Restarting Efforts on Demo Fusor II

For posts specifically relating to fusor design, construction, and operation.
User avatar
Ed Meserve
Posts: 54
Joined: Wed Oct 19, 2011 8:51 pm
Real name: Ed Meserve
Location: Maine, USA

Restarting Efforts on Demo Fusor II

Post by Ed Meserve » Mon Jun 15, 2020 2:43 am

It has been a few years, since I last had done anything with the fusor. After all the participants in the USM fusion group graduated, the lab manager informed me that he needed the storage room emptied and that all of our gathered materials and equipment were going to be disposed of unless I took it. So I inherited the lot.

I have recently been working on getting everything running again and this post is intended to document my progress.
20200510_184312.jpg
Demo Fusor II
The chamber is a repurposed MKS Bellows valve, with a view port scavenged from an electron microscope.
20200510_184410.jpg
Viewport View
At the moment, the Welch 1400 pump is plumbed direct to the chamber. The power supply is a NST lent by Stephen Hansen, of belljar.net fame. Stephen, if you happen to read this, and all go well, I will be able to return it to you when I get the new power supply up and running, ala Chris Bradley (see post viewtopic.php?f=11&t=4887 ).


I pieced together a meter panel to display voltage, current and pressure.

For pressure measurement, I am using an MKS Deadlock 901P-11040, that I picked up from Ebay for about $40. Powering it with a 24VDC power supply I scavenged from a discarded control cabinet, I am reading the analog output of 0-10V for pressure in Torr; Pressure = 10^(Vout-6).

For current measurement, I am using a 100uA analog meter with a 28 gauge ni-chrome wire shunt. I was able to calibrate the meter using a Fluke 787 process meter, normally used to test 4-20mA signal loops.
20200606_154517.jpg
Calibrating shunt (1)
20200606_171742.jpg
Calibrating shunt (2)

For Voltage measurement, I am using a 50uA analog meter with 200 MOhm voltage block for 0-10kV. I have not been able to verify that I am accurately reading voltages on this yet, but I have a Fluke 6KV probe on order.

My plan is to have two web cams arranged so that I can monitor the grid through the viewport (pointing downward) and one web cam monitoring the meters. I will need to up grade the meter cam as it is not of great quality. Below is a the first metered run, with web cam capture and a cleaner shot of the meters with another camera.
20200510_184511.jpg
Web Cam Placement
demo2_run-2020-06-12-2205.jpg
Web Cam Displays
20200612_221514.jpg
Better Shot of Metering
This first run I was only able to pump down to about 400 mTorr, so I've got quite a bit of work to chase down leaks.
Last edited by Ed Meserve on Mon Jun 15, 2020 12:52 pm, edited 1 time in total.
Ed Meserve

User avatar
Richard Hull
Moderator
Posts: 12433
Joined: Fri Jun 15, 2001 1:44 pm
Real name: Richard Hull

Re: Restarting Efforts on Demo Fusor II

Post by Richard Hull » Mon Jun 15, 2020 4:53 am

Welcome back into the fray, Ed.

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
Retired now...Doing only what I want and not what I should...every day is a saturday.

User avatar
Dennis P Brown
Posts: 1996
Joined: Sun May 20, 2012 2:46 pm
Real name: Dennis P Brown
Location: Glen Arm, MD

Re: Restarting Efforts on Demo Fusor II

Post by Dennis P Brown » Mon Jun 15, 2020 7:58 am

Glad to see an old system getting put back on line. I look forward to your continued efforts & posts. You've certainly made a good start getting a lot of critical gauges up and operating - that is the key to all diagnostics. An area of importance that I do not think is as appreciated as it deserves but certainly essential to all successful fusors.

I really like your home build gauge system; impressive put together. More details on those, I'd think, would be a nice thread in of themselves and really help newbie's. Just saying.

Hopefully, my eventual attempt to get my old system (a deuterium accelerator) back on line, as well - just have to finish the building construction and then … . Sorry, I've derailed too many threads this way and need to stop. ;)

As for leaks - double ugh. No greater bane when it comes to vacuum work. Though that level appears to be in the more "massive" catagory. Such large leaks tend to be a bit easier than most - certainly not likely to be out gassing - lol. Sorry to ask (but we all can miss the obivous, at times) but did you rule out the pump?

Looking at your system that foreline and its connection to the fusor looks to be a bit suspect and might need to be upgraded. Epoxy on SS tends to get brittle and leak very easily.

User avatar
Ed Meserve
Posts: 54
Joined: Wed Oct 19, 2011 8:51 pm
Real name: Ed Meserve
Location: Maine, USA

Re: Restarting Efforts on Demo Fusor II

Post by Ed Meserve » Mon Jun 15, 2020 1:17 pm

Thanks Richard / Dennis for the kind words.

I would be glad to go into more detail with the gauge system; I'll step through in detail how I followed Richard's excellent FAQs on the subject.

As far as the vacuum leak goes, at this moment everything is suspect. I basically slapped everything back together and ran it; the pump is due for an oil change and while I do that, I plan to connect the Deadlock gauge direct to the pump to check performance. Then I'll work my way back to the chamber.
Ed Meserve

User avatar
Ed Meserve
Posts: 54
Joined: Wed Oct 19, 2011 8:51 pm
Real name: Ed Meserve
Location: Maine, USA

Re: Restarting Efforts on Demo Fusor II

Post by Ed Meserve » Sat Jul 11, 2020 11:14 pm

With everything plumbed up as previously pictured, the best I was able to pump down to was 251 mTorr (pretty terrible). I tried to use a Doug Coulter trick and introduce helium around the various suspected leak points but, no luck as it appears that it was still pulling in more air than I could introduce helium. (And those party tanks don't hold much helium).
Testing for leaks with Helium.jpg
Testing for leaks with helium
I thought maybe that the spark plug I am using as a feedthough was leaking around the ceramic, so I sealed it up with the vacuum Loctite. No change.
curing epoxy on spark plug.jpg
Curing Epoxy on spark plug
The only other thing it could likely be is the viewport. The chamber started its life as an MKS baffle valve. It was repurposed and mated to a view port scavenged off a salvage electron microscope. It was machine as best as we could manage, but the chamber has two key ways in the viewport opening that is problematic. I also replaced the PVC with some flexible stainless steel line and KF25 Fittings.
New Pump Fitting.jpg
New Fitting, Pump
New Chamber Fitting.jpg
New fitting, Chamber
Starting the leak chase, the first thing I did was test the pump on it's own to make sure that it wasn't the issue.
PumpTest.jpg
Pump Test
It pumped down with just the gauge to 16mTorr in about 5-10minutes, and with just the replacement flexible stainless line with KF25 fittings it pumped down to about 18 mTorr in 5-10 minutes. So it seems the problem is in the chamber.


The original though was to seal it with High Temp RTV gasket material, and two part epoxy. As far as I know , neither are rated for vacuum, plus the whole job was done about five years ago and has been moved and stored and bumped around multiple times.

So I decided to tear it apart, clean it up, and re-seal it with some vacuum rated Loctite from Ted Pella.
Vacuum Epoxy.jpg
Vacuum Epoxy
Before Image.jpg
Before Image
In process-1.jpg
In process - 1
In process-2.jpg
In process - 2
In process-3.jpg
In process - 3
In process-4.jpg
In process - 4
In process-5.jpg
In process - 5, all clean
In process-6.jpg
In process - 6, mixing Epoxy
In process-7.jpg
In process - 7, mixing Epoxy
In process-8.jpg
In process - 8
In process-9.jpg
In process - 9
In process-10.jpg
In process - 10

This stuff takes three days to cure fully on its own or 2-3 hours if you bake it at 140 degrees. So I setup my heat gun a little ways off from the chamber and just bath it in warm air. It's far enough away from the chamber not to melt anything, but close enough to heat it up.

Once the vacuum epoxy has cured, I'll run it again to see how it goes.
Attachments
Waiting for the Cure.jpg
Waiting for the Cure
Ed Meserve

User avatar
Dennis P Brown
Posts: 1996
Joined: Sun May 20, 2012 2:46 pm
Real name: Dennis P Brown
Location: Glen Arm, MD

Re: Restarting Efforts on Demo Fusor II

Post by Dennis P Brown » Sun Jul 12, 2020 11:09 am

A large two stage pump like that with essentially (I hope) no load (i.e. leaks in those few components) should never take five to ten minutes to pump down to just 15 mTorr. That isn't acceptable. That shouldn't take even a minute - the oil in the pump or seals in the connections or a combination might be the issue.

Also, epoxy sealing large metal plates like that isn't a good vacuum seal in the long run - in time the thermal expansion of the metal and aging of the epoxy (gets very brittle; as you can see in your repair) will lead to leaks. I'm sorry but that isn't a good idea. If it was leaking (and not certain of that yet) re-surfacing the plates and/or replacing the o-ring would have been my first step. Vacuum epoxy is generally used for thread connections - not large area, flat metal surfaces.

I guess it is too late to reverse; I do not think that will holdup for a long time (assuming it isn't leaking already.) Hopefully, it holds for now.

User avatar
Ed Meserve
Posts: 54
Joined: Wed Oct 19, 2011 8:51 pm
Real name: Ed Meserve
Location: Maine, USA

Re: Restarting Efforts on Demo Fusor II

Post by Ed Meserve » Sun Jul 12, 2020 4:46 pm

For the original pump-only test I recorded three entries:

7:24pm -> Atm
7:25pm -> 63 mTorr
7:38pm -> 16 mTorr

I don't remember why, but I didn't note anything between the 2nd and the third reading.

I've since swapped out the KF16 flange on the pump to a KF25 for the flexible line, so I couldn't put the gauge back on the pump properly without some hassle.
I did have a rubber sheet, so I poked a hole in it, and held the gauge against it by hand, just now, and the pump brought it down to about 40 mTorr in about thirty seconds, 25mTorr in about a minute and a half.
VacuumPumpQuickTest.jpg
Vacuum Pump Quick Test
Does that sound reasonable, or should I investigate the pump further?


I should also note Dennis was spot on: I ran the pump on the chamber, and saw no improvement in depth of vacuum. I think I will need to fall back to plan B which is to abandon the Valve/Chamber and pick up something with standard ports.
The cost to have the viewport opening machined down (I don't have much in the way of that kind of equipment) would likely be more expensive than buying a used ISO100 T from BMI surplus, Ebay or the like.
Ed Meserve

User avatar
Dennis P Brown
Posts: 1996
Joined: Sun May 20, 2012 2:46 pm
Real name: Dennis P Brown
Location: Glen Arm, MD

Re: Restarting Efforts on Demo Fusor II

Post by Dennis P Brown » Sun Jul 12, 2020 5:07 pm

While that is still slow for such a good pump - more than likely the methodology you are using to couple the gauge sensor to the pump is the cause. Short of bad oil, I think that pump is capable of under 5 mtorr (5 microns) in under a minute pumping on a vacuum sensor that is directly connected via proper vacuum flanges. Those are very reliable pumps and have a good design (overhauls are very simple.)

As for the absolute vacuum (what is your zero vs. a true vacuum that is under a micron), it is possible that the 16 mTorr is your bottom pressure due to the gauge sensor calibration (so you are reading well under 5 microns.) So that is something that can only be determined by an outside reference. That said, again, if the oil is clean and you have proper connections, I'd think that pump can do under 5 microns in under a minute.

User avatar
Ed Meserve
Posts: 54
Joined: Wed Oct 19, 2011 8:51 pm
Real name: Ed Meserve
Location: Maine, USA

Re: Restarting Efforts on Demo Fusor II

Post by Ed Meserve » Sun Jul 12, 2020 5:51 pm

As always, thanks for the insight, Dennis.

I changed the oil in the pump with some Welch Duo-Seal Oil just before this last test. I'll flush it again and retest.

I think you're onto something with the gauge calibration; I just powered up the gauge at ATM, and the analog output read 8.65V, which per the manual should be [ Pressure (Torr) = 10^(Vout -6) ] => 447Torr, which ain't right (atm=760Torr).
I also just read that the gauge analog output factory default is Pascal, not Torr, which is 10^(Vout - 4).

So I need to do the following:

-re-verify my meter is reading the analog output voltage correctly
-establish an RS232 serial connection to the gauge and verify output units (Torrs vs Pascal)
-retest gauge direct to pump (when proper KF25 to KF16 fitting arrives)
Ed Meserve

User avatar
Richard Hull
Moderator
Posts: 12433
Joined: Fri Jun 15, 2001 1:44 pm
Real name: Richard Hull

Re: Restarting Efforts on Demo Fusor II

Post by Richard Hull » Sun Jul 12, 2020 7:02 pm

Gauge inaccuracies are as common as the sea sickness. Regardless of cal issues, your pump is probably OK. You will know more with it is hooked to a chamber. In air, use the images in the vacuum FAQs to get an idea of what your pump is really doing. The lower the pressure, the more voltage it takes to even get a glow mode going. I am fortunate to have so many different gauges, but still, it can be confusing.

viewtopic.php?f=24&t=2795

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
Retired now...Doing only what I want and not what I should...every day is a saturday.

Post Reply