Looking for some GLASS high voltage fast diodes. / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Looking for some GLASS high voltage fast diodes.

previous 1 2 3 4

Move Thread

LAN_403

Pinky's Brain

Tue Sept 06 2011, 04:01AM

Registered Member #2901 Joined: Thu Jun 03 2010, 01:25PM
Location:
Posts: 837

How do they turn out with a ground plane at a realistic distance?

Ash Small

Tue Sept 06 2011, 10:02AM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Ok, I'm with you now, JP. In your previous post I thought you were still in favour of having the multiplier outside of the vacuum system.

That's pretty much the basic idea. a multi-stage electro-static accelerator, with the capacitor plates and diodes spread out along the length of the accelerator, similar to the Siemens design, but probably without the electron-stripping foil.

As it will be for accelerating protons I'll need a negative voltage, but that's just a question of reversing the diodes.

I've not yet started on the 'detail design', but some preliminary 'concept drawings' I posted on another forum a couple of years ago also had the toroidal stages. (since then I took a job for a year in the vacuum/wafer processing industry to gain some more relevant experience, but gave up work a year ago due to health issues). For the past year I've been starting to work on the details of the ion source (mainly the RF matching network 'or' magnetron, still not decided which to use, might try both), the HV transformer, and 'putting it all together'.

Putting the multiplier inside the vacuum system means I'll probably need a bigger pump, due to the increased volume inside the chamber. I've a couple of Diff. pumps, 2 1/2" and 3", but I also have a couple of large turbo pump controllers (untested), I might just get a larger diff. pump. I also have a couple of rotary vane pumps, a dozen or more assorted guages, some brand new, two and a half complete MKS pressure control systems, and pretty much all the other vacuum stuff I'll need (I have a few mass flow controllers too, but none designed for hydrogen. I might try 'trading' a 'new' argon or nitrogen one for a second hand hydrogen one). I obviously can't build the chamber itself until the design is finalised. (I've a few ideas about layout). I've spent thousands (UK not US) on this so far.

As you are aware, I'm concentrating on the transformer at the moment (but all my spare cash is going on my truck at the moment). It now looks like the transformer will be a conventional H-bridge, I've finally been persuaded to use TL494's to control the GDT's. Transformer will 'probably' be a resonant design, maybe with a ballast inductor if required (I've a few ideas about this, but I'll knock up a basic circuit and take it from there). Heatsinks for the IGBT's are the biggest problem at the moment, although I might look into oil cooling, as the transformer will be under oil anyway. (I already have pumps, pipe, swageloks, etc.)

I've various ceramic tubes for the body of the ion source, although most people use quartz glass for this. The rest of the ion source will require quite a bit of machining, and my lathe is still in storage since moving house last year, but I'll be moving again soon, so that's 'on hold' for now. I've a 600 watt 13.6MHz RF generator for which I need a 'matching unit', (well, two big variable (vacuum?) capacitors,) so I might just 'knock up' a magnetron based ion source instead (this might be simpler than building the matching unit). There are several threads on here already about this.

I've still a few months (at least!) until I start assembling the chamber/accelerator itself, but I can't finalise the design until I know what parts I'll be using (hence this thread regarding the diodes)

BTW, do you know of any good 'free' rendering software for DXF or DWG files? (or MDL files, for that matter) I've lost the installation code for the one I used to use, I've downloaded a few 'trial versions' but they all expire after 30 days or so. The only other one I have is nearly 20 years old and doesn't round things off.

jpsmith123

Wed Sept 07 2011, 10:06PM

Registered Member #1321 Joined: Sat Feb 16 2008, 03:22AM
Location:
Posts: 843

Actually that crude model I posted was mainly intended to illustrate what I said in my earlier post about the grading and the shielding (i.e., it shows that there is no HV feedthrough per se, and the electric field is uniformly graded with a value that wont cause problems with a glass tube or whatever. I should've made it more detailed, but I was lazy. If I get some time later I will add a tube/vacuum envelope to it.

Ash Small

Wed Sept 07 2011, 11:27PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

jpsmith123 wrote ...

Actually that crude model I posted was mainly intended to illustrate what I said in my earlier post about the grading and the shielding (i.e., it shows that there is no HV feedthrough per se, and the electric field is uniformly graded with a value that wont cause problems with a glass tube or whatever. I should've made it more detailed, but I was lazy. If I get some time later I will add a tube/vacuum envelope to it.

So you now agree it's better to put all the 'really high voltage' inside an earthed chamber, and just have ~50kV outside?

Here is a link to a thread that raises some interesting points, with plenty of suggestions for further 'googling':

BTW, It was Doug Coulter who first suggested to me by E-mail that the ideal solution would be to put the multiplier 'inside' the chamber, but that the problem would be outgassing from the capacitors. That's why I came up with the idea of using some sort of vacuum capacitor, and then you posted that Siemens design, which got me thinking.....

I'm posting this because I know you are also interested in accelerator design, and you may not have seen that thread before. It's quite informative.

(I've been told that Carl Willis is a safety officer for a company that builds accelerators (he is also chief moderator on fusor.com, but isn't identified as such.) He often talks out of his arse though, for example, he once 'insisted' that a target I'd installed for neutron spallation, in an 800MeV proton accelerator, was solid tantalum, where as it is actually a series of tantalum plated tungsten discs. I can give plenty of other examples, but he 'is' quite knowledgable in 'some' areas. The only way he can cope with repeatedly being 'proved wrong' is to ban people who 'prove him wrong' from posting on that site. (I hope the mods here don't consider this infringes the 'flaming' rules.

))

EDIT: I apologise for the 'rant'.

jpsmith123

Thu Sept 08 2011, 04:06AM

Registered Member #1321 Joined: Sat Feb 16 2008, 03:22AM
Location:
Posts: 843

Unfortunately I haven't built an accelerator yet so I can't speak from specific experience, but relying on general experience, several things come to mind.

Having the accelerator in a grounded enclosure would be good in terms of convenience and safety, and it would be better to only have a few vacuum joints to worry about rather than the multiplicity of joints you'd have with a conventional accelerator tube, yes, but I can see a few disadvantages too.

You'd need a big, strong, vacuum tight metal container of some kind, most likely aluminum or stainless steel, and that would
be expensive. But then if you run into problems, you wouldn't be able to see what's going on. Maybe viewports would help, but that adds to the complexity and expense.

And then of course with so much surface area under vacuum, your vacuum system would have to have a relatively high pumping capacity, and you'd want it as clean as possible, I would think.

And you'd still have to be careful with the design and construction to minimize field enhancement and avoid things like insulator flashover and field emission.

I think the place to start the design would be to get an idea of the size of the metal enclosure you'd have to have. For example if the HV end (ion or electron source) will be inside an aluminum sphere, pick a size (diameter) to start with, say 6" in diameter, then pick a canister that's somewhat bigger in diameter, say 12" ID, find the Efield, and see how it compares to whatever published information you can find for the DC vacuum breakdown potential for a positively charged (or negatively charged, whatever the case may be) aluminum conductor. IIRC it'll vary a little bit depending on whether it's polished and/or anodized.

Let's say the threshold (for "breakdown" in some meaningful sense) is 120 kv/cm, and your calculations show 145 kv/cm...then you have to increase the size until you get below the threshold (plus some safety margin). This would set the ID of the vacuum enclosure. Then you would do something similar to find the approximate length, in this case determined in part by the maximun safe Efield over the surface of your insulators (plus some safety margin). I think this would get you started.

Edit:

I wonder if you could fit it inside of the following linked stock pot:

previous 1 2 3 4

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob