Remote viewing of fusor in realtime with raspi/cam

This is somewhat of an admission of failure. You can't easily pigeon-hole everything, and most real projects use commercial software, homebrew, and hardware all at once. So, for you makers out there (including me) - this is where to put whole projects that don't fit well in the other forums.

Re: Remote viewing of fusor in realtime with raspi/cam

Postby Doug Coulter » Wed Nov 12, 2014 5:04 pm

I need to grab just one window out of a 4 monitor huge system, and an audio input of choice (I was using a USB headset for its mic). ffmpeg as a CLI isn't even available anymore for Mint-mate 17, it's just a .so used by a bunch of things.
the problem with record my screen appears to be that it does record the whole screen, and by some trickery only marks the window you want "playable" which FWIW, drives my video editor crazy - it's not meant to handle multiple vid streams in one container. Kdenlive can split audio from video so I can use audacity to notch out the hum no problem. But then it can't seem to get the right vid stream added back to the new audio...It might wind up being a case of getting into the source code of record my screen (name or program is actually all one word) and fixing it there. Of as you say, finding another way to do this entirely.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: Remote viewing of fusor in realtime with raspi/cam

Postby Doug Coulter » Mon Nov 17, 2014 1:32 am

Oh, the real name is recordmydesktop. Oops. Still has the hum, and another bug. I split out the audio with kdenlive, good so far, and yup, him on the fft, a pure sine. But in putting it back, kdenlive messes up. It appears the capture was somehow the entire screen, but as different streams in the container file, and kdenlive then uses the wrong one...bad day at the office...

I had to re-edit this post twice, finally on something other than my nexus 10 running kitkat, due to it's insane auto-correct changing hum to him and recordmydesktop to record my desktop *after* those words went off-screen...worth it to turn on a real computer (this nuc i5) even on a bad power day. Gheez what a load of junk that autocorrect stuff is, even the voice recog does better most of the time, but both provide, I suppose, "plausible deniability" that you said what you said. What a steaming heap.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: Remote viewing of fusor in realtime with raspi/cam

Postby Doug Coulter » Thu Nov 20, 2014 3:13 pm

Thought I'd add a few related pix of the pi cam setup here. I improve the optic a big (Thanks BillF for finding the glass) and locked things down better, while adding additional X ray attenuation. Not that the latter matters, it was already so good most of our X rays were actually neutron capture-gammas from all around, not from the fusor directly, but hey, it's nice to be complete. So, here's what it looks like at the camera mount.
PiCamMount.JPG
Here's the pi cam looking through a 2.5 diopter lens into the tank from above, with lead covering the unused part of the viewport. The extra IR Filter is just stuck on with black tape, and some hot glue.

Baling wire! It works, and is nicely adjustable with pliers. Here's the view it sees (more or less, I re-aimed it a little I think).
View.jpg
A little green from the extra IR block, but good anyway.

You can see my main grid, and at the lower right, the ion grid, which is pretty gross as a fusor grid, but it works fine for making ions when the gas pressure is so low the main grid won't "light off" at all. In fact, you could think of all this as a very high voltage enhancement mode p-fet - it takes 5-40kv on the ion grid (depending on pressure), and from 1 to about 12 ma to bias the main grid "on" at 50kv and any current I want - the limit isn't the power supply here, it's what would melt. Power gain is around 100 used as a "fet" of "reverse polarity vacuum tube". But transit times are very long compared to an electron tube.
I made a lead hat for all this so the X rays from there don't leak out into the room where the air scatters them around the shield between this and the operator position.
RightSidePi.JPG
Showing the "hat" and also the pi itself in it's little box.

This also shows the pi in the plastic box. In that box are also a laptop walwart (overkill) and a 5v 2a switcher (more overkill) to run the pi, with wires soldered right to it's board (not overkill, the microusb was a very stupid choice on their part).
The pi is running a high power wifi (2.4ghz) from adafruit on one of its USB ports. I got tired of running wires and switches, and this is safer for my network if something should pick up a big spike anyway. You don't actually need that little display, or even to log in for this to work. It is nice to see it if it crashes so you know why, sometimes, though. I suppose you could get that by SSH'ing into it, which does work fine too. All this stuff for the streaming via raspimjpeg and apache gets run at boot anyway, and runs if you surf to the pi's page.

Of course, it looks better when the fusor is really running. This one is cropped down to just the action here:
PiCamStill.jpg
My line focus is just to the right of one of the grid wires, but I need to rotate the grid so you can see it better next time I break vacuum.


So, this one's pretty much a done deal now. I don't plan on fooling with it much more unless I get into the source code for the thing and fix a few bugs it has with settings not all working as they should and so on. Mechanically, we're there except for a cable tie to hold the box on there no matter what. It's better than this in most regards. I can't seem to get my handheld camera to NOT focus on the screen just past the glass, and of course, there are reflections from the 3 layers of leaded glass, which is why I often take pix at night...this time not.
EndOn.JPG
End on with normal camera that likes to misfocus on the screen next to the glass.


The beams and focus are a lot smaller to the mark 1 eyeball than this shows, misfocus by the camera, not my fusor in this case.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: Remote viewing of fusor in realtime with raspi/cam

Postby Doug Coulter » Thu Nov 20, 2014 3:33 pm

And of course, it's not complete without a movie. Someone is going to have to fix the bug that won't let you set white balance off auto, even though the current software claims to do that. The colors didn't actually change at all, other than the ones due to the parts heating - the plasma is always purple.
http://youtu.be/64NkTcqjXqc


I'll have to dig into this some more.

Oh, I left out the picture of the cool stuff that helps block near-IR for the camera. Things still look a little hotter than they really are to the eye, but it's close enough that I can "see through" that.
PiOptics.JPG
The other stuff you need for this.


I carefully pried this interference filter off a welding goggle "instant dark" plate that had failed and was destined for the trash anyway. Looks green through it, but is a mirror if the angles are right - classic Fabray-Perot.

Here's a slightly better video, spliced from 3 taken during the test run, then audio commentary recorded while watching it, then splicing that in, with audacity and kdenlive. BTW, Donovan...I now have hum in ALSA in Audacity...You might have been correct, but I'm not sure the reason. I was able to use audacity to take most of it out in this, didn't bother to learn how to really kill it, but it's kinda OK here.
http://youtu.be/J-NVvIMeSIU
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Previous

Return to Combined projects

Who is online

Users browsing this forum: No registered users and 1 guest