Hot Ion gauge filament repair

[Peter Schmelcher]

My used gauge arrived with only one good filament, and then an unexpected mechanical clamp failure vented my chamber… darn.

I had purchased a few meters of 0.15mm diameter W wire thinking that one day I might repair the missing filament, so that day arrived. I did a few quick spot welding experiments adjusting the voltage of my switching power supply. Two large output filter capacitors make nice big sparks when one is careless.

The welding technique that worked best for me was using a solid copper wire and pushing the W into the SS support with it. Solid copper wire elsewhere was also used as an improvisational high current switch contact. The loop on the filament ends keeps the W in position and that makes the spot welding task much easier. I did 3 spot welds connecting the W loop to the SS support (belt, suspenders and rope). All the welds have low resistance and work fine but I was never that happy with the mechanical strength, adequate for the job comes to mind.

6 hours later up it’s and running.
-Peter

Start

forming helix

loop

welding setup

finished job

[steve_rb]

I also have a burned gauge with same look but only single filament. I might try to reppair it the same way. Thank you for sharing.

[Jerry Biehler]

I have a capacitive discharge welder I am assembling right now for welding filaments for my SEM. I will post pics sometime, it is a commercial unit, a Hughes vacuum tube based power supply and a Unitek weld head.

[Richard Hull]

I found a Raytheon vacuum tube capacitive discharge spot welder made for manufacturing vacuum tube components at a hamfest a few years back. It was left as a freebie in the grass as the event was shutting down. Two of its 7 tubes were missing and spiders had built nests in the thing. After a bit of debugging (literally) and picking up a couple of rather rare mercury thyratrons at a subsequent hamfest I got it in working order and have used it ever since. There is a variable joule impulse knob on the front that lets you select between 5 and 80 joules. Perfect for assembling grids or tacking together any number of small metallic components and elements.

Making a capacitive discharge welder is a rather simple task for the electronically adroit. Given successful assembly of a good CD electronics welder, the hand piece and its electrodes are the key to a successful spot weld.

Richard Hull

[Jerry Biehler]

That is almost exactly how I got mine. All the tubes were missing, luckily each socket was marked as to what was supposed to go there. A voltage regulator tube, a couple little thyratrons, a pretty generic valve and a timer. It is pretty cool to watch in the dark, I took some video of it when I got it working:

https://www.youtube.com/watch?v=5KVTrHBW-K0

[Peter Schmelcher]

I enjoy making tools and tooling but the first version is usually just the first iteration.
Fixing equipment is sometimes frustrating, often fun, and typically rewarding; however, I long for a smaller to-do pile.

My uncle gave me an old 1940’s band saw blade welder. It took a few weekends spread out over years to refurbish, align, and learn some tricks to use, and then finally to produce good blade welds that don’t break when cutting SS.

I think the new filament (probably pure W), runs hotter that the previous filament. The cross arm branch reaches a touch me quick temperature after a ½ hour warm up. I think if I oxidize the filament slightly it will improve the electron emission and possibly drop the filament temperature. I haven’t figured a low risk way to oxidize and as they say if it ain’t broke…

[Bob Reite]

Was the original filament coated?

[Peter Schmelcher]

I didn’t take any photos. It was just that I remembered the work function for W oxide is slightly lower than pure W. Thoria is much lower but is also a minor health hazard so I just don’t know what a prudent manufacturer would do.

[Bob Reite]

The ion gauge that I have has a Yttria coated Iridium filament. It is more tolerant of higher pressures than a pure tungsten filament, although if your system suddenly goes to atmosphere all bets are off.

I still think that it is totally cool that you did your own repair. I probably would have just sent mine off to the rebuilder.

[Peter Schmelcher]

I read a 50’s paper on filament designs some time ago. Over the W thorium alloy, a layer of W carbide was created improving the lifetime of high power transmitting tubes, very leading edge technology in its day. Today the lowest work function materials were called a girls best friend, diamond, graphene and other family members are also good and even heating is optional. Must have some catch 22.

I have been pondering ion guns for some time and one design I have been considering has a filament. Learning this now reduces unknown variables in future design choices, and I had some fun doing it as well.

[Bob Reite]

Some transmitting tubes were done with thoriated tungsten, but don't know if you can get those filaments or the material to make them without a license. Come to think of it you could buy thoriated tungsten welding rods, but they are coming under increasing scrutiny

[Roberto Ferrari]

Hi
I use to spot weld W filaments and after several over oxidized, I turned to put a drop of isopropilic alcohol at the spot, getting a nice weld.
Roberto

[Jerry Biehler]

3% Rhenium/97% Tungsten Wire is also used for filaments on mass specs and SEMs.

Yes, you can get thoriated tungsten, it is commonly used for tig welding, though finding wire might be a bit harder. I don't think you really need it for this though.

[Rich Feldman]

I have noticed the general availability of a couple of tungsten-rhenium alloys in wire form.
For example, W3Re and W5Re.
They are used in ultrahigh-temperature standard thermocouple pairs (see types G, C, and D).
http://www.omega.com/pptst/XTA-W5R26.html
http://www.omega.com/temperature/pdf/tor-bw.pdf

Are those compositions popular because of the thermocouple application,
or were they adopted first for physical properties like workability or corrosion resistance?
How do they compare with pure W in, for example, spot weldability?

[Peter Schmelcher]

One could harvest filaments from a 34W T12 fluorescent tube, locally available 24-7. The big catch is removing any trace amount of mercury and not contaminating anything or anyone. Aside from the mercury part almost perfect.

[George Schmermund]

Thoriated tungsten wire suitable for making filaments is readily available on eBay. The high emission characteristics of thorium are not really available in this wire unless the partial pressure of residual water vapor is considerably less than 10^-10 Torr. Most all of what's left in a properly cleaned and baked system in the UHV range will be H2 and H2O. The lifespan of the single atomic layer of Th at 10^-10 Torr will be less than 3 hours and then the emission will drop to that of pure W (3 hours is not enough time to even get through a proper bake cycle). This decrease in emission current will drop by a factor of 10,000 or more if the temperature of the filament is held constant.

Th/W filaments are only useful when used in a gettered, hermetically sealed environment (like a vacuum tube). Filaments must also go through an activation cycle to bring Th to the surface as a monatomic layer after it is thermally dissociated from the thoria component of the wire. Carburization of the tungsten surface can dramatically increase the lifetime of the Th mono layer, but any residual H2O will still doom the high emission properties of the filament.

[Jerry Biehler]

The filament in a fluorescent lamps is probably too heavy to be used in an ion gauge.

These guys sell wire for filaments and rebuild them as well.

http://www.sisweb.com/ms/wire.htm

There is tungsten wore all over ebay. I picked up 500m of .089mm W wire for about $20.
http://www.ebay.com/itm/321462456687?_t ... EBIDX%3AIT

I got everything put together on my setup tonight and welded up a bunch of filaments for the SEM. I am still working on the technique to bend the hairpins.

This is the setup, old Hughes stored energy spot welder and a Unitek weld head. One thing different between this and the welder the OP made is that the output of the capacitor is dumped though a large transformer so you have low voltage, high current. I am willing to be that is what makes a difference in the quality of the welds.

Capacitive discharge spot welder by macona, on Flickr

[Peter Schmelcher]

After a few hundred hours at 1E-7 the filament has settled in and the temperature rise is back to 15C, which I think this is normal. I switched to the unused back up filament and it was also much better with a rise of 20C, something that I was not expecting. The fresh filament had a temperature rise over 50C and you need to start thinking about your electrical signal cable insulation melting. With some aging and these things work great.

Next time I would use a bolt in a drill and form the W wire helix in the grooves of the bolt.

-Peter

[richnormand]

Nice work Peter,

Was the W 0.15mm diag representative of the original diameter of the filament or just what you had at the time? Or did you buy that diameter specifically because it is appropriate for this?
Was the gauge pressure indication OK afterwards? One would think so if the controller automatically adjusts current for a set emission. This could explain why you were running much hotter with an uncoated filament initially. Nice to know it settled afterwards. Perhaps I'll forgo the search for coated wire then.

I too have a nude gauge with only one filament working. Using the spare now. This is the type of repair that might be able to get me back on-line quickly.

Cheers
Rich

[Peter Schmelcher]

It is hard to know when all you have is debris but I believe the new W wire was slightly thicker. I scaled the length slightly to compensate for the lower resistance per length. If the diameter was very different the controller might not have enough current or voltage to operate the filament correctly.

Exactly the accuracy was not altered by the new filament. The controller just operated the filament at a higher temperature to get the preset emission current. Then over time chemistry happened to the W filament and the work function dropped and the controller compensated by reducing the temperature.

I didn't set out to compare filament materials. I am just saying that I am happy with pure W now that it has aged. If I did it again I would pass a current through the W wire and make it glow a dull red in room air for a few seconds, and then put it into vacuum service. This repair was not hard to do and if you buy that $20 500 meter W spool above you'll become proficient in the first few meters;)

Have fun
-Peter