<fltrz>
SpeedEvil, thats so counterintuitive :) is there an article that delves into and accentuates this result?
<fltrz>
SpeedEvil, or does that condition prevent heat conduction, or does it prevent convection?
<Degi>
So thermal resistance goes high when two surfaces get near each other but don't touch? Kinda like paschens law?
<SpeedEvil>
fltrz: broadly - if the distance is >1 mean free path, all that happens is that air molecules bounce off air molecules, and the overall conduction doesn't change much, as the total velocity of conduction does not change much
<SpeedEvil>
It's only once you start to get to 1 mean free path spacing that you start reducing the conduction, as now the thermal conduction starts to drop linearly with pressure.
<fltrz>
SpeedEvil, but if the walls are closer the molecules bounce with high velocity from hot wall and with lower velocities from cold wall?
<SpeedEvil>
Bringing the walls closer together at a given pressure does indeed increase thermal condictuvity
<fltrz>
SpeedEvil, I respect your knowledge, it's just that I'm better understanding something when its written out in paper form, so I was hoping there was a relevant counterintuitive puzzle paper?
<fltrz>
oh, ok so then I misunderstood
<SpeedEvil>
The pressure vs thermal conductivity graph has a major bend at the MFP=spacing
<fltrz>
SpeedEvil, it a totally unrelated question, but can you recommend a good text on calculating thermal and shot noises in "arbitrary" circuits?
<SpeedEvil>
Alas not.
<SpeedEvil>
I have mostly been pretending that electrons don't exist and it's all water so far.
<fltrz>
I had this app note from on of the major opamp MFRs, which seemed to explain quite deeply, but some of the steps seemed off or unfollowable when they applied it to some simple opamp circuits
<fltrz>
lol
<SpeedEvil>
It works for a surprising number of things :)
<fltrz>
SpeedEvil, would you be interested like I am? I am looking for perhaps the equivalent of network theory to do it manually / analytically, or alternatively a monte carlo method so the noise can be simulated for arbitrary circuit...
<SpeedEvil>
Sorry - utterly exhausted ATM and no energy. Trying to concentrate on getting this place watertight/... ATM
<fltrz>
whenever you end up trying to calculate feasibility I find I need such a thing
<fltrz>
your place is flooding? or hermetically sealing from coronavirus?
<SpeedEvil>
general house DIY - rain/roof
<nmz787>
fltrz: if you need a beam of gas, you already have most of the setup for a neutral atom microscope
<nmz787>
Degi: ben krasnow made his own SEM... so it's def possible... also a few good papers out there on much higher resolution methods of DIY scopes
<fltrz>
nmz787, the first paper on the SEM design, they start of with noting the need for atmospheric operation for biological samples, and then the rest of the paper doesn't talk about atmospheric operation?
<fltrz>
I am curious about the OPERA 3D code though, if its open source
<fltrz>
regarding the neutral atom beam, its exactly because fine apertures and optics seem hard with gas beams that I was designing a way for passive gas molecule collisions to be visualized instead of trying to generate a controlled beam of gas particles
<Degi>
Hm I found that for making approx 100 µm sized apertures, putting copper foil onto a glass plate and then hammering in a needle works.
<fltrz>
Degi, they use a 300nm aperture
<fltrz>
hmm, I wonder if I can use my original idea to measure the surface collision density optically in atmosphere, while having an acoustic diffraction grating spray a spectrum of pseudorandom pressure oscillations so one dimension of the 2D sample is encoded by filering the optically measured collision density fluctuations by frequency... that still leaves the second dimension... perhaps a second beam of accoustic waves with other wavelength band,
<fltrz>
orthogonally to the first?
<fltrz>
i.e. for pixel position X=10,Y=21 correlate over time the optical collision signal with [accoustic signal of X=10 * accoustic signal of Y=21]
<fltrz>
I should check that the coding scheme works regardless of if the physics works though... I never saw correlation using 3 signals... its interesting if that could be extended to higher dimensions...
XgF has quit [Ping timeout: 265 seconds]
XgF has joined #homecmos
<nmz787>
fltrz: you need less than high vacuum for /some/ biological samples, it really isn't related to the electron gun... it's more differential pressure vacuuming of the chamber vs gun column.... i.e. you need a small aperture that the beam travels through, to limit diffusion into the gun, so the chamber can be at appreciably higher pressure
<nmz787>
fltrz: there really aren't any gas beam optics, you move the sample to scan the image AFAIK
<nmz787>
and aperture isn't really correct either, you want a skimmer (the funnel shaped thing)
<fltrz>
nmz787, oh, but on their diagram theres more pumps for the sample chamber which is why I thought it was harder vacuum
<fltrz>
oh were talking about the other paper
<fltrz>
nmz787, so the SEM can work at atmospheric pressure with a skimmer?
<fltrz>
how much energy would a plasma window consume, and could a SEM work through that?
<fltrz>
hmm, still would prefer some optical method if it could be made fast
<nmz787>
umm, I think we've confused papers a few times
<nmz787>
the skimmer is only for the neutral atom microscope
<fltrz>
nmz787, the weird thing is the SEM paper also mentions desirability of atmospheric operation, but then leaves this apparent goal dangling for the rest of the paper?
<nmz787>
a SEM basically doesn't work at atmosphere... unless the path from an aperture that's containing the high-vac is very very close to the sample... but then you basically just have an STM
<fltrz>
right
<nmz787>
well because environmental SEM has nothing to do with gun design
<nmz787>
and that paper is about gun design
<fltrz>
yeah
<nmz787>
the more gas there is, the more chances for the electrons to interact with it, and scatter... thus reducing your imaging resolution
<nmz787>
so environmental SEMs make this tradeoff carefully... you want better than optical resolution, but also don't want to dry your sample too much because it will start mechanically deforming
<nmz787>
and thus won't be true to nature, as much
<fltrz>
nmz787, yea its in their conclusion: "The environments of the processing, such as the vacuum, the impurity concentration, etc., should be taken into consideration in the choice of the type of focusing lens. For a low vacuum process, a magnetostatic lens is preferred because electric arcs are likely to occur in an electrostatic lens at low vacuum. For other cases, however, an electrostatic lens seems to have many advantages owing to its compact size."
<fltrz>
so the contribution of the paper is about a compact column for higher pressures
<nmz787>
well that's also likely only important for the final lens... the objective lense
<nmz787>
(Assuming you could enter the mass and charge for your "ion")
<fltrz>
nmz787, cool thanks
<fltrz>
btw, theres a new Applied Science video out, where Ben shows how he uses antqiue large format camera with ortho litho film and gets great results
<fltrz>
nmz787, perhaps its these lines? e1=17000.#SO110-EL
<fltrz>
e2 = 23502.*.9523*INPUT[7] #BI-EL
<fltrz>
if that represents energy then it also represents mass for these CERN SI unit droppers
<fltrz>
hmm probably not given they add the input parameter...
<nmz787>
anyone know what a reasonable price for an RF match network might cost, for like 1kW power rating? an eBay seller is offering $850 shipped for the match network, plus two controllers, plus a vacuum chamber with (presumably) ICP coil wrapped around it
<fltrz>
no clue
<fltrz>
nmz787, this is for impedance matching?
<fltrz>
I have always wanted to try taking N parallel coax cables, put them in series on one side (i.e. one cables shield attached to next ones conductor etc), and put them all in parallel on the other side (all shields together, all conductors together))
<fltrz>
not sure how much coax $800 buys... and probably really sloppy solution full of reflections...
<fltrz>
so impedance on left hand side is N*Z, and on right hand side its Z/N
<fltrz>
huh, perhaps I can try that for EDM lol
<nmz787>
yeah impedance matching
<fltrz>
nmz787, whats the impedance ratio?
<nmz787>
no idea
<nmz787>
I have a 1kW RF generator for 13.56MHz
<fltrz>
output port is what impedance?
<fltrz>
what are you building?
<nmz787>
plasma vacuum chamber
<fltrz>
how is the EM energy applied? electrodes in partial vacuum? or external electrodes through glass?
<fltrz>
I worked with a ballast when generating plasma, but from DC... plasma is essential conductor
<fltrz>
or are you using a waveguide into the chamber?