Fusor Forums

Found this on ebay, wasn't even listed as a vacuum component, only as a "tee with 2 3-inch flanges, 2 2-inch flanges and a 1-inch flange"

Turns out it has 2 KF50 (which are almost 3" OD), 2 KF40 (which are about 2" OD) and a KF16. Looks unused, stainless steel.

Should go great with my diff pump with KF40 inlet. Need to find an appropriate gate valve - the only one I have is 2" with non-standard flanges. I'm sure I can improvise something though.

It's going to be a tiny fusor.

That is a real nice find. perfect combo of pump, feedthrough, gauge, viewport, etc.




If your diffpump just has a KF40 on it, I would not bother hunting down a gate valve -- I don't think that tiny gate valves like that are even very common. I would probably just use a normal angle valve like this: https://www.ebay.com/itm/NEW-Huntington ... Sw3yVZiVVH which will also have the side-benefit of making it less likely for an accident in your fusor to get dropped into the pump. My system uses a valve like this.

(this is more important for turbo-pumped systems while I think diff pumps are tolerant of having small metal bits dropped in them while operating, but still its easier to fish things out of the valve than out of the pump.)

These valves are extremely common, and frequently come up at a very good price on ebay. If you can drive a pneumatic valve those are much more common and cheaper as you have probably seen. Of course this would have the side effect that your fusor is now offset from your pump and slightly lower conductance, which I doubt matters. Personally I like the aluminum ones because light weight.

In fact I'm not planning to put the chamber on top of the pump but rather use it in the position it is in the picture with the connection to the pump being a 90deg elbow. But it does make sense to replace the elbow with a 90deg valve, thanks.

I did indeed see KF40 gate valves but a bit pricey and usually pneumatic. I'm pretty sure even the pneumatic ones could be modified to make them mechanical/manual but I'll look into the 90deg ones first, thanks.

As for port usage, I think the KF16 will be used for vacuum gauge and D2 intake, the KF40 next to it for the pump, the KF50 opposite the pump for HV feedthrough and grid support, the KF40 opposite the KF16 for viewport, and the top KF50 unused initially, later probably will have a ion gun.

I would probably try to use a KF50 for a viewport for maximum visibility, if it were my project. Subject to viewport availability and cost of course. Also you never know what feedthroughs, etc you will find.

An ion gun (usually? often?) will have its gas inlet *in the gun*.


How do you plan to combine D2 inlet and a vacuum gauge on the KF16?

ian_krase wrote: ↑Sat Nov 11, 2017 4:22 am I would probably try to use a KF50 for a viewport for maximum visibility, if it were my project. Subject to viewport availability and cost of course. Also you never know what feedthroughs, etc you will find.

I plan to use a camera, not direct observation. As such even a small viewport will do. I have a feedthrough already but I need a CF 2.75 to KF50 adapter to mount it.

An ion gun (usually? often?) will have its gas inlet *in the gun*.

Mine doesn't
(well I could adapt it so it does but it's not necessary)

How do you plan to combine D2 inlet and a vacuum gauge on the KF16?

With a small extension tube on the KF16. Ideally I would need one that has a KF16 at one end, a 1/4NPT at the other and a compression port on a side. That's a bit specific though and I may have to improvise. A 3way KF16 tee would suffice as I'll be able to fabricate custom adapters from there.

Not putting a valve between the fusor chamber and the top of the diff pump is a big mistake for obvious reasons, talked about many times before in these forums.

Richard Hull

My suggestion would be to use the KF-16 port for vacuum. In such a small chamber you will want to reduce the pumping speed as much as possible as this is the only way you will maintain a stable pressure when deuterium is to be emitted. Therefore, I would invest in a KF40 to KF 16 reducer as well as a KF-16 bellows valve (right angle valve), to attach the diffusion pump to the KF-16 port. When I built a KF-50 5-way cross fusor I reduced one of the KF-50 ports to KF-16, then used a KF-16 bellows valve and butterfly valve which seemed to work relatively well.

Here is the fusion claim for my KF-50 fusor: viewtopic.php?f=18&t=10277

As for the other ports, I would use one KF-50 for HV feedthrough, the other KF-50 for vacuum gauge, KF-40 for viewport and the other KF-40 for gas.

Best of luck. I am looking forward to seeing how your project progresses.

Scott Moroch

Is a KF40 or KF50 angle valve really that difficult to control at the low end?

A fusor is easy (for me, once "seasoned or clean") to control having a main gate valve; such an 'extra' control allows one to set a fixed and reasonable exhaust rate for the chamber. This allows one to 'nearly close off the chamber' and thus, use very little deuterium gas to obtain the required pressure that your system will need to get a stable maximum voltage & current. Besides vastly reducing deuterium use, once I set a voltage, I control my current solely via the deuterium valve - this is possible because I have my system nearly closed off using the main gate valve. Anything that simplifies the control of a fusor makes operation far easier.

I would think Scott's idea, for a small chamber, would work very well and such small "gate valves" (KF-16) should be fairly low cost compared to the larger ones.

Thank you everyone for the extra comments.

I never said I wasn't going to use a valve. The choice was between using a flat gate valve followed by an elbow, vs a 90deg valve.

The idea of using the KF16 port for vacuum is interesting; a reducing flange KF40/16 is easy to find, however KF16 valves with manual control appear to be, at least right now in the used market, actually more expensive than KF40 equivalent ones.

I may still use the KF16 for vacuum instead of the KF40; however my plan would be:
1. either this: diff pump(Edwards, KF40) -> gate or butterfly valve (either KF40 or smaller with adapter) -> 90deg on/off valve, electrically actuated (to shut off chamber from pump without having to change the gate valve setting) -> chamber

2. Or, reusing more stuff that I already have: 2-inch diff pump (different from the one above, CVC with cold baffle) -> 2-inch gate valve (which I already have, and matches the flange of the pump's cold baffle) -> custom adapter/reducer which I will have to find or make (aluminum with bulkhead clamp to some KF) -> 90deg shutoff valve as above -> chamber.

I also plan to use a PWM controller on the diff pump's heater to allow for some extra control of the pumping rate.

This is all theoretical so far and needs to go through practical testing. Finding a chamber was the hold-up so far but now I can go ahead with actual testing.

Controlling the current to the diffusion pump heater sounds like a waste of time and effort. Controlling the vacuum rate via a gate valve is extremely fast; controlling current to a heater takes a long time (even for major current changes.)

My main gate valve is a manual butterfly type and works extremely well.

Eh, it's not complicated to do (I have a PWM already built with the appropriate rating) and at worst it won't be useful. I was thinking that it may be handy for long runs at reduced pumping speed (not to make sudden changes during a run). Especially since it looks like I'm leaning more towards using a pump that is oversized for my chamber's needs.

Throttling the boiler electrically sounds bizarre. I would think the oil is either hot enough to form the needed jetting in the stack or it is not. This is what I observe at pretty much a single temperature level in my diff pump. It starts working at a specific temperature, dropping the fusor's pressure to a given sub micron level and a temperature climb of another 40 deg C does not deepen the vacuum one iota. It is kind of an on-off senario. One moment It's not pumping at all then, suddenly, it is going gang busters. Diff pumps do not seem to incrementally pump with increasing temperatures.

Once you actually run a diff pump with a pressure gauge capable of seeing it start to pump, you will observe this. Once it starts to work that is as low as it will go. There is no middle ground of variable pumping speed. The throttle valve on top of the diff pump to the fusor controls and can vary the pumping speed.

Richard Hull

Diff pumps are a semi closed loop feedback system
DC 704 will run at one temp with a standard heater DC705 at another temp albeit higher
If you use Santovac it will be an even higher temp.
The diff pump is a self regulating system using the cooling in the upper parts and the super heated fluid in the lower parts ( giving the vapour flow) to establish the correct operating conditions all of this due to pump design and operating oil characteristics.
do not try to circumvent this setup it will lead to tears!!!
a burping unstable pump is not another variable you need to introduce to a fusor setup.
side effects will include contaminated vacuum gauges and chambers from this burping (farting) phenomeom and silicon oil coated surfaces are a B#$%^tch to clean properly.
Been there
Done that
Got the Tee shirt

OK scratch that part of the plan. Everything else stays the same though. The HV feedthrough definitely needs to go into one pf the KF50 because that is the only way to adapt mine.

Quick update, KF40 manual angle bellows valve. I got it for a good price (the cheapest similar KF16 valve was about double the cost).