items, there will be a limited quantity available for lending at the workshop.
Some ideas of sensor shields to bring? Accelerometers, capacitive input sensors, potentiometers, joysticks, pressure sensors, Arduinome, and more! We could also help you assemble your own shield out of junk electronics (bring some!) at the workshop, if you want.
This workshop is kindly hosted by Betahaus/Planet Modulor, as part of the People in Beta festival lasting the whole day. You are invited to take a look around.
Sébastien Bourdeauducq is an inventor passionate about science, electronics and open source. After working for several small companies – which included developing the Wi-Fi driver infrastructure for the Nabaztag/tag – he founded Milkymist in the summer of 2007. The project combined his interest for the world of music with the desire to learn about and open up system-on-chip design, and it has now grown into a full-fledged open source project and commercial venture. Since 2011, he is also providing electronics engineering services for the CERN’s open hardware repository. Sébastien holds an engineering degree from Supélec and a MSc in SoC design from KTH. His hobbies include traveling, urbex and hackerspaces.
Fabienne “fbz” Serriere is a hardware person who likes spreading open hardware through hands-on workshops. She has worked on realtime low latency controllers for artists and musicians, massively multichannel audio, hardware reverse engineering, and open source hardware. In the past she has worked on writing hardware how-to’s for popular online publications. She organizes a yearly hands-on event for hardware called hardhack (http://hardhack.org). She documents her work on her personal blog (http://fabienne.us).
Date: October 1st, 2011
Price: free, just drop in
Duration: 4 hours (but feel free to come and go)
Location: Planet Modulor (very close to Betahaus), Prinzenstr. 85, 10969 Berlin
Please repost & share!