Filament Tests Debunked

Shoutout to my homie Sam Johnson for coming by today and shooting video of the project!

Today I tested the filaments near a langmuir probe to see if I could pick up any signal of emitted electrons. I tested the filaments with a low resistance langmuir probe and got clear signal on the voltmeter of electron emission! The electron potential consistently gave voltages of -0.3 to -0.7 volts and peak at -1 volts. The filaments are DC driven as well, so their emission should not vary with negative and positive cycles of AC current.

The low resistance langmuir probe gave a clear reading compared to the high resistance probe I had originally mounted in the chamber. High resistance of the probe could’ve been a major issue in previous trials.

Mark Suppes, a former polywell experimentalist, helped answer my questions in the previous post about electron detection. I only need to connect the voltmeter directly to the probe with no follower circuit. This makes sense; the langmuir probe is essentially an extension of the voltmeter. It can be used give a voltage reading between the negative electrons in the chamber and ground. I thought this was the way to do it initially, but confused myself and complicated the problem after failing to detect anything.

Basic electron cloud measurement schematic. Langmuir probe is the extension of a voltmeter

Unfortunately, I could not translate my voltmeter readings to the oscilloscope. The oscilloscope gave consistent noise on the 200-500mV scale, yet it can normally resolve these voltages with minimal noise. I messed around with voltage and timescale settings with no success of dampening the noise. For a non-obvious reason, most of this noise went away when the electron filaments were powered on, although the voltage was still centered at the same vertical position of ~0 volts. When I turned the filaments off, a negative voltage peak of similar scale to the potentials measured earlier on the voltmeter appeared. Strange!

Langmuir probe reading on oscilloscope (time on x-axis and voltage on y-axis). Negative peak occurred when the filament electron source was turned off. Following this peak the oscilloscope gave an increase in noise with the filament supply off. Why?

Ahah! After some thinking I can explain the noise issue and odd oscilloscope readings. Grounding was the problem. The oscilloscope was plugged into a power strip with two roughing pumps that were pumping down the chamber. The pumps’ ground is likely connected to their exterior metal frame, and this is connected through metal flanges to the vacuum chamber. Thus, the oscilloscope is ultimately connected to the same ground as the vacuum chamber. To take measurements, I attached the oscilloscope ground cable to the electrically isolated langmuir probe, while the hot cable was connected to the chamber for a 0 volt reference. In reality, the oscilloscope cable connected to the chamber was completed an electrical connection all the back to the oscilloscope ground and picked up noise from the pump and chamber components it was traveling through!

Pumping system ground connects the reference/hot cable of the oscilloscope to the langmuir probe creating a closed circuit with noise.

This explains the noise when the filaments were turned off. Why was there less noise when the filaments were turned on?

When the filaments turn on, the electron cloud is sensed by the langmuir probe which connected to ground through the oscilloscope ground cable. However, this cable is also (unfortunately) linked to the other cable on the oscilloscope through the vacuum chamber ground. So voltage stays at 0 volts on the oscilloscope reading because both cables equally pick up the negative Langmuir reading. This also overrides noise from the chamber and pumping connections.

Lastly, why is the spike in voltage caused when turning off the filaments?

The reason for this is hard to thoroughly explain, but it is probably due to a decrease in electron cloud received by the langmuir probe. The oscilloscope cable directly connected to the langmuir probe senses this decrease faster than the other cable, giving a negative spike. When the chamber connected cable senses this shortly after a small positive bump is seen on the oscilloscope display.

Since the spike occurs when there is a change in electron density, why doesn’t this spike occur when the filaments are turned on too?

Hmm… The only cause I can think of is that electron density changes more rapidly when the filaments are turned on compared to when the filaments are turned off. The Langmuir probe was placed < 1 cm away from the electron emitting filaments for these tests, and the filaments take a longer time to power down than turn on.

To avoid all of these grounding issues I can ground the oscilloscope and pumps separately or take differential (two channel) measurements of langmuir probe and alternative ground reference. This is a tactic frequently employed for safety purposed when measuring high power supplies with an oscilloscope because it prevents shorting a low resistance ground. Basically, both channels of the oscilloscope are added together and one of them is inverted (causing subtraction) and only the hot cables of each channel are used.

To answer another question from yesterday’s post:

The iridescent coating surfaces surrounding the tungsten filaments occurs due to tungsten vapor deposition. Incandescent and halogen bulbs are filled with inert gases to reduce tungsten deposition on the walls of the bulb. When these filaments are placed in a vacuum, they have shorter lifetimes because tungsten deposition occurs at a faster rate.

A photo of tungsten deposition on aluminum foil that was placed near the filaments

Leave a Reply

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