I just disassembled and acid washed all the parts for the coil former and electron gun. This degreases and cleans the parts.
I used a 15% HCL solution.
The aluminum started bubbling, so I didn’t leave it in long. The Stainless and ceramic I left for 15 minutes.
The parts will dry overnight.
Got a Tektronix 2445 4 channel analog oscilloscope for sale:
Starting $1, no reserve.
Comes with 4 HP probes.
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=220747335752
I’ve decided to adjust my research by taking smaller steps more frequently.
Yesterday I installed the RGA to get a clean baseline for the vacuum system.
It didn’t detect much! Which is good news:
I also spent HOURS organizing the lab. Getting wires off the floor, shelving, etc:
And got a stethoscope for future vacuum diagnostics:
As readers have pointed out, it’s time to put a lid on the power supply. Took care of that today:
I also added a fuse for the AC side:
Added a chassis ground for all surfaces:
Insulated this diode which could have shorted the bus bars:
Added a permanent voltage meter (steampunk):
Permanently attached bleed resistors to slowly drain the capacitor bank by default.
I wanted to look at the current flowing into the coils from the coil powersupply.
I passed the coil current through this ammeter shunt resistor. 100 mV = 100 A.
With the capacitors triggered at 400 V, we see a > 1,200 A current spike lasting about 10 ms:
That’s some current!!!
Let the pump run all night.
The chamber with blanks got down to 98 nanotorr:
Pretty great actually.
Now we have a baseline best case performance for the vacuum system.
I found the fucking leak. It was the glass viewport. The passivation ate the metal between the glass and the steel:
For the record: do not passivate glass viewports.
I found the leak by listening. Late last night the building was quiet enough to hear the leak. I didn’t hear a hissing so much as I heard the sound of the pump emanating from the surface of the glass.
NOW THE EXPERIMENT CAN PROCEED!!
I will let the pump run overnight to see how deep a vacuum we can get with just blanks.
Previously I though I had broken my pump.
So the pump is back from repairs. But the strange thing is when Pfeiffer received it, they said it was working fine! Although I did manage to knock the rotor out of balance and it needed fresh oil. Here is the pump capped with a blank conflat running at full speed. GOOD.
But the problem was never the pump! The problem is a gross leak in the vacuum chamber.
Once I attach the pump to the chamber, the pump won’t spin up… indicating a gross vacuum leak.
The copper gaskets are all new, and I even got the torque wrench out to make sure the conflats were tight enough.
I can’t hear any hissing.
How do I troubleshoot this…
Here is a good overview on leak detection.
UPDATE:
I though I head a hiss near the bellows. So I tested just the bellows and it’s working fine. So the leak is somewhere in the chamber.
This project has been an ongoing lesson in electrical engineering. Now with an oscilloscope I can finally see a circuit’s behavior.
I’ve spent the last week reviewing my assumptions about the most basic circuits and components. Sometime my hunches are correct, but just as often I am confounded by what I see. Here is my test setup:
The most interesting behavior happens with an AC signal. Conveniently the oscilloscope has a built in square wave generator intended for calibration. I am passing this square wave through test circuits to see how the wave changes.
The oscilloscope generates a square wave that goes from ground to +0.4 V. The frequency ranges from 50 Hz to 5 MHz depending on the time setting. I start with 50 KHz. I use two probes. The first probe is connected to the signal source, the second probe is connected to various other points in the test circuit.
The first thing to note is the signal generator doesn’t provide much power. If you overload the signal generator, you will see it’s voltage sag.
Both probes ground to the oscilloscope chassis, so choosing appropriate ground points is crucial … incorrectly grounding a probe can drastically change the circuits behavior.
I started by looking at a single capacitor. I tested this circuit:
In this capacitive coupling configuration a capacitor removes the DC component from an AC signal. Probe 2 shows the same signal as Probe 1, except shifted down. Probe 1 goes from 0 V to + 0.4V whereas probe 2 goes from -0.2 V to +0.2 V. So that’s what it looks like to block the DC component.
Next I passed the square wave through a transformer. I’m using a variable resistor to limit the current into the transformer.
This is what I see:
signal generator at top. transformer output bottom
It doubles the voltage as expected, and adds quite a bit of color to the waveform. It also draws enough current to make the signal generator’s voltage sag… even the ground line as seen in this video:
 |
| From 2011-02-12 |
At some frequencies, the transformer really changes with waveform:
signal generator at top. transformer output bottom
In the US the mains power runs at 115V, 60 Hz. With an oscilloscope, I can see this mains hum everywhere. It radiates from the AC power lines throughout the lab and the whole city. Appearently our bodies make great antaneas for 60 hz: touching the oscilloscope probe shows a 9V RMS @ 60 Hz. Like a nine volt battery! From my body!
I also noticed that the using the probe clip picks up more mains hum than just the probe tip.
I’m learning how to use the scope. I suspect it will become my main measurement tool.
Today I received a Tektronix p6109 general purpose probe compatible with my scope:
Prometheus Fusion Perfection 