The Phantom LN2 Evap/Pot
Special thanks to Gigabyte for the MA790FXT-UD5P
And also AMD for the CPUâ€™s
And a special thanks again to Aaron Schradin for all his help and ideas !
The Phantom Dual Wall Clear LN2 Evaporator â€“ 2nd Encounter
I had a chance to make some addition runs with the Phantom pot, I had a set of shorter tubes made ~6â€ in length to do these runs and they work much better than the longer 15â€ tubes. All though the 15â€ tubes do work fine they are really to long for anything other than a LN2 show. I ran the 15â€ tube without the inner tube also and it worked fine.
Running the longer tube the setup looked like this. I wrapped a piece of sheet foam around the tube and cut to fit, then a layer of blue shop towel. This setup worked very well but was a tad tall to be sitting and trying to pour LN2 in the tube so I had to stand up often. You could really load this tube up with LN2 tho and run for a long time without even looking at the tube.
Here is that setup
When the shorter tubes came I ran them first with just the outer tube, and this is that setup.
This proved to be a very good setup and held temps very well, drop in temps from .2c were the norm for me with the probe mounted on the pot. It was very stable, could run for a long time without reloading, plus frost buildup was well under control. In fact the board would start to frost up well before this pot did.
With these kinds of pots swing is not a real issue and as long as you have the pot at max temps you are good to go. Simply keep LN2 in the pot and temps are stable. It required just about no baby sitting and was very nice to use.
It had been said to me to just keep LN2 to about 1â€ deep or near the base copper for normal running for best temp control. I did not see any real difference in keeping levels at the 1â€ level or much deeper. This pot was that stable in temps.
When you are at max LN2 temps you can hear the difference in the sound the pot makes, it goes from a rapid boiling while pulling down to a somewhat calm slow boil that makes very little noise. The bubbles are much smaller and in this pot because of the design of the base you can see a flow of LN2 as the level drops to the top of the base. I have tried to show this in the video that I did. You will see the center of the base boiling faster than the edges. I believe that this is because of the massive surface area of the base, the way it was machined to channel a flow of LN2 to the center and the very large number of sharp edges which heat from the CPU radiates to these edges causing flow.
This is how the single outer wall looks before setup
Now letâ€™s look at the Dual Wall prepping and setup
All the parts laid out and ready to be put together
Teflon tape is wrapped around the inner tubes threads and around the base threads. Wrap tape around many times to assure a nice tight fit and no leaks.
Set the inner tube and washer assembly on top on the base.
Now screw the outer tube in place. Tighten everything up good and snug to prevent leaks.
I then took this completed assembly and filled the outer ring with water to check for leaks. I only had 1 small one which was easily fixed by just tightening the inner tube just a bit more. I then let this sit for awhile just to make sure I had no leaks. All was good :clap:
Now insulated and installed on the motherboard.
Insulation was very easy for both methods. For the single tube setup a simple foam sheet cut to fit work great. It could be removed very easly by simply sliding it off to go for dual wall setup.
As you can see I just used two layers of foam with the circle are cut out for the pot to fit into. This could be cut to fit any motherboard easly.
With the single tube setup the layer of foam around the tube really didnt interfer with any parts on the motherboard, except for the first ram slots.
The foam I used was 1/2 inch think which insulated well and provided plenty of clearance for parts on the motherboard.
Here is a short video of the Phantom in operation.
This run was about 4 hours long and ran just fine throughout the whole session. The only problem was it began to start frosting over after about an hour to an hour and a half. It just let it go to see what was going to happen. As you can see it became very, very frosty by the time I finished up. The motherboard was getting very, very cold and was becoming very hard to keep stable OCâ€™s towards the end.
Watch in the video for the LN2 in the base and how it looks from a boiling to down to no LN2. This was not easy to video as light was not that great and focus was hard to maintain but I think it came out ok.
Some additional photos of the session.
My conclusion is that with the 80 proof alcohol that I used here my run time with a clear viewing time was somewhat limited to maybe one to one and half hours. It started frosting up and the alcohol began to freeze close to an hour into the session. After awhile the alcohol froze up solid and frost became rather heavy. This did not appear to impact the function of the pot in anyway other than you could not see the insides. If a higher proof alcohol is used such as Everclear which I cannot get in California I understand that the freezing of the alcohol takes longer, I am not sure if that would keep the frost away or not tho. A blow dryer might have helped and could be useful in a demo or event type of setting to keep things clearer, but frost melting off of the pot might become an issue.
From what I could tell the difference between the two setups was not great in running, in fact they both seemed to work the same to me. So getting this setup with dual tubes has its advantages of being run with just the outer tube for normal use, then being able to use the dual wall setup for events and thing of that nature. It really works out very nice.
Use of this pot with AMD chips is very simple and this pot performs very well.
I do have a few mods in my mind for an event type setting such as lights and what not, I will update this thread when I get around to those
I think with a pump and radiator you could keep the outer layer warm enough to avoid condensation.
Maybe even cool the NB or a GPU while your at it
I freeze things.Ask me about vapor phase change if you have questions, I'm always around.
How much does surface area over volume of copper actually take part in the equation used in the design?
The man never feels the want of what it never occurs to him to ask for. - Arthur Schopenhauer
Thanks to a spam post, I came across this thread.
That is one awesome pot!
|Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)|