So! Looking around, BobR found one at HVstuff.com, and BillF found one on Amazon, so I got both. You never know, and I don't have a spare bunch of K$ laying around to buy a new DC supply (Spellman supplies are not cheap, just very very good).
Thanks, guys.
The HVStuff one is certainly a thing of beauty - it looks very well made- their catalog pic doesn't quite do it justice (I haven't done a beat on it test of anything here yet), but they don't come cheap, and that's not as much C as I'm maybe going to want...(the other side of the pi network has 1kpf to ground, and I'd like a bigger ratio here).
I didn't believe the caps on Amazon could be 50kv till BobR told me they are, but just need encapsulation. Their picture on amazon is very misleading, they're an inch in diameter - and 3/4" thick! Ok, they won't stand off 50kv externally, but hey look good if you can encapsulate them, so here we go...
One peeve of mine with most doorknob caps we've acquired here is that the thread depth is super short, making it hard to find the right screws, and having to change screws based on how thick whatever else is (things do change around here). So I resolved to make that better here. I don't need the terminals to be at the center on both ends, so to accommodate a longer nut, and generally make this easier, I offset one of them to get a longer internal path length for preventing arcs.
Looking around the shop as to what to make a nice deep nut out of that would be solderable, my eyes fell on some old .223 brass with about the right size primer hole (small rifle). The hole is slightly large for 10-32, but not too far off, and there's a lot more brass thickness, so a #22 drill and a 10-32 tap and I have my nut. It's too long now but the metal saw makes quick work of that, and it even has a built in transition to metal thin enough to solder to. So, after a little fooling around to see how to set this up, I cut and bent capacitor leads for the ground side, greased up a 10-32 SS screw to plug the tapped hole, filled the hole full of solder, tinned the caps and dropped them into the space around the screw. I did this with the caps spaced off the bench a little with some 1/4" plywood so the nut would extend below them, which will be important later.
- Bottom view of assembly with ground nut
- Top view, same stuff
H/T Tamar at 3X3 custom for the "use both fast and slow glue to make your life easier" tip.) I used hot glue to tack this together long enough to finish the build. It won't matter later on.
I made another nut the same way, but this time it was easier to just solder a capacitor lead to one side, and tie the caps together down on the caps themselves. Less tricky soldering in a place hard to do upside down.
- The top, testing the fit in the container
I used a container that once contained chicken salad from my favorite deli. A 3/8" hole in the bottom was close enough to force the slightly bigger brass through, then that hot glue sealed it up well enough to not leak the encapsulant.
- Test fit and caulking
Looks like I'll have a nice 8800 pf capacitor at 50kv once I encapsulate it - assuming the silicone works out and I didn't botch something. I got it at the recommendation of CliffS at Spellman a few years ago, but McMaster-Carr appears not to have this exact type anymore, though there is a comparable GE product, and yes, this stuff is in the range of $100/lb. I've also bought some less expensive stuff at Amazon, we'll see if it works out. You need somthing not-corrosive, but also flexible as otherwise it will shear off bits of whatever you're casting inside it when the temperature moves, or fail some other way, and that flexible stuff isn't cheap.
Here's the manual for using it, for grins.
SilPak_R-2374.pdf- Goop specs and instructions
- (28.5 KiB) Downloaded 277 times
I expected a few bubblies, but nothing like what I saw. They stated it might expand 4x. That's a huge understatement, I wound up using lots of little containers and letting them bubble in vacuum till they settled down before pouring, then putting the whole shebang under vacuum, and it seemed like it never quite stopped. It was goofy enough that I took a video of it. I think, given the very few tiny bubbles I saw on a test cure I cut apart, that under vacuum it wasn't only air, but something volatile in their stuff boiling, it was the energizer bunny of bubbling - and that pump still seizes up when hot, so...I hope I got it done well enough.
- Toil and trouble
Note this level of bubbling was after each sub batch had been degassed for 10 min or so and quieted down pretty well.
I'll be back to post a link to the totally non-pro youtube movie I made of the bubbly. You can tell I only do that as an afterthought...I just don't think ahead enough, and I'm not doing it for clicks or money, so...it is what it is. It does look pretty cool, in a tar-pit sort of way.
Here's the video of that:
https://youtu.be/T4e1ABYUgWM
