EEVblog #836 - Tour Of The Australian Synchrotron

[gadget73]

Particle accelerators are neat stuff.  The company I work for runs a couple.  One facility has an old Lineac in service that is used largely to color gemstones and do some sort of processing on silicon wafers.  They also have a thing called a Rhototron, which is like a much smaller version of a Cyclotron.  It has a couple of beam paths, one is strictly electron beam, two of the others aim at a tantalum target to produce X-ray.  Its used for commercial irradiation of things.  The cavity of the unit is a circular enclosure, probably about 5 feet across,  surrounded by magnets to bend and accelerate the beam to the proper levels.  It uses some large vacuum tube as the RF source as well. 

[jancumps]

Dexterslab did a tour of the LHC a while ago, might be worth a look at his video.

The only reference to "Dexter's Lab" I can find is some goofy cartoon.
I presume you are referring to something serious?

dexterslab is one of our fellow forum members.

[SeanB]

Dexterslab did a tour of the LHC a while ago, might be worth a look at his video.

The only reference to "Dexter's Lab" I can find is some goofy cartoon.
I presume you are referring to something serious?

dexterslab is one of our fellow forum members.

https://www.eevblog.com/forum/profile/?u=83

http://dexterslab2013.blogspot.co.uk/

[JoeLaBidouille]

Pretty cool video Dave.

Not so far from the CERN, there is also a synchrotron in Grenoble (French alps). It works on same principles and this big scientific tool is shared between multiple E-U countries
The ring is longer with 844 meters of circumference compared to 214 meters for the Australian one.
https://en.wikipedia.org/wiki/European_Synchrotron_Radiation_Facility

I attach some picture from my last visit (all visit are free and guided, this is a real enjoyment)


This synchrotron is only few hundreds of meter from an other toy called ILL Neutrons for science. This is a beam source of neutron, the most intense in the world.
Thunderf00t used to work there:

[Richard Crowley]

I am assuming that big light-tone metal cast bracket is aluminum (for its non-magnetic properties).
But just at the end of that bracket there appears to be an ordinary steel-strap, the kind used for packing and shipping.
How do they keep that conductive (and ferrous/magnetic) loop from becoming a secondary transformer winding?

[Yansi]

Kystron... whats that? I know only klystron. Misstype or is it anything else? https://en.wikipedia.org/wiki/Klystron

The proffesor was very good at responding to Dave's biting questions. A smart man he is!

Nice and interesting video, thank you.

And is also interesting those two guys really look they know every cable and machine in their facility. Very nice!

[gadget73]

I am assuming that big light-tone metal cast bracket is aluminum (for its non-magnetic properties).
But just at the end of that bracket there appears to be an ordinary steel-strap, the kind used for packing and shipping.
How do they keep that conductive (and ferrous/magnetic) loop from becoming a secondary transformer winding?

Likely a non-magnetic stainless steel alloy if I had to guess.

[Red Squirrel]

I was thinking, could that "storage ring" concept actually be used to store significant amounts of energy, like say for power generation?  You have a solar power plant or other plant that runs during the day, then extra power is stored to be used throughout the night.   Or is this really only limited to small particle energy?  Of course you would need a way to convert it back to electricity as well.

[Zad]

I only have one question....where were they hiding the flux capacitor? 

The storage ring is one huge flux capacitor!

Lets see ya get that rig up to 88 mph!

[Richard Crowley]

I was thinking, could that "storage ring" concept actually be used to store significant amounts of energy, like say for power generation?  You have a solar power plant or other plant that runs during the day, then extra power is stored to be used throughout the night.   Or is this really only limited to small particle energy?  Of course you would need a way to convert it back to electricity as well.

Even if you had a 100% efficient way of converting the energy back into commercial-grade power, the whole thing seems horribly inefficient as a "storage" device.
IIRC they said the beam was 1MeV (?) @ 200mA = 200KW  But then they said it took 4MW of power. Even if only half of that power were used directly for the storage loop, the cost/benefit ratio is hugely negative.  Which is why we are still using electro-chemical gadgets like batteries, and electro-mechanical gadgets like flywheels, etc.

[g.lewarne]

I was thinking, could that "storage ring" concept actually be used to store significant amounts of energy, like say for power generation?  You have a solar power plant or other plant that runs during the day, then extra power is stored to be used throughout the night.   Or is this really only limited to small particle energy?  Of course you would need a way to convert it back to electricity as well.

Even if you had a 100% efficient way of converting the energy back into commercial-grade power, the whole thing seems horribly inefficient as a "storage" device.
IIRC they said the beam was 1MeV (?) @ 200mA = 200KW  But then they said it took 4MW of power. Even if only half of that power were used directly for the storage loop, the cost/benefit ratio is hugely negative.  Which is why we are still using electro-chemical gadgets like batteries, and electro-mechanical gadgets like flywheels, etc.

Use my new, upcoming "Beamerizer" and make it 800% more efficient!



But in all seriousness, that was a cool video dave, thankyou

[Red Squirrel]

I was thinking, could that "storage ring" concept actually be used to store significant amounts of energy, like say for power generation?  You have a solar power plant or other plant that runs during the day, then extra power is stored to be used throughout the night.   Or is this really only limited to small particle energy?  Of course you would need a way to convert it back to electricity as well.

Even if you had a 100% efficient way of converting the energy back into commercial-grade power, the whole thing seems horribly inefficient as a "storage" device.
IIRC they said the beam was 1MeV (?) @ 200mA = 200KW  But then they said it took 4MW of power. Even if only half of that power were used directly for the storage loop, the cost/benefit ratio is hugely negative.  Which is why we are still using electro-chemical gadgets like batteries, and electro-mechanical gadgets like flywheels, etc.

Yeah was thinking that too, I guess there is probably not a more efficient way to put the energy in... we'll have to wait till the batterizer comes out I guess and daisy chain a bunch together.

[ECEdesign]

Am I the only one that noticed the copper strips around the walls?  I assume that is their ground bus 

[Steffen]

Exactly, the copper strips are there for grounding the concrete blocks as well as devices connected to it.  But it is more for potential equalisation. Usually the real grounding (according to VDE) goes to special joints in the building structures which connects to the metal mesh in the concrete structures.

[German_EE]

Grounding seems to be something that the Germans are very good at. Yesterday I was standing in front of TK Max waiting for SWMBO and I noticed that someone had fixed a ground strap to the metal window frame of the shop. The chances of that frame getting energized are tiny but someone grounded it anyway.

[Richard Crowley]

How do they detect the electron "bunches" going around the loop? I presume they need to know where they are so that they can synchronize the operation of the gadgets that keep the energy "topped off".  And to synchronize the "pump" electromagnets, etc.  But how do they do that without putting something into the path and presumably attenuating the energy?

[SeanB]

Very easy to detect high energy electrons, especially as you are bending them in a magnetic field. Simplest is a silicon photodiode to detect the rather large radiation pulse, or simply use a probe to detect the RF radiation as it passes by. Then you use some digital delays so the next time they come past ( you know how long it takes to go around) the acceleration energy is just in the right phase to pump up the energy. If your pulses are all a known length and timing you skew the pump pulses so that every x number of revolutions through the same pulse will get pumped, but in the interim your pumping will have done them all.

As well remember that the experiment tapping points are in line with the beam bending magnets, the beam itself does not lose much energy in a straight line, only when it turns a bend in a magnetic field. Thus there is a single pulse the length of the beam as it passes, and as it is a magnetic field I would also guess it is polarised somewhat and not a random polarised beam, but somewhat coherent phase wise with a very sharp rise and fall time.

[jahonen]

How do they detect the electron "bunches" going around the loop? I presume they need to know where they are so that they can synchronize the operation of the gadgets that keep the energy "topped off".  And to synchronize the "pump" electromagnets, etc.  But how do they do that without putting something into the path and presumably attenuating the energy?

Probably by same manner than in LHC, using a beam current transformer, which is a toroidal transformer arranged so that the beam travels through the center of the toroid. Then, the secondary current gives number of bunches, and if the transformer is fast enough, then it can sense the current contribution of each individual bunch.

For any storage ring or synchrotron needs very precise RF system to hold the bunches in so called RF-buckets (there is a phase loop which measures the beam and adjusts the RF), and that also means that RF frequency is set to such value that the RF wavelength is a exact multiple of the ring circumference length. That system give all the timings for all those who need it.

There is a description of BCT systems in LHC design report: https://edms.cern.ch/ui/file/445861/6/Vol_1_Chapter_13.pdf

BTW, those dipole magnets have pretty much constant field strength, so having shorted turns is probably not an issue, since magnet current is essentially DC.

Regards,
Janne

[daqq]

Nice video Dave, thanks!

Now... where's the teardown?

[SeanB]

Nice video Dave, thanks!

Now... where's the teardown?

I think they confiscated his screwdriver at the entrance...............

[electrophiliate]

Interesting hardware! Regarding efficiency, that is difficult to guess. I assume the 200mA is referring to the total current in the storage ring. I had a quick look again and don't feel like watching the whole video again, but my impression was that the initial 90kV (electron gun) is boosted to 100MeV using RF amplifiers (peak power 35MW, 10uS pulses, low duty cycle), then the booster ring increases that to 3GeV. The display was showing an injection top up current on the order of a few uA at most. The total average power for the whole facility was 4MW.

[Richard Crowley]

Doh!  I forgot that simply bending/turning the beam "accelerates" it. So, right at the bend you can detect the bunches quite simply without interfering with the main beam.

[TinkerFan]

I really enjoyed that video. This stuff is really fascinating, especially the variety of things they use it for. I actually did my work experience last year at the company that built (parts) of that Synchroton (FMB Berlin) and I looked every morning at a piece of a beamline from the Australian Synchroton. And guess what, the vacuum pumps they use while the thing is operating are produced by - Agilent!
http://www.agilent.com/en-us/products/vacuum-technologies/high-vacuum-pumps#1
I think they appeared somewhere in the video, but I cannot find them anympore.

But what I didn't understand, is how they produce the diffrent wavelengths. Do they insert the 'light' at diffrent wavelengths or do they 'divide' the wave in the beamlines?

Now after 2 years of silent watching I finally arrived in this Forum...

[dexters_lab]

Dexterslab did a tour of the LHC a while ago, might be worth a look at his video.

The only reference to "Dexter's Lab" I can find is some goofy cartoon.
I presume you are referring to something serious?

dexterslab is one of our fellow forum members.

https://www.eevblog.com/forum/profile/?u=83

http://dexterslab2013.blogspot.co.uk/

did someone mention me? 

the cern openday was awesome, there is supposed to be another long shutdown in the next year or so... hopefully they will run another openday program when it happens, Dave should make an effort :-)

been putting off watching this vid as it's an hour long so needed to find the right time to watch it!

[lpickup]

Wow, that was a great video!  Thanks Dave.

This is not my field, and it's been quite awhile since college physics, so I may have some really basic questions, but I would really appreciate learning more about this.

1)  What is the ultimate purpose of using this giant machine to generate essentially IR, visible light, UV and X-rays, as opposed to say using an X-ray machine at the dentist's?  Is it essentially the fact that you can get a much more energetic ("brighter") output?  And that higher energy is necessary to perform the type of experiments that are being done there?

2)  So I followed what they were saying that the beam attenuates (loses energy) due to non perfect vacuum, and more importantly due to the bending acceleration of the beam.  So would that mean that a larger ring, with more subtle bends, would attenuate less?  Or does it all come out in the wash because ultimately the beam has to bend  a full 360 degrees anyway?  (and thus the non-vacuum source of the attenuation would actually be greater in the larger ring)

3) Did I understand the beam timings correctly?  The pulses are at a 2GHz frequency and each pulse is 23ps long?  So in other words, a new electron bunch comes along every 500ps and are 23ps in "length" (giving a gap of 477ps?) 

3a)  In the scope display at 49:05, my understanding is that it was displaying a single beam measurement device, and each of the large "blobs" represented the electron bunches passing by the detector.  But what is the gap?  Is there an intentional "quiet" part of the beam with no bunches?  Is this where new bunches are injected into the beam?

4) So when the main beam is tapped off for one of the experiments, I assume that one (or more) bunches of electrons are redirected off the main ring and into the experiment's beam line.  Presumably these bunches would have to be replaced by new bunches to maintain the total ring current of 200mA.  How is this managed?  Are bunches stripped from the beginning (or end) of the pulse train and then injected at the opposite?

Fascinating stuff!