Solving Noise Issues

I repeated the tests described in last post to confirm electron detection. The graphs below show electron well results with about 11kAT (kilo-amp-turns) on the polywell coils. The top line of the oscilloscope shows the initial pulse of the high current discharge, while the bottom line shows floating electron potential from the langmuir probe.

Test at 11kAT
Another test at 11kAT

There is considerable noise in the langmuir probe reading in these graphs. Additionally, the oscilloscope reading makes it seem like the electron well sustains itself beyond the high current pulse. Really, the oscilloscope is not picking up the full duration of the high current pulse — it should be longer than 1 microsecond (the timescale of the graphs above). The electron well should be sustained throughout the high current pulse for 10s of milliseconds. This brings me to another key point: after doing these trials I realized the timescale should’ve been zoomed out to a millisecond scale. The noise in these graphs may appear significant due to the small timescale.

However, not all noise was reduced by zooming out the timescale. High current pulses powering the coils induced noise in the langmuir probe channel of the oscilloscope. This was an issue when trying to resolve electron well potentials. I took apart the circuit to look closer at the causes of this problem, and realized some circuit components were not correctly isolated. I ended up remounting the high current pulse circuit on a control panel shelf.

Rewired high current discharge circuit
Zoomed out high current discharge on 50 microsecond scale. Bottom channel (langmuir probe) picks up induced voltage temporarily despite no electron well.
Electrical feedthrough and oscilloscope connections

With the circuit rewired and adjustment of oscilloscope settings I was able to resolve the high current pulse better. The graph below indicates the current discharge lasts about 7 milliseconds at peak intensity, and 15 milliseconds overall. Here the filaments were not turned on, so the bottom channel shows no electron well. This looks much better!

Oscilloscope graph of high current pulse

Unfortunately, accidental shorting of a power resistor caused a large current into the filament circuit, blowing out a few filaments. Despite getting the oscilloscope right, I could not get anymore electron well results. I opened up the chamber to take a look.

Filaments shorted and run at too high pressure (oxide coating)
Iridescent tungsten deposition from filaments on chamber wall
Filament sputtering on polywell and chamber walls

Leave a Reply

Your email address will not be published. Required fields are marked *