Fusor remote hardwire safety

Stuff you made and how to make cool stuff

Fusor remote hardwire safety

Postby Doug Coulter » Thu Aug 25, 2016 5:06 pm

Almost....there.
This is part of a massive undertaking to be able to run the fusor remotely - and safely - in the new mode that is dangerous for a human (and some computers) to be near.
I will detail more of the data aq and computer based stuff elsewhere, but from experience - computers sometimes crash and leave things in funny states, particularly x86 - it seems the smaller the geometry of the bits, the more likely EMI or a neutron or gamma hit will take it down in some bad way (else I would be done with documenting/patenting the new mode - it caused my existing PC data aq to instantly crash from flux alone).

Knowing this...well, it seemed wise to have some kind of backup on the otherwise handy computer stuff, some way to for sure, and instantly, turn it off, shut it down, make it safe to go back in the building it is in.
As luck would have it, a friend had long ago donated some really nice telco burial cable - shielded-armored, silicone filled, 6 pairs, so I've buried that between the fusor building (shop) and the op position (my normal dev station).

Some things that look simple, aren't. I wanted to be absolutely sure that almost no matter what failed and which way, I could turn it off - make it fail into some safe mode if possible.
I don't yet have a drawn schematic (I worked from lists of pinouts) but I'll get one together and post it, it's not complex - the design did consume some skull sweat to make it come out simple, though.

We run two power supplies on our fusor. One, the main high voltage, is a Spellman Sl2Kw (which is way overkill, but very very nice). 50kv at up to 40ma, though we rarely have used more than around 16 ma and that is an extreme case, usually we're down closer to 10 or under - at 50kv for about 500w input. More doesn't seem to really help, less is often better in our setup, and heat removal is less when there's less input.
We also run another, smaller Spellman, a ptv200n, which is modified to up to 12.5 ma at up to 40kv for our ion source. It is usually run around 20kv and around 10 ma or less. We have had no issues with the ion source making enough fusion to be really dangerous (tops it's about 1 million neutrons/second, kind of like fusor.net stuff).

I wanted to be sure I could shut this down remotely, no questions, no problems. So, one of the pairs in that buried cable is power from my end to be "coil power" for a pair of relays that switch 240VAC for the big Spellman. Power has to come from here - and any open circuit or short to ground will open the relays. Even if one sticks - we have two. That would be the left red covered switch on the panel (which BTW also has a led you can't see in the picture as it's more or less invisible unless lit). The SL2Kw also has a safety interlock and return, which gets another pair in the telco wire. That's the second switch and led. Either switch open kills main power, and again, an open circuit or a short of any kind works out to "it's off".

Now, there's a lot of wires left, and I'm Scottish, so of course I figured out some other useful things to do with all but one of them (keep a spare, you never know). I have a Blue Yeti microphone over there that the remote data aq already records via USB during a run, but why not have a zero=-latency audio feed back to the op position? There's a couple pairs, and that particular mic provides a nice low impedance high power headphone jack as well as the USB interface...
To use that at this end, I got one of AdaFruit's little class D audio amps and some tiny speakers from Marlin P Jones, so I can hear what's going on. As an old audio-phool, I made sure the imaging is good enough that I can kinda tell where some noise is coming from as well - those tiny speakers are surprisingly good for this..

As I had the wires, I added some signals back from the power supplies, for example their "I'm on and making power" signals, and the "clapboard" signal from the main raspberry pi over there that indicates the time sync at the beginning of the run.
All of the data aq slave computers see this signal when I hit go on the software GUI so their time stamps can be synced together for later teasing apart of causes and effects. It's not perfect, but perfect is going to have to wait (believe me, this is not a trivial problem to solve as I found out when working for a government customer, but for our short runs it should do - the various clocks won't drift that far apart during a run).

I mounted a perfboard on the panel as "you always need a transistor or something" but this time, not yet. I left it there blank till we get some operational experience. Rev 2 is the first one of anything that really works, right?
I made the panel so it'd fit in my op position spot by cutting down a perfectly good 19"" rack panel...sigh.

Here's some pix of the op position end.
100_3103.JPG
The back side. The legs and lead weights are to make it stable on the table.

I used some of the aluminum I cut from the main panel to form legs and weighted them so it'll sit nicely on the table here. Scottish, you'll understand.
Looks fairly nice in place. All I had handy that wasn't black paint was brown, so it's brown. I won't be looking at it much, or if I do, it'll be to see the leds, which are various of red, yellow, and green for various things.
100_3106.JPG
It's at its home now.


I (duh) didn't know this before. Those sexy "safety cover" switches auto close when the cover is closed. To turn them on, you lift the cover and flip the toggle, like any switch. But from there, you can either flip them off, or just close the cover (even just punch the front panel like the Russian cosmonaut in Armageddon). I'm guessing that since you also see stuff like this in fighter planes, they were designed for high adrenaline situations when people might be somewhat distracted, like a fighter pilot - of a fusioneer who is having great success.
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: Fusor remote hardwire safety

Postby Donovan Ready » Fri Aug 26, 2016 4:59 pm

Of course, ripping flacs isn't a distraction... :mrgreen:
Donovan Ready
 
Posts: 239
Joined: Thu Apr 17, 2014 1:22 pm
Location: Austin, Texas

Re: Fusor remote hardwire safety

Postby Doug Coulter » Fri Aug 26, 2016 5:24 pm

Ripped long since - I actually own all that crap as CDs. I was making a backup of my shop share system (on a pi with 2tb) to another huge machine - took 12 hours to do. I just figured it would look better than the wallpaper.
It only took a couple seconds to type in the rsync command, so not that much of a distraction.
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: Fusor remote hardwire safety

Postby Donovan Ready » Fri Aug 26, 2016 8:41 pm

I had no doubt. I was jus' teasin'...
Donovan Ready
 
Posts: 239
Joined: Thu Apr 17, 2014 1:22 pm
Location: Austin, Texas


Return to Machining and Fabrication

Who is online

Users browsing this forum: No registered users and 1 guest