Morphing Air Conditioner into Autocascade System

[n00b 0f l337]

Are you sure your not having a problem getting r290? Sounds almost like your adding air to the system.

[mytekcontrols]

Adam give me a break

No I verified that I was getting liquid propane spurting out of the cylinder when held upside down. And I charged it into my evacuated 3x9 addition tank using an electronic scale, being sure to purge the lines with propane to get rid of any air. Also had a brand new supco filter dryer inline to absorb any water that might have been in the propane. Stuff was getting cold past the first phase separator, and I'm sure that if I had added more that I could have seen the temperatures drop in the first stage, but unlike the R123/R22 combo, the R600/R290 combo tends to reflect the pressure of the R290, which can get quite high when a saturated liquid is present (the R123 actually reduces the pressure of the R22 or R410a when mixed). This wouldn't leave me much room to add the other gases.

I will leave it up to others to puzzle the HC mystery.

[n00b 0f l337]

Heh thats odd indeed, I'll have to get back to testing my autoc once I get more time.

[mytekcontrols]

Heh thats odd indeed, I'll have to get back to testing my autoc once I get more time.

So you have been making progress on this project. I was beginning to wonder. Pretty stealthy I'd say... NOLTECH = AREA 51

Any more details you can divulge?

[n00b 0f l337]

Not going to use a rotary

[yngndrw]

Not going to use a rotary

My god, the flood of detail is going to drown me.

[mytekcontrols]

My god, the flood of detail is going to drown me.

Not going to use a rotary

Any reason not to? From my recent experiences, it seems like a rotary makes for an excellent pump. Twice as efficient as a reciprocating, able stand much higher static pressures, great start-up torque, and compact size. What else could you ask for?

Perhaps you're going to use a Scroll instead?

Hey wait a minute... AREA 51... recovered alien technology... Hmmm....

[godmod]

A reciproc can fit into a normal SS case. Maybe he wants to build an autocascading system in a SS case.

[n00b 0f l337]

Any reason not to? From my recent experiences, it seems like a rotary makes for an excellent pump. Twice as efficient as a reciprocating, able stand much higher static pressures, great start-up torque, and compact size. What else could you ask for?

Perhaps you're going to use a Scroll instead?

Well, simply put, height, though not so much in this case, but also noise. Rotary's are lovely and all, however though this version will be much too big for a single stage case, I do want to get it down to that sort of height.

[n00b 0f l337]

Damn lost the auction on that Danfoss SC12CLX... Maybe rotary will occur in this build after all.

[teyber]

what was your highest bid and how long left?

[n00b 0f l337]

Um bid at $45, someone took at $52.50, both with 4 seconds or so left?

[Xeon th MG Pony]

Bumping thread due to sheer awsomnes of it!

[mytekcontrols]

Thanks MG

This also served to remind me that I still need to post an update on this project.

What's been happening:
I did quite a few tests with R410a, until I finally decided that it just isn't anywhere as good a refrigerant as R22 (of course I kinda suspected this all along, but now I'm completely convinced). Although I did get some interesting results by eliminating the argon, and boosting everything else (almost got it to do -90C at 225 watts).

Anyway I shelved it for quite a while, having other projects needing my attention. But as of last week, the AC-2 unit now has a Sub-Cooler, and is undergoing tests to see how cold it'll run (as of today; -139C under a non-loaded condition, and about -115C with 50 watts applied --- probably close to the maximum load). It wont suit a PC cooler application, but it'll do just fine for use as a water vapor cryo-pump on our shop's helium mass spectrometer (5-10 watt load).

I'll be putting the finishing touches on the case, adding a front panel, and basically wrapping up this project over the next 2 weeks. It was great fun building this little pup, and I hope you guys enjoyed it as well. I'll be posting pics and specs in a few days, and will also show you the final form in a couple of weeks.

[n00b 0f l337]

That we did, that we did.

[mytekcontrols]

I think I am done with the refrigerant charge tweaking.

Here is the Cool-No-Load results:

[Raadster]

So Michael why is it you still use R22 instead of R290, is it because oil return isn't an issue or the flammability of propane?

[mytekcontrols]

Flammability is not an issue (already using ethane in the charge). It is more the fact that I haven't had good success with using R290 alone, or in combination with R600. Or for that matter, R600 alone.

I figure if I've got to use R123 anyway, at least in order to obtain the best overall results, then I may as well use it in combination with R22. Yes there probably is an HC substitute for the R123, but I don't have anything on hand. And we just have a ton of R22 that was recovered from an ice skating rink for free, so I really can't beat the price.

[n00b 0f l337]

Your early charges with R600 in this thread point to good performance though, then you found a blocked captube. Is it possible that too many HC's makes purification hard? Or is R600 too low a BP? Though its only 0C, and r114 is very ideal...

[mytekcontrols]

I have attached what I certainly hope is the final piping diagram for this project (assuming I haven't made any errors). It is a vector pdf, since the resolution of a bit mapped image would probably not capture the detail.

It is a modified version of earlier diagrams that I posted, and shows the AC-2 unit in its new application dress. That being a water vapor cryo-pump for a helium mass spectrometer.

Please note that the offset as shown in the Final vs. the Sub-Cooler captube coming out of the bullet strainer is intentional. Having the SC captube slightly lower then the Final captube insures that the Sub-Cooler always gets fed some condensate, no matter how poor the quality or quantity of what is available. This makes sure that the Sub-Cooler is always trying to operate in an active mode, and thereby continues to create an environment where the Argon will dissolve into a sub-cooled R14 stream that runs through it.

[wdrzal]

Polycold wasn't purchased for 22 million dollars in 2003 for a outdated cfc/hcfc mix that's no longer usable.

The present invention overcomes the need for using CFC refrigerant mixtures in a refrigeration system by utilizing refrigerants R14, R134a, R508a or R508b, R142b, and R740 in a component mixture. To achieve desired properties, these refrigerants may be used in a special mixture.

[mytekcontrols]

Nol -- good observation.

Come to think of it, I probably didn't explore this avenue as well as I should have, but I did do a follow up test with R600 after fixing the captube, and was still seeing some strange behaviour. At the time I attributed this to too good of separation within the various phase separators, possibly causing a starvation effect to occur in the later stages. But then again, I was also trying to make a 4 stage autocascade work more like a 3 stage at the time. Now that the unit has a sub-cooler, and I am using a much greater quantity of Argon (I'll post the final charge soon), I am obviously using all 4 stages (-140C). Perhaps R600 and/or R290 combo would work well in this situation.

Perhaps you'll have better luck with your AutoC project.

[mytekcontrols]

Good point Walt. But Polycold has access to refrigerants that I simply don't.

Polycold charge as stated in their patent:
R236fa, R125, R23, R14, Argon (EDIT: forgot to list the R23)

Requires POE oil, and extremely good oil separators (Polycold had some gigantic ones custom made to their specs).

EDIT: BTW; I no longer work for Polycold, and this project was not an attempt to make what they have made, it was simply meant to serve as a tutorial on autocascades (which I think it has done quite well).

[wdrzal]

EDIT: BTW; I no longer work for Polycold, and this project was not an attempt to make what they have made, it was simply meant to serve as a tutorial on autocascades (which I think it has done quite well).

I realized that, and it was a great build and informative.

Getting below -120c oil becomes a major issue. Thats why we need a new compressor design.

Maybe a 3 stage axial oilless design......30 to 40 bar pressure differiantal @ 40,000 rpm.

[n00b 0f l337]

Interesting with r236FA, boils at -1.4C. Then a jump to R125 which is rather high pressure, almost like r410a.
I think from a boiling point sense, they've tried to make like a r114 with r236fa, but then are skipping the r23 like refrigerant in favor of r125 saturated with r14 maybe?


EDIT: Looking over and overlaying some charts, it would seem that R236FA is well, hell almost like butane exactly, except three times the mass, and a tendency to really mess with oil. That's odd indeed.
Wish you had another one mytek of the AC2 to work with, could with your three phase seps, probably try r600, r290, veryyyy small amount of r170, then jump to r14, and maybe some argon.