Emma - Fusor Number 1. First deuterium run

Emma Black · Post by **Emma Black** » Sat Sep 03, 2022 2:34 pm

Whist I have a few threads going with the fusor, as the first deuterium was run though it this week, though it was about time for a dedicated thread. Huge thanks to everyone on this site for all the help. Looking forward to lots more experiments in the future.

Power Supply - 110v to 33kv oversized potential transformer, rectified and controlled via variac.

Pumps – Edwards Diffstak 700 diffusion pump & Edwards E2M18 two stage vane pump.

Chamber and valves - 4.5” / 2.75 Conflat cross, manual valve for diffusion pump throttle & sapphire precision leak valve for gas inlet. Deuterium generated via the Horizon PEM kit and held in a 60ml syringe. Grid uses a tungsten carbine ring.

Metering – voltage via high voltage probe and meter. Current via nichrome wire shunt, calibrated manually. Low pressure via Edwards Penning gauge. “High” pressure via Edwards Pirani gauge.

Neutron detection via Ludlum 2363 with prescilla probe. CPM/graphs calculated via Andrew Seltzman’s standalone program.

Run details:
First Run with Deuterium
Duration 10 minutes
Voltage 29KV
Current 11ma
Pressure 8 microns
Recorded CPM (With Andrews code): 37.5CPM

Using the Ludlum over several hours the background count seems stable at 5-6 CPM. After several previous runs this seems unaffected when in proper position next to the fusor and operating on hydrogen only.
The 37.5CPM is potentially low and maybe not sufficient to decisively prove fusion is taking place yet, but this was the very first run with deuterium.

Jim Kovalchick · Post by **Jim Kovalchick** » Sat Sep 03, 2022 5:05 pm

Nice work Emma!

Post by **Richard Hull** » Sat Sep 03, 2022 5:44 pm

Emma, A first first rate report. Just the thing we are looking for.

Now: A proton recoil detector is not a great detector for fast neutrons at amateur levels specially at 29kv which is just starting to "get there".
I used this same type of Bicron detector in 1998-99 to detect and claim fusion at 30kv back then. I know its terrible 0.5% to 1% efficiency as well as its gamma reject capacity.

As such, coupled with your claim of a net 10X increase in numbers over a careful background count, I am going to place you in the neutron club. I would hope you can get that voltage and pressure up a bit in future as long as you are strapped to the Prescila probe. I rapidly graduated to the BF3 tube in an Eberline PNC-1 and got far better results, (higher numbers over background, 100X). Finally with the 3He 4 atm. I can get reliable neutron detection at 15kv. (30,000 X over background at 40kv)

You squeaked into the neutron club just like I did using proton recoil detection in a good probe made for a much higher flux of fast neutrons.

Richard Hull

Matt_Gibson · Post by **Matt_Gibson** » Sat Sep 03, 2022 8:54 pm

I’ve got this Ludlum and can detect neutrons at that voltage, so I’d also agree that you’re seeing them. Congrats!

-Matt

Liam David · Post by **Liam David** » Sat Sep 03, 2022 9:46 pm

Congrats! No doubt you're doing fusion. Welcome to the club.

When you push the voltage higher and counts start to increase, do be aware that Andrew's program will lose accuracy due to how the meter accumulates and calculates the dose. You'll want to calculate TIER from the dose directly. Also, the hammer is somewhat sensitive to x-rays. At much higher voltages (>50kV) my unshielded Prescilla probe gave 1-2 orders of magnitude too many counts. Probably not an issue for you yet, but something to keep in mind if you run at a higher voltage in your current configuration. I've managed to push those 30 kV feedthroughs to 70 kV in air with a toroid, so you have lots of headroom. Use low-profile socket head screws to get that extra little bit, and keep things clean.

Emma Black · Post by **Emma Black** » Sun Sep 04, 2022 4:05 am

Amazing I’m so pleased- thank you Richard!

Upgrading the neutron detection is definitely on the list of projects. I want to learn more about the different setups, as my knowledge of detection tubes, amplifier modules, signal shaping etc is very limited and I’m quite keen to get some hands on experience. Plus being able to experiment with moderators and how they effect the counts is really interesting.

Liam, that’s a very good point - sounds like I have some interesting maths to do to when the counts hopefully start to improve. Before I added the torrid you could literally start to hear what I assume was corona discharge. Good to know that you can push these things. Adding a voltage multiplier is also on the future list as well as some cooling, this thing gets warm!

Also I think there may be a minor leak on the deuterium inlet side so think some air is still getting in diluting the deuterium, so plenty of immediate work to do.

Jim Kovalchick · Post by **Jim Kovalchick** » Sun Sep 04, 2022 8:02 am

Pushing a feedthrough is not without risk. I pushed one designed for 30 kV in the 40 kV range and ruined it. A small blow through in the ceramic at the throat of the connection with the chamber made the feedthrough permanently unusable. It's a crap shoot. You can try it but don't be surprised when it fails.

Dennis P Brown · Post by **Dennis P Brown** » Sun Sep 04, 2022 9:02 am

Congratulations on measuring neutrons and building a very good system. I also get neutron detection at 30 kV but yes, a 3He detector should do better. Jim raises a good point - those feed-thru's have a rating level that should not be pushed. I isolate my feed-thru by mounting it on a large glass disk that is used as the vacuum flange for my chamber.

Emma Black · Post by **Emma Black** » Sun Sep 04, 2022 9:21 am

Yes fair point - I do have a couple of spares but would rather not blow one. Wondering how difficult it would be to make something custom from scratch. A friend of mine is excellent at pottery with a kiln etc. Initial idea would be to 3D print the shape of the ceramic needed, make a mould and then slip cast (liquid clay) the ceramic, before finishing, firing and glazing etc.

Anyway. This morning think I have fixed the little leak, todays short run was a bit better:

Voltage & current - 32kv 10ma
Pressure 9 microns
Average CPM= 64.5 over 6.75 min

Matt_Gibson · Post by **Matt_Gibson** » Sun Sep 04, 2022 10:46 am

I also push my 30kV insulator to 70kV using a toroid on top and a half toroid over the bolts.

What mR/hr values are you seeing on your Ludlum?

-Matt

Liam David · Post by **Liam David** » Sun Sep 04, 2022 10:50 am

When pushing the 30 kV feedthroughs much beyond their rating, cleanliness is paramount. That includes a lack of any metal scrapes or deposits, as well as loose dust. A bakeout can also help tremendously. For those not wanting to go the acid cleaning route, I've found that 30 min - 1 hr in a heated ultrasonic bath works to remove all metal from the insulator. Add Alconox or some other strong cleaner, and after a rinse, it's UHV ready.

Rolling your own feedthrough is definitely a possibility, and I'd certainly follow along with interest. Just be sure that the parts are non-porous and don't have any residual binders, or anything other than ceramic, for that matter.

Matt_Gibson · Post by **Matt_Gibson** » Sun Sep 04, 2022 12:09 pm

Does current (increasing heating) come into consideration when pushing a feed through way beyond its rating? I’ve pushed a 20kV to 54kV and a 30kV to 70kV without any problems, but I also don’t get much more than 10mA at these higher voltages.

-Matt

Liam David · Post by **Liam David** » Sun Sep 04, 2022 12:22 pm

I haven't had issues up to 20 mA. The plasma heating load certainly goes up, but these feedthroughs are rated to something like 350 or 450 C.

Emma Black · Post by **Emma Black** » Sun Sep 04, 2022 2:17 pm

Matt - I'm getting 2-3 mR/hr and the hammer is 34cm from the centre of the fusor.

Matt_Gibson · Post by **Matt_Gibson** » Sun Sep 04, 2022 2:57 pm

That’s pretty good for 30kV from that distance. I have mine around that distance and get around 8mR/hr pushing 70kV (8mA).

-Matt

Liam David · Post by **Liam David** » Sun Sep 04, 2022 5:28 pm

Your cathode doesn't look very symmetric within the chamber. Fixing that might help improve rates a little, although with that geometry it might not make that big of a difference. With 2-3 mrem/hr and 34 cm, I'm calculating 2.3-3.4e5 n/s. Not too shabby at that voltage/current.

Emma Black · Post by **Emma Black** » Sun Sep 04, 2022 5:52 pm

Yep I know. It's off by a few degrees although it looks worse in the photo than it is. I can only think I must have knocked it during a clean, but didn't initially notice. With the chamber now getting down to 3.5x10-7 mbar I'm reluctant to open it again for the moment.

Feel I should know this but out of pure laziness in asking, what is the equation used for calculating the n/s in this situation? Edit: I should have looked at the FAQ's...it's been a long day.

Luca Aldridge · Post by **Luca Aldridge** » Wed Sep 07, 2022 4:27 am

Emma,
Well done. The build looks professional and something I’d aspire to. Good luck on improving your flux!

Emma Black · Post by **Emma Black** » Wed Sep 07, 2022 6:07 am

Thanks Luca, you will get there soon

I've been running it few times a day for the past few days now and it seems to be improving by 5-10% better counts each time.

There still must be either a leak or lot of off gassing happening though, as when left overnight, it takes a while for the current to drop at the higher pressures as the deuterium takes over inside. The plasma is an interesting whiteish colour during this stage. This seems to be getting quicker each time as I assume the chamber gets cleaner.

Hopefully I can have a go at some simple activations in the next few weeks.

The raw data from the Ludlum for a short run:

Luca Aldridge · Post by **Luca Aldridge** » Wed Sep 07, 2022 12:19 pm

That’s good, I guess it’s all down to practice.
Data looks good!

Dennis P Brown · Post by **Dennis P Brown** » Wed Sep 07, 2022 2:08 pm

Is the plasma white or rather a faint blue white? Deuterium tends to be red or a light blue in my experience.

Doubt you have a leak if the system is improving with time; that tends to indicate out gassing and that issue is being reduced by pumping and plasma interactions.

Emma Black · Post by **Emma Black** » Wed Sep 07, 2022 5:14 pm

Yeah it was a blue white glow and was decreasing with each run.

Another one of the rubbish diodes has blown anyway. So while I wait for a better quality replacement to arrive, I opened the chamber to sort the cathode alignment.

Have a look at the colours in the steel, being a small chamber with welds, the heat is very intense.

Dennis P Brown · Post by **Dennis P Brown** » Wed Sep 07, 2022 5:19 pm

Sputtering of the cathode is normal with a fusor. In mine, the window was so coated it appeared as a high quality mirror. Needless to say, I added a protective glass plate to stop that. Cleaning off the window wasn't fun nor good for the window.

Those diodes are, at best, rated to 20 kV but likely that is optimistic and considering that the voltage rms is a good bit higher then what we measure on a voltmeter, even using two in series (40 kV) isn't really much insurance.

Emma Black · Post by **Emma Black** » Wed Sep 07, 2022 5:55 pm

I'm not even sure how you would go about cleaning that!

Yep I used 3 of the 20kv ones in series but they still failed in the high 30's. The failure took me a minute to figure out what was going on.

All running fine and the plasma suddenly died. No current. As I increased the pressure in the chamber, the plasma lit up again. Then current suddenly spiked massively and would probably have popped a fuse etc had I not had my hand on the variac to immediately turn it down. The power supply is now outputting AC only and is only able to support plasma at higher than normal pressures.

Matt_Gibson · Post by **Matt_Gibson** » Wed Sep 07, 2022 9:43 pm

I might have missed it, but are you using a ballast resistor? They will save your diodes when current spikes like that. Something like this is a good one:

https://www.ebay.com/itm/114714898756?m ... media=COPY

Also, try to control current via pressure control. Run at lower pressures and very gradually increase to increase current.

-Matt

Emma Black · Post by **Emma Black** » Thu Sep 08, 2022 2:46 am

Yep there are two 50k's in parallel, normally its remarkably controllable. Those sapphire leak valves give you such fine control over the pressure.

However with some failed diodes its power supply was putting out AC only, it was behaving exactly like a fluorescent tube. Nothing for ages then it suddenly lit up with more current than normal and the pressure jumped at the same time.

Dennis P Brown · Post by **Dennis P Brown** » Thu Sep 08, 2022 6:50 am

Sorry for the diode fails - again. I keep a NST on hand to test my diodes - very easy and fool proof to determine their status.

What Matt suggested is how I control my fusor. I get it to high vac (under 10^-4 torr), then throttle the turbo (or diffusion pump) to it is almost closed position (but the system does not experience a pressure rise.)

I then set my variac to about 10 to 15 kV. After this I slowly increase gas pressure via the leak value (For my system this is always four turns. But everyone's system is different but should be consistent once you set up your system.) When I see current flowing, I bump up my voltage up slowly. As I do that, I notice my current increasing. I stop increasing my voltage when I'm at max. current for my x-former. If below 25 kV, I lower my deuterium pressure by slightly turning the leak value down.

The current falls.

I then bump up the voltage a few kV. I repeat this process till get to full voltage (30 - 31 kV) and normal max current (30 ma.) Since my system is clean this tends to hold steady; however, after it has been exposed to air, I'll often get current spikes. Again, like you one has to be fast on the variac.

Aside: there is zero reason (and a lot of reasons not to) clean your fusor. Those metal deposits will do no harm and even are helpful to act as a "getter" on the surfaces. The window, however, is an other matter. Acetone and a soft lint free rag should work. If stubborn, I use a very fine steel wool (but that can scratch so be careful.)

Matt_Gibson · Post by **Matt_Gibson** » Thu Sep 08, 2022 7:47 am

What’s the peak current rating on your diodes? 25kohm ballast might be low enough to allow too much current during a “short”.

I bet your system is still burning off “junk”. I’d keep voltage low and let it do some cleaning runs where you let the cathode heat up, and remain hot, for longish runs. Eventually you’ll see just the faint reddish glow of deuterium.

I cleaned my viewport with a dremel, some 50k grit powder, and a felt bit. It did a great job in not much time.

-Matt

Dennis P Brown · Post by **Dennis P Brown** » Thu Sep 08, 2022 10:17 am

Two 50 k's in parallel? That is a problem - two resistors in parallel operate such that most all current runs through the resister with the slightly lower resistance value - so using two doesn't buy you any reduce load on the resisters - one is already carrying 99+% of the current anyway. Those resisters need to be in series giving you 100 k-ohms. As you are seeing, 50 k-ohms is just too low.

Matt is likely correct that your system isn't clean enough, yet. Running the plasma at lower voltage (with some deuterium) is a good idea.

Liam David · Post by **Liam David** » Thu Sep 08, 2022 12:21 pm

Two resistors in parallel is not a problem... the current will flow in inverse proportion to their resistances. In fact, there is even a limited degree of thermal regulation. The lower resistance will dissipate more power (V^2/R), thus heating up, increasing in resistance, and so reducing the fraction of current it carries.

Dennis P Brown · Post by **Dennis P Brown** » Thu Sep 08, 2022 1:32 pm

Not an electrical engineer - simply learned from one that identical resisters in parallel do not carry similar currents. Rather one gets the majority of the load. Maybe not - never did the experiment myself but if you did, then that is the case.

Regardless, 50 k-ohms has proven inadequate (obviously the resisters in series have failed to protect in runaway conditions.) so it would appear they need greater resistance so placing the resisters in series is the better idea in order to protect her transformer and diodes.

Matt_Gibson · Post by **Matt_Gibson** » Thu Sep 08, 2022 1:42 pm

I think it’s diodes in parallel that don’t share load equally.

-Matt

Emma Black · Post by **Emma Black** » Thu Sep 08, 2022 2:33 pm

Yeah I think i'll add this 100k monster while I'm diving into the oil tank, just to be on the safe side, as clearly 50k is not limiting the current enough. The old diodes said 2amps but take that with a huge pinch of salt.

Correction from earlier - It was two 100K's in parallel 50k total.

New ones are microwave type, but from a proper supplier with full spec sheet etc. 12kv @ 500ma. I will use many.

Anyway original issue it turns out was entirely due to my ineptitude. A usb cable had physically fallen and brushed past the feedthrough on its way down. Did not notice until I had a look today. There are tiny burns on the cable and the camera and power adaptor are both dead. The resistor value issue still applies though.

The fusor itself is fine and after a pump out I'm getting down to 2x10-7mbar now, the lowest its ever gone.

Denis - my previous operating procedure was pretty much identical.
Let it hit below 10^-5torr. Throttle back the pump. I have also my moved my deep pressure gauge between the fusor and the pump, so you have an idea of how much gas is leaving the chamber. The other gauge will stay put. Then slowly build the voltage and pressures adding the deuterium, while watching the current.

I was more interested in how oddly a chamber filled with low pressure deuterium behaves, when exposed to HV AC current. Nothing for ages then brightly flashes to life in a fraction of a second and acts like a dead short.

Matt_Gibson · Post by **Matt_Gibson** » Thu Sep 08, 2022 2:52 pm

2A peak rated diodes would definitely struggle with a short, especially if you have a capacitor in there.

Take a look at the 2CL2FM diodes. They are rated for 20kV, 100mA, and can run up to around 35kHz. They are able to take up to a 10A spike (non repetitive).

-Matt

Emma Black · Post by **Emma Black** » Thu Sep 08, 2022 3:08 pm

Thanks Matt - the ones I have picked up are 12kv 500ma and can deal with 30A surge. A local store had these for £18.50 ($21) for a 100 pack.

https://www.tme.eu/en/details/by12-dio/ ... ctor/by12/

Matt_Gibson · Post by **Matt_Gibson** » Thu Sep 08, 2022 3:16 pm

That’s a solid deal right there!

Edit- Btw, sorry for your loss, the Queen was a truly classy woman and great leader!

-Matt

Emma Black · Post by **Emma Black** » Thu Sep 08, 2022 3:30 pm

Thanks Matt, yeah I know. I only met her the once very briefly but she came across as genuinely nice with a very quick sense of humour.

Post by **Richard Hull** » Thu Sep 08, 2022 4:53 pm

Diodes in parallel are indeed a no-no.

Resistors of identical value100k @5w, in parallel, share all currents perfectly 50:50. Their total would be 50k @10watts. With 10ma flowing through this combo would subtract 500 volts from the supply voltage to the fusor. The 10 watts would be double what a 10ma would demand (5W) and would be nice.

Two 25k 5w resistors in series would do the same thing yielding a 50k @10watts. Again, this combo would do the same thing as the above example.

Resistors of any equal value but of a fixed wattage will have their wattages in either series or parallel combination sum.

A100k resistor would drop 1000volts from the supply at a fusor current of 10ma. The wattage would be 10 watts demand so a 20 watt resistor would be better.

Richard Hull

Dennis P Brown · Post by **Dennis P Brown** » Fri Sep 09, 2022 12:03 pm

Well, that settles the matter on parallel resisters.

Emma Black · Post by **Emma Black** » Fri Sep 09, 2022 8:11 pm

Though I had replied already, thanks for the clarity Richard Is there any real benefit for large wattage resistors for use in fusors i.e. 100W+ ?

Have also been getting ready for some activation attempts - lots of HDPE, from Amazon cutting boards. I was going to use wax, as I have a massive bag but the fire and melting risk is not worth the hassle.

Post by **Richard Hull** » Sat Sep 10, 2022 2:32 pm

Wattage only comes in handy with dead short conditions. 25 to 50 watts will handle most anything up to a longish short like 3 seconds or more. (time to reach the kill switch.)

Most fusor supplies are whimpy. 20 ma is a death point at full tilt. Monster supplies that will handle 60kv @100ma will need the higher wattage ballasting as the supply is quite happy vaporizing a central grid in pressure bumps or electron runaway conditions.

Compute a good resistor value that you can tolerate under a load of say 30ma to a running fusor. Let us say 50K . Now figure the power rating.
P = I^2 x R .030 X.030 = .0009 X 50,000 = 45 watts Double that to 100 watts or so. Now, few power supplies will supply 30ma at full voltage. Still be prepared for near short conditions. Most likely 20ma would be more normal as max current before the supply blows up. .0004 X50,000 = 20 watts and a 50 watts would be OK.

Richard Hull

Matt_Gibson · Post by **Matt_Gibson** » Sat Sep 10, 2022 2:53 pm

The higher wattage resistors are probably larger/longer and might handle higher voltages than smaller ones. That might be a reason to go for one of those.

-Matt

Bob Reite · Post by **Bob Reite** » Sun Sep 11, 2022 11:08 pm

During an arcing event, figure the impedance is close to zero, even though the duration is short, so you need at least a 100 watt resistor. You can find 225 watt wirewound resistors at hamfests. I plan to bring some extras to sell to HEAS.

Post by **Richard Hull** » Sun Sep 11, 2022 11:35 pm

That's great Bob! I dare not think how many fusors have benefitted and been brought to life within a budget via the hamfest route. Hamfests are hit-and-miss related to fusor construction, but very viable. The HEAS flea market is the only place where you can see, touch and feel what is for sale and also be talking to a knowledgeable person selling his fusion related wares. E-bay is far more vast and has most anything you might need, but remote at best, with sellers ranging from idiots to the very helpful, but rarely knowledgeable in your area of need.

Fortunately, most arriving and staying on fusor.net know how to use real tools in their hands to assemble materials and build things. What they don't know, they pick up through self-directed learning, reading, seeing and then the doing. The materials needed range for hardware store found materials to the most complicated of scientific materials and instruments. As always, the true go-getters, go and get.

Richard Hull

Emma Black · Post by **Emma Black** » Mon Sep 12, 2022 11:34 am

I'm off to a hamfest next month to take a look for useful stuff, never been before we shall see what crops up.

Anyway - rebuilt the bridge using 6x4 12kv diodes, installed a new 225w resistor and we are back in business. Pressure and current is nice and stable again. Neutron counts are around 0.6-1.0 mrem/h, so some I think some reconditioning needed.

Emma Black · Post by **Emma Black** » Tue Sep 13, 2022 5:21 pm

I did not appreciate how much the weather can play a part in HV systems. Today we have thunderstorms and the humidity in the workshop is crazy. All sorts of discharge sounds coming from cables and connectors.

The video (for sound only) was taken at around 30KV (with the fusor disconnected) via a iphone recording on top of my potential transformer, sound is mostly from the transformers HV connectors.

Emma Black · Post by **Emma Black** » Mon Sep 19, 2022 5:50 pm

Update
Somewhat predictably my feedthrough has started arcing at voltages over 30kv. Out of the other two I though I had, one arcs even worse and the other wont hold a vacuum.

Its failed where the ceramic goes through the conflat, as seems to be common with other posts on here about failed feedthroughs. Running things in high humidity can't have helped at all. I now realise what may have been obvious from the previous post and that the discharge noise was being caused by condensation on the cables, due to the humidity and low temps.

Fixes & Changes
I've added a dehumidifier & heater to the workshop and re-routed the HV cables. They are now all sitting in PVC pipes filled with oil and wrapped in some shielding tape. Also spent some time dealing with all the sharp points, now no more audible corona. I've also changed the oil - was previously using fully synthetic motor oil, now switched to proper transformer oil (it was on offer on amazon)

New cable routing:

So figure it's time to build a new better feedthrough.

Thinking of building something close to the excellent post by Liam David around building a custom feedthrough using quartz and compression fittings. I could just use borosilicate glass and heat to seal. Making my own ceramic is also an option, but the main issue I think is going to be the material shrinkage 10%+ after firing and I'm not sure how I could get that to seal. Maybe something like this:

Or get another 30KV feed through and modify it to cope with higher voltages, perhaps by adding an "oil jacket" to the air side of the ceramic. Either way I'm currently playing around with Ansys Maxwell to experiment with some simulations.

A few of the posts around very high voltage feedthroughs discuss two stage biased designs (I'm not thinking of building this) maybe a silly question but how does this work?

Matt_Gibson · Post by **Matt_Gibson** » Mon Sep 19, 2022 8:06 pm

That’s too bad. I’ve got three of these feed throughs and have been able to way exceed their voltage ratings. I do use some 99% isopropyl alcohol on them before use…

Were you getting your feed through really hot? I haven’t pushed much beyond 700w and I have a fan blowing across, so maybe that’s a thought?

-Matt

Post by **Richard Hull** » Tue Sep 20, 2022 12:25 am

I have a large box fan blowing across my insulator so that any small corona ozone/air ions that might cause or support an air arc via rising heat columns are blown laterally away from the insulator. Corona in air can be subtle and unseen even in the dark. It doesn't take much in a rising heat column to establish an air arc. The fan also helps cool the fusor, too. Many tricks out there when pushing the limits of materials.

Richard Hull

Liam David · Post by **Liam David** » Tue Sep 20, 2022 12:48 am

Could you post a picture of the insulator failure, and the one that arcs worse? My personal sample size is one, but even with a toroid, the failure mode is typically arc-over on the air side. Getting the voltage high enough to puncture the insulator, or arc along the vacuum insulator surface, would typically require an oil socket. Unless something in your commercial part is much different from mine, I don't see how anything other than contamination, or perhaps a problem with the stalk, can cause a problem.

Emma Black · Post by **Emma Black** » Tue Sep 20, 2022 6:11 am

Richard - I did see mention of you using a fan to help blow away the corona, this is on this list to implement! It's perhaps worse in this setup being all inside a box. Also I could hear the corona, but even in the dark could not see it. I tried a "bat detector" to amplify any ultrasonic sound, which I think is similar to how commercial corona detectors work. It worked slightly better than the mark 1.0 ear but not a massive improvement.

I think Liam you are quite correct and its not indeed shorting through the ceramic but forming an arc all the way around the end due to contamination. The reason I thought it did was because when it flashed over the ceramic lit up brightly like a lightbulb, it's fine on the air side. Also both of the other used ebay feedthroughs leak it turns out. The feedthrough and stalk are absolutely filthy now.

I risked a very non-definitive test in air with the feedthrough on the bench and the arc indeed forms around the end of the ceramic not through it. Perhaps I was jumping to conclusions too soon. Maybe, after a deep clean and a stalk re-design it may actually be ok.

Matt - I was running it fairly warm for short periods of time - the cooling has become quite an issue.

Emma Black · Post by **Emma Black** » Thu Sep 22, 2022 3:59 pm

Feed though has been cleaned by carefully using 1500 grit paper followed by an ultrasonic bath, with water based solvent to clean any sanding residue. Finally finishing with IPA and wipes to remove anything else. The feedthrough itself seems ok, yay!

As for the second part, posting as a failure for what not to do.

For the stalk re-design I should probably have actually done some simulating or indeed any calculations rather than throwing something together to experiment with. More thought less rush etc..... This sparks excessively in one place, even after a long cleaning run at low voltages. Issues seem to include:
-Glass is very thin & close to the stalk, meaning its insulating value is likely near zero.
-Stalk diameter is too large and the junction with the glass creates a horrible area of high field.

Damage to the glass - post removal

Liam David · Post by **Liam David** » Thu Sep 22, 2022 4:58 pm

Yep, that would be a triple junction problem. The glass can also trap contaminants at the metal surface, shielding them from effective cleaning by the plasma. It'll do much better without the glass.

As you suggest, there's an ideal diameter for your 2.75 conflat port. Too small and the high theta curvature (r, theta, z coords) enhances the field. Too large, and proximity to the wall increases it again. This ideal region is actually pretty flat, so for the often dirty, crooked, or otherwise nonideal stalks in most fusors, its diameter never really becomes the limiting factor. For an ID of 3.65 cm, which is pretty typical for 2.75 conflats, the ideal diameter comes out to 1.35 cm. This diameter gets modified by the presence of a dielectric like feedthrough ceramic, but it remains analytically solvable.

Post by **Richard Hull** » Thu Sep 22, 2022 6:21 pm

You want all the high field in the system at the central, grid, ring, sphere, etc. You have a lovely, smoothly radiused and polished ring!
The large diameter stalk would be just the thing, if, you had more clearance around it and the feed through insulator. I bet a sharp edged thin cylinder would cure many issues as long as you haven't "tracked". permanently. the insulator. It is forever all about low field control along the stalk merging to extremely high field conditions at your central toy, whatever that is.

Richard Hull

Liam David · Post by **Liam David** » Fri Sep 23, 2022 12:15 am

A high field at the cathode doesn't help, generally speaking. Quite the opposite, in fact. I'd argue it hurts the performance in most systems here. It also has little to nothing to do with where the plasma will form. My first suggestion would be to increase the cathode length several times and decrease its diameter somewhat, but the one you have now is nicely made and it would be worth collecting data on it.

Matt_Gibson · Post by **Matt_Gibson** » Fri Sep 23, 2022 5:28 am

Is that a tungsten wedding ring?

Emma Black · Post by **Emma Black** » Fri Sep 30, 2022 1:43 pm

Friday update time

The cathode has been remade. I've downsized the stalk diameter by 2mm and pulled the glass right back, so its now just being used to keep everything central. I cut the glass using a simple tube cutter then melted the ends to smooth it out and get rid of any dust. Have only done the first couple of re-conditioning runs but so far no arcs. I'm picking up a new lathe soon so it should be a lot easier to make my own stuff going forward, be fun to get back into machining things.

I've also had to replace one of my PM1029B panel meters as the previous arc issue seems to have destroyed it. It now only displays nonsense segments. It's now in a metal box and the cables shielded.

Matt- yep from Amazon. I cut a hole in the side to thread the stalk on using a little diamond bit.

Before cleaning:

Grid is now much better aligned to.

Emma Black · Post by **Emma Black** » Mon Oct 17, 2022 9:54 am

I wanted to experiment with using continuous albeit lower voltage in between runs to help keep the chamber clean, bit like an ion pump.

Not wishing to use my main supply for this I took a cheap ebay AC power supply, connected it to the fusor after it had sat overnight allowing for the pressure to rise. No pumps were turned on. Plasma struck at around 3.8kv and over a couple of minutes the pressure slowly dropped until the point at where 6kv plasma would flicker on and off as the pressure rises and falls. When the plasma is lit it's using no more than 0.5ma.

I tested it in this state for a couple of hours with no pumps running (see video) and the pressure stays put. I wonder if 5 minutes per hour in my setup would be enough to maintain a 6kv or higher plasma for long durations. More testing to follow.

Dennis P Brown · Post by **Dennis P Brown** » Wed Oct 19, 2022 7:54 am

Unless you have a leak (real, not virtual) your system should hold vacuum and remain clean. Running continuous power (a plasma) really isn't necessary - my system holds a fairly good vacuum and after three months (all of 3 torr), fired up and ran flawlessly on its first attempt.

Post by **Richard Hull** » Wed Oct 19, 2022 3:41 pm

Emma, What you observed is the norm for a tightly sealed system. If you vacuum to a very low level and are almost perfectly sealed and let in enough gas to strike a plasma, then keep that voltage on, over time, the plasma should go out as you are burying the gas. (especially hydrogen). with more voltage the plasma can restrike but ultimately, the system will go so "hard" in vacuum, no voltage will restart the plasma. (like you note, an ion pump like operation is occurring.)

With a leaky system the above scenario will not work out or ever go totally "hard" as gas will always leak in allowing the plasma to strike.

Richard Hull

Emma Black · Post by **Emma Black** » Wed Oct 19, 2022 6:06 pm

3 torr in 3 months is pretty cool. Richard I assumed that was the case, this is very much a learning experience, I'm still new to vacuum systems. After another hour or so of playing with it you indeed couldn't strike plasma at all. It's also possible that there is minor leak still but its on the order of microns per hour and not worth looking at while things are otherwise working nicely.

Rubbish photo but I had the tungsten grid glowing today. Interestingly as soon as the grid starts to glow the neutron numbers improve even when the current is dropped subsequently.

Matt_Gibson · Post by **Matt_Gibson** » Wed Oct 19, 2022 7:20 pm

What power levels are you seeing before the ring begins to glow?

-Matt

Emma Black · Post by **Emma Black** » Thu Oct 20, 2022 2:25 am

It starts to glow around 300W, 8-10ma at 35-40kv but only after about 20 minutes of bombardment. It's obviously more visible on the camera than by eye. There is no glow on the stalk at all.

Jim Kovalchick · Post by **Jim Kovalchick** » Thu Oct 20, 2022 7:23 am

Emma,
When I was using tungsten carbide wedding rings for my grid I found that they took a bit of doing to get them glowing, but it could be done. Attached is a picture of one of mine glowing to the point that it's incandescence dominates the color in the chamber versus the plasma.

I think this image was at over 400 W, but I would have to check my notes.

The other image is another TC ring grid after shutdown showing that it still glowed red for a while.

Jim K

Dennis P Brown · Post by **Dennis P Brown** » Thu Oct 20, 2022 11:50 am

Emma, a few microns an hour is almost certainly a "virtual" leak (i.e. simple out gassing.) As pressure climbs some, it will slow; ditto when you keep it under vacuum. Running a continuous high voltage voltage, as you found, doesn't remain lit. So your system is likely very sound.

Emma Black · Post by **Emma Black** » Fri Oct 21, 2022 1:29 pm

Jim, those photos are very cool. I was just reading through your original thread theorising on the reasons for the increase in neutrons once the grid glow starts. In terms of thermionic electrons generating plasma more efficiently.
You would think this would be the reverse and the electron emissions would start to ruin the fusion. However this is not what I seem to be seeing so far in this setup, it seems to increase non-linearly, which is confusing.

Post by **Richard Hull** » Fri Oct 21, 2022 2:09 pm

All of this oddball behavior is due to what I have preached from the beginning over the 23 years here. This is a simple system! It is a simple system that is not directed in a specific single manner to do fusion. It does fusion over a wide range for an equally wide range of possible reasons in many zones. Just as it thwarts fusion for also many reasons and in many zones. You just can't simplify this simplistic fusion system. Suffice it to say, it does measurable fusion and that specific increased fuel pressure at elevated voltages and currents suited to a specific reactor will increase fusion.

It is generally found that to get the most fusion out of any design, the central cathode element will heat and glow or even nearly incandesce.

Richard Hull

Emma Black · Post by **Emma Black** » Sat Nov 05, 2022 4:36 pm

Things have been busy but finally had some success and managed (I think) managed to activate some fine silver sheet to a credible level. This was at 40kv 6ma, with a 10 minute run, using cut up chopping boards as a moderator. The ludlum was showing between 5-7m/rem hour and has not been moved from its previous position.

Within the first 15 seconds I can get 2k cpm with a pancake probe, with the silver pressed between a sheet of lead, it then drops rapidly.

Bad news however is that the leak seems to be both real and getting worse. Thanks to the very helpful "air duster" trick I believe I have managed to narrow it down to the valve stem on the valve I'm using to throttle the diffusion pump.

Post by **Richard Hull** » Sat Nov 05, 2022 11:38 pm

Emma you have made the elite fusioneer club. I have added your name.

Richard Hull

Emma Black · Post by **Emma Black** » Sun Nov 06, 2022 3:36 am

Amazing, thank you Richard. The amount of reading and effort involved with getting the system stable enough at for me the higher neutron numbers has been considerable, this place is a great resource.
Each time I stacked the moderator up, it felt like something would go wrong, from blown diodes to grid issues. Hopefully now I have appeased the problem gods enough to allow for some more experiments. Sticking with silver I would perhaps like to try some silver nitrate in solution next.

Secondly whist I'm sticking with "no upgrades", until all has been had out of the current system. It's hard not muse on of what to do next. More voltage would be nice so the power supply is the next likely target. Want to stick with the potential transformer PT so my initial thoughts/options are:

-Keep the PT at 50hz and use some monster caps to build a voltage doubler. Expensive and dangerous, but it would be bulletproof.
-As above but use increase the frequency of the PT via perhaps an audio amp, allowing for smaller caps.
-Use a second smaller transformer in between the PT and the bridge, that I wind myself. No idea about how well this would work or if it would be practical,

Edit - Thought I added this to the previous post, but here is the graph from one of my initial conditioning runs if anyone is interested. Voltage and current was maintained throughout at 35kv and 4ma by slowly increasing the D2 pressure.

Post by **Richard Hull** » Sun Nov 06, 2022 4:37 pm

Excellent work and data reporting! You also have a good grasp of what to do to prevent problems while expanding your neutron numbers. Your electrical knowledge is advanced. While I think another HV transformer would be possible between the PT and the bridge, but as you note, would be a very special wind to stay on the 50hz. Likewise, I have considered big caps and a doubler on my X-ray transformer but the stored energy is a big nasty.
If, as you note, you wish to keep what you have and just add as needed to improve it before moving to a more advanced fusor, you will find you continue to learn. This is mostly in operational experience. Wring the system out to its max. I always say as a personal maxim, " never seek the better without first securing the good and all that it allows."

Finally, if it is not too personal, the name Emma is traditionally a woman's name. Is this the case? If so I would like to add this to the fact that you are the first woman to become an elite fusioneer in your note on becoming a fusioneer. It is kind of a big deal for me and I hope all who are here. As you will note you have not been treated with any preference or assumptions in our helping out here, nor will that change.

If a woman, I will have to stop using the "old boys" appellation to my future reference to those having done fusion and activation.

I really enjoy when we have a winner fusioneer and they acknowledge, as you did, the fact that these forums really assisted you in your winning effort. I do hope you will stick around and go as far as you feel you can in the amateur fusion effort. I hope for this with all the winners, especially the elites.

Richard Hull

Emma Black · Post by **Emma Black** » Sun Nov 06, 2022 4:58 pm

Ha! yeah don't worry Richard at all, I've worked in the technical side of IT for many years and am very used to being the only girl. Things are changing slowly in the industry now though, but I amusingly remember being outnumbered 100 to 1 or 2 at various conferences. Before specialising in the security side, I spent a long time running various small IT & AV departments, the result being I know a little bit about a great number of things.

Something to think about upgrade wise, I would quite enjoy building a custom little transformer but need to figure out how that would look. The idea of having some huge amount of stored energy in a flashover situation is somewhat scary.

Post by **Richard Hull** » Mon Nov 07, 2022 10:39 am

I was never worried, just didn't want to do the elite writeup wrong. I have now updated the elite message. IT nerds, fusion nerds, and amateur science nerds are special people regardless of gender. All of us tend to go far beyond the norm into varying degrees of obsession once we latch onto an area that captivates us. Curiosity unchained and unbounded, coupled with the verve to do and driven by the "hands-on" imperative force us into projects few could accomplish.

Keep the great effort and reports.

Richard Hull

Jim Kovalchick · Post by **Jim Kovalchick** » Mon Nov 07, 2022 11:21 am

Emma,
Congrats on making the Elite section of the Neutron Club. Your work shows that you're applying first rate skills and knowledge while you are also learning. Thanks for sharing while on your journey. We are all better for it.

Regards,

Jim K

Emma Black · Post by **Emma Black** » Mon Nov 07, 2022 2:02 pm

Thank you for the kind words, I think the childish excitement of getting something that only just possible (at home at least) to work, or puzzling on a strange result is very addictive.

That valve is leaking to an unusable amount now so I have treated myself to a brand new one. While the fusor is down for maintenance I'm taking the opportunity to add some cooling blocks to the end caps. In addition to moving things around for more moderator space. Beam pattern is interesting and not what I expected:

Post by **Richard Hull** » Tue Nov 08, 2022 3:35 am

Beam patterns are indicative. symmetrical is the ideal. I would suspect 1. a sharp point at the copper gasket that is raised ever so slightly off the plane of the flat surface. Note: there could also be a sharp edge on the mating internal connecting receptacle or 2. A very minor and unsuspected leak at the gasket. Inspect that area of the gasket or knife end. something in that extension of the pattern to the edge is acting as an electrostatic beam attractor.

You might start a new thread now that you have you fusor fusing. This one is rather long now. Typically, each time something new comes along in your effort like this beam pattern you might start a new post with a title like "Odd beam pattern in my fusor" and let it develop. Now that you are truly one of the elites, we would rather your future posts deal with a single image or question in the various forums available that relates to the issue involved.

Remember, image du jour is not the only place for images. This beam post and its image might go better in the construction forum or in the vacuum forum. You will get the hang of this as you go. In short if you have something curious or puzzling with a image or not post to a forum that is related. In this manner it can be searched in a forum where folks might expect to find it and related subjects that also have images associated to the topic.

A posted thread should go on only so long as replies to it original thrust are helpful, valid and until the subject is effectively exhausted.

While you are a relative "newbie" you are also a very intelligent, capable and advanced newbie having done so much in so short a period of time.
Spread you wings and be a regular poster in such forums as seem most appropriate with an illustration or image there to instruct us or question us.
Remember, you can teach us a few things if you find something amazing and test our abilities to answer conundrums.

I attach a classic example of using the image du jour by myself to talk about hamfests and to show the booty of electronics and detection gear to be found at such events. As it is all about hamfesting and not the attached images it is an outlier which is posted to get electronics that might be used in fusor work at a bargain price, in person where you and touch what your are buying and bargain for it.

viewtopic.php?t=14581

Richard Hull

Emma Black · Post by **Emma Black** » Wed Nov 16, 2022 1:09 pm

Quick photo update: the re-build has done wonders for the cleanness of the chamber, aside from the beam its crystal clear and neutron counts are up.

Secondly - I've discovered that those Hoizon PEM cells are highly temperature dependant, as you would expect. Below about 17c not only do you get very poor gas flow but also, if left overnight, visible condensation on the outlet tube. 30 second very gentle blast with an heat gun to pre-heat works a treat while I'm waiting for the workshop heating to get going.

Post by **Richard Hull** » Wed Nov 16, 2022 6:52 pm

I have all of my fusor lab unheated as an attachment to my home with almost the same square footage as my home. I suffer from the winter doldrums so far as fusor work is concerned. I can work in my lab in the winter, but only in an isolated area where I use 20lb propane tank heater to take the chill off a small local work area at a bench area at the opposite end of the lab from the fusor.

I feel for your having to warm your PEM cell and enjoyed your report of its getting balky at 17 degrees. While I use only tanked deuterium, I am sure it is of interest to those using the PEM cell that you report some issues at lowered temperatures. I think most fusioneers tend to work in areas that are either in their homes or heated garages or out buildings. My large lab is just too big to heat in the winter. Propane has gone up. A $29.00 refill on my tank is now $39.00. Fortunately, I tend to only need 3 refills during the winter as my simple, non fusion experiments tend to move to my heated up stairs lab area within my home.

Richard Hull

Matt_Gibson · Post by **Matt_Gibson** » Wed Nov 16, 2022 7:51 pm

I abandoned the PEM cell due to condensation forming, bad enough that it made its way into my chamber. I now use bottled deuterium. A lot more $$$ but much much easier.

-Matt

Emma Black · Post by **Emma Black** » Thu Nov 17, 2022 6:48 am

Condensation that bad, you would think the membrane had burst! Warming it first seems to help a lot. I have been looking again at prices for bottled D2, quite like Richard's arrangement of filling a very small cylinder for each run to a known pressure.

We use woodstoves both for the house and workshop, the house being from the 17th century has terrible insulation, it takes ages for things to get warm. I also have a small diesel heater pointed at the main work area, which helps and is not too bad to run.

Post by **Richard Hull** » Thu Nov 17, 2022 2:18 pm

Keep warm this winter, for sure. When growing up, we heated with coal. Being the older boy, I had to get up at 3AM and throw a few fresh lump or two in the stove in our bedroom and shake down the grates so my younger brother and I could keep warm. Lots of bedding and blankets helped.

Running a fusor with cold fingers and toes is no fun. Still, having a large 700 sq. foot lab is more than most of us have, even if unheated.

All the best in future with your fusor.

Richard Hull

Emma Black · Post by **Emma Black** » Sun Nov 27, 2022 11:33 am

After more debugging with the PEM system, I hit a new record today 13.29 MRem/H showing on the Ludum at 35kv 4.5ma, which I make 385,410 n/s.
I'm hoping a few wall loading runs will yield the 500k mark, but we will see. 3600 CPM on the silver activation, improved slightly after messing around with the position of the moderator blocks. Have some Indium foil on its way to experiment with to, but not sure if the neutron output would good enough for a detectable activation.

On the PEM setup. I'm finding that these ground glass syringes start to get stuck at the merest suggestion of dust, which seems to result in air getting sucked into the system, presumably around the plunger. So swapped back to using the plastic ones and moved it all into a box with dust filters, that combined with warming the cell and using some silica gel to dry the gas all seems to be helping.

Post by **Richard Hull** » Sun Nov 27, 2022 11:11 pm

You are honing your operational skills. As they getter you results will follow. while fusion is technically easy with a fusor, the operation can be tricky which quickly falls into a well honed routine.

Indium is as easy as silver to activate. You will have no trouble. While also a good beta emitter like silver, Indium is a far better sturdy of gammas if you have a gamma ray spectrometer.

Richard Hull

Emma Black · Post by **Emma Black** » Tue Nov 29, 2022 1:40 pm

Yes its defiantly getting easier to operate, it's now very routine. After swapping out the lamps I'm using for primary ballast to higher wattage ones, what I think should be the 500k mark (blue line) was pretty easy to reach. We now have a bit more current to play with.

Likely don't need these at all, but with 4 bulbs in parallel it's good to be able to select the current limit by switching some off, all 4 on = 12ma, 2 on 6ma etc. While operating I keep the filaments just at the point of barely visible glow, in this state they are quite responsive to any undesirable increase in current and impossible to ignore.

Post by **Richard Hull** » Tue Nov 29, 2022 4:48 pm

Excellent report with graph showing the rise in neutron count. However, what are the details on the bulbs? What are their type? do they have a number or wattage, what are their rated voltage in normal use? key details? Where are they in your power circuit?

Richard Hull

Emma Black · Post by **Emma Black** » Tue Nov 29, 2022 11:56 pm

Useful details indeed missed there, my apologies. I'm using 400w 110v rated bulbs connected as below. They don't use anything like the full wattage to start to glow of course and as soon as they do the resistance ramps up rapidly. As I said though maybe not needed any longer or an ideal solution.

Post by **Richard Hull** » Wed Nov 30, 2022 2:43 am

So you have 1600 watt lamp load bank there that is switchable. 400 watt 110 volt lamps are big! Or did you mean 4 -100 watt lamps?

Regardless, it is pretty neat idea to have a visual variable ballast on the input side of the AC line. Cool. This is interesting and never put forth here.

Richard Hull

Emma Black · Post by **Emma Black** » Wed Nov 30, 2022 8:12 am

Yes they are massive! D-Cell for scale:

Emma Black · Post by **Emma Black** » Fri Dec 02, 2022 12:42 pm

First indium foil activation, this was a fun one and worked first time. 15mins in the fusor averaging 4.2e5 n/s 36kv 6ma.

I did a video of the first 15 minutes, compressed into 1 minute, the survey meter was on 0.1x. This is making me very much want to get a gamma spec setup.

https://youtube.com/shorts/CYHtGPmaW-8?feature=share

Emma Black · Post by **Emma Black** » Sat Dec 17, 2022 7:21 pm

Since switching over I've found these little 11l Hydrostick cartridges are great, no more messing around the pem cell. Constant pressure for as long as you need. The switch has sorted my condensation issues as well and neutron counts are up as a result.

Also built a simple arduino based stepper motor system for the leak valve, for some super precision control, which I need to install and test.

Matt_Gibson · Post by **Matt_Gibson** » Sat Dec 17, 2022 7:43 pm

I could be wrong, but this might be the first example of using the hydrostick for deuterium supply. Could you give some more details about it?

-Matt

Post by **Richard Hull** » Sat Dec 17, 2022 8:29 pm

Great work Emma! We look forward to your reports on this system and how it works out for you.

Richard Hull

Emma Black · Post by **Emma Black** » Sun Dec 18, 2022 5:20 pm

Matt - there was an excellent thread a while ago by Andrew Seltzman with a great deal of technical details on both the cartridges and the filling station.

viewtopic.php?t=13292

Dennis P Brown · Post by **Dennis P Brown** » Tue Jan 17, 2023 5:56 am

Really top rate work, Emma. I understand the issue with a 17th century house - part of my house was built in the early 18th century; yes, there was no such thing as insulation then but certainly horsehair plaster, semi-finished tree shaped beams using wood dowels and a few handmade metal nails are certainly there as I've found out the hard way!

Emma Black · Post by **Emma Black** » Mon Jan 30, 2023 7:31 am

A little update. Now that the power supply has been upgraded to support a higher voltage (60kv+) the feedthrough has started having arcing issues. This is not that surprising given that its only rated for 30KV.

After some investigation, this seems to have actually been on the air side and exacerbated by moisture/dust on the ceramic. It was a little better after a clean. Corona noise was still present over about 50kv though, so I am currently experimenting with an added oil tank to improve the feedthrough.

This is a simple ABS print. Now ABS starts to deform at fairly low temperatures, so even though there is a thick rubber seal between the fusor and the plastic for some thermal separation, the temps will still need to be watched.

Post by **Richard Hull** » Mon Jan 30, 2023 12:00 pm

Be advised, regardless of seals, USP mineral oil (non scented) has a way of getting around many types of seals. You do not want to come out to your fusor and find it and the table it is on or the floor covered with oil.

I had a real battle when making my own rolled capacitors for my Tesla coil work back in the early 90s.
Good luck

Richard Hull

Emma Black · Post by **Emma Black** » Mon Jan 30, 2023 12:46 pm

Thanks Richard, that sounds like a lesson learned from harsh experience!

The mess of it leaking is a bit worrying. 900ml of transformer oil all over the place would not be fun to clear up, at least it wouldn't short everything out like water would. Its been a week and nothing has leaked out so far.

If this functions ok performance wise I will look at constructing an all metal tank. Possibly by welding a SS container to a double sided cf flange, so I don't have to weld directly to the feedthroughs CF.

Post by **Richard Hull** » Mon Jan 30, 2023 5:38 pm

Sacrificial double sided flanges are a great solution if a weldment is needed on an insulator. I hope your seal holds. Some rubber/polymer seals can deteriorate over time and become permeable. The heating and cooling cycling, as you note, might speed up issues with the seal.

Richard Hull

William Turner · Post by **William Turner** » Mon Jan 30, 2023 8:31 pm

Viton or Buna N (Nitrile) should be fine with transformer oil. Viton A will have lower permeability with <30% compression.

Natural rubber and EPDM (very common with fittings intended for use with water) both are terrible.

https://www.coleparmer.com/chemical-resistance (Select Oils: Mineral)

Be careful with SLA/FDM printed parts, they tend to be porous and can leak over long periods of time without sealing (vacuum impregnation with a low-viscosity epoxy should work).

Emma Black · Post by **Emma Black** » Sun Feb 05, 2023 4:27 am

Thanks William, I was initially using a Vitron seal but I wasn't happy with the amount of force needed on the printed part to compress the seal against the top of the CF fitting. So I switched to using some high temperature automotive liquid gasket, it will be a bit messy to remove but I have a lot more faith it wont leak.

For the prints normally resin impregnation would be the way forward, however the thing was too big to fit in any chamber I had on hand. So version 2 was printed with a double wall and resin poured in-between.

Lab assistant for scale.

Emma Black · Post by **Emma Black** » Mon Mar 06, 2023 5:58 pm

A few little improvements:

After a good deal cropped up on ebay I bit the bullet and switched over to a little turbo. For now its just using the existing hose from the big diffusion pump and the adaptor that came with it, so far so good. It's connected to a bench 80v power supply and an analog 0-10v voltmeter to measure the speed. Compared with the with the cooling issues, startup time and power supply requirements of the old pump this little thing is so much more convenient to use.
The grey cable wrapped around everything is reptile heating cable, which I've been as a heater to help bake things out.

Second change was to switch over to two 901P vacuum transducers from the edwards pirani gauge I had been using before. Following several of the very useful posts on here I connected them via serial and was able to instruct the transducer to emulate the old unit. I think this just changes its output voltage curve but it meant I could use the my existing Edwards analog gauge controller. This simply involved adding an RJ45 plug to to the new gauge and it was plug and play from there.

Finally everything has been getting some TLC, the leak valve controls got a nice little enclosure, the variac got a new dial and a ton of cable management has been happening to overall tidy up a bit of the mess.