Of course, I could be wrong...

We can't get the beam to light. There is this tube up inside the tank (the seven ton steel shell around the accelerator body,) that uses an RF coil to ionize the gas so it then goes down the 4MV gradient and speed toward the target. Because everything up there is at 4 million volts (or 1 million because that's what we're trying to run it at,) there is no actual connection between those controls and the knobs on the console. There is a pair of linked 'selsyn' motors that talk back and forth with each-other, then a big long Lucite rod that then actually turns the variac controller up inside the tank. Sadly... something seems broken. And its looking more and more like we have to pull off the tank.... which we may not be able to do at all.

I love it when a plan comes together.

Or at least when you are trying to fix things and actually succeed. New power supply was there when I showed up Tuesday morning.

Tuesday: Soldered on new leads as it is much smaller than the one it replaces. Hook it up, turn on the power to the board on and presto! As expected, we needed to re-tweak the offsets on the op-amps but having practiced the procedures and not having to locate malfunctions it went smoothly. We had the thing back in, the lab cleaned up (like any lab not involving germs or hazardous chemicals it had got a bit dirty,) and ready to try out the next day.

Wednesday: Tried to fire the beast up. The new board works perfectly, and even corrected some of the anomalous meter reading we were getting (like the fact that we were reading 0.3MV on the charge plates before we powered anything up.) We get a charge, the stabilizer works exactly the way its supposed to. We dump some hydrogen gas into the beam-line and crank the ionizer up and... nothing. No beam. The issue seems to be with the ionizer rather than the gas, but its hard to tell and we ran out of time before we could test further.

Thursday: Since the other ongoing project is the relocation of all the labs from the Physical Sciences building to the Biological Sciences building (which used to be the Physical Sciences building back in the old days apparently,) we're going through all the stuff in the lab and packing it for transport. I was looking through the stuff in one cabinet, trying to identify various materiel samples in jars and envelopes. Open the thing and look at the hand-written note inside, which read "The white stuff is calomel (Hg_2 Cl_2) and the red stuff is Hg_2 I_2." These chemicals were contained in a small glass tube which had broken and a quantity of the white stuff was now on my hand and clothing. So now I have a bunch of mercurous chloride on me. Thankfully it washes off and doesn't go into your skin, so it isn't a significant exposure, but ick!

Increment

New part arrived, so the plan is to install it, make a few adjustments and try to fire the Beast up. I give it better than even odds we have it sorted out by tomorrow afternoon.

The story so far.

We are busily re-rehabilitating a 4MeV VanDeGraff particle accelerator. The energies are too low to do any interesting work in nuclear or particle physics, but perfect for irradiating crystals or semiconductors to see what happens.

I'm coming in partway through the project, and the initial baselines readings on our samples have been taken and analyzed. Now, we have to bombard then with 1MeV protons. Which means we need to have a source of said protons, and thus the accelerator.

This machine was built in the mid 1970s and last brought into operation in the mid-90s; there is a spring that flows into the basement where the two-story tall beast lives so the humidity is rather high; someone actually sabotaged it at one point in a number of subtle and difficult to find ways. Last week, we'd managed to get some function from it, but the moment we tried to switch the voltage stabilizer on, it went haywire and started pulling a huge current off the accelerator plates. We shut it down and dug the manual out, and there was this maintenance procedure for the control circuitry that we resolved to get into on Monday.

Monday: The procedure pretty much consists of iterations of: Ground a few circuits; Connect a meter to this other spot; Adjust a trim-pot until it reads zero. We start the first adjustment and we read about -9.5V, and can't adjust it more than a quarter volt away from that, and it needs to read zero. Problem!!! Easiest thing to do was replace the socketed op-amp, but no luck. We traced the thing back and the only thing we could find was that the power supply was all norky. We locate another power supply that seems to be the right sort, but it's in another device and has solder all over the pins. Desoldering wick is located, but there isn't an iron to be found. We decide to pick it up the next morning and I'll bring my kit in.

Tuesday: Solder removed, power supply module in place and... its only 10V not 15. Still, the op-amp adjustment is in the right range and I can zero it, so it looks like the theory was sound. We had another supply, so we pulled it out, had to solder on new leads and... right voltage but not enough power to drive the circuits.

Today: We bring in an adjustable power supply down from one of the labs and start working our way through the procedure and we locate a bad amp. Thankfully we have a spare and are able to finish everything up to the point where the adjustments require a beam. We're ordering a new supply and should have it when we get back into it on Tuesday.

I suspect we'll need to go in and re-tune everything with the new supply but it should go fast. Hopefully we can fire the beast up and start in on beam profile characterization.