Morphing Air Conditioner into Autocascade System

[mytekcontrols]

Haha sorta as I pictured it. Looks nice, I love watching you dig in the foam

Yes its kinda like being a sculptor

I also had to specially grind a drill bit that would cut-out a plug (sort of like a mini hole saw) in order to make a spot for the return of the subcooler to enter the top of Cascade #3. Also kept a nitrogen bottle attached with the regulator set to about 30 psi to blow out any chips.

[n00b 0f l337]

Yep good idea. I would suggest a concrete drill bit, its more flat then anything, and if your good and have internal pressure, no need to worry about chips, you wont puncture till the last of last miliseconds and then you'll be through.

[mytekcontrols]

Never thought about that, but I can see how a concrete bit would be very similar to what I custom made.

Sorry to not reply sooner to your other question (too much thinking will short out your brain).

I should have some more pics tomorrow with the cold probe attached, and the case partially assembled.

[n00b 0f l337]

LoL your good man, its just fine, why not checkout the cascade thread of mine
Tons of pics for you there.

[n00b 0f l337]

Hey again,
Was pondering as I normally do about this stuff ( yep I'm a freak, laugh all you want, I love it ). And was thinking about condenser sizing, you had said it seems like you don't need as big a condenser since it's all done in the first third. So got me thinking, with well, why.

Odd theory, but let me know what you think.

In a given situation with a hot discharge temperature X entering a condenser, the refrigerant will condense when at that pressure it achieves the temperature needed to condense. In some situation A, this may require the full rated and normal condenser to achieve said temperature. Now in situation B, our autocascade, for a given temperature X of input, let's say the same temperature is achieved at the end, however, with refrigerant like R123, the temperature required to condense the refrigerant, is much warmer then temperature X, and thus it condenses much earlier in the coil. Thus when R123 or similar begins to condense, it transfers to a liquid obviously. Now as something changes phases there's an energy transfer as well, with condensing it gives off heat, thus making the area around it "hotter", and thus we perceive the hot condenser. This also now "enters" the coil as load much earlier, versus towards the end.
Now the rest of the coil you are still reducing temperature, however condensing is done much earlier of the refrigerants that will condense, and the rest are simply being cooled on there way to ambient.
Sound possible? Logical?

EDIT: Also! When in this top area of the coil, or earlier shall I say, the refrigerant becomes liquid, heat transfer also increases!

[mytekcontrols]

Adam -- I think this makes very good sense of the what is perceived to be happening. And yes heat transfer is increased, hence the reduced requirement for condenser area as compared to a conventional single stage unit for a given compressor mass flow.

This theory has been proven over the years that Polycold has been building AutoC's (you should see how small the condenser is on their air-cooled 1.5hp gas chillers).

I think it would be safe to say that at least a 50% reduction in surface area can be implemented (and maybe as much as 65% is possible).

[n00b 0f l337]

Of course, but then again why not go with a big one like a window AC anyway
But the reason I bring it up, is if your looking at a 200W heatload, how much condenser do you need. Your window AC has like a 1.7kw condenser!

[Raadster]

Michael what did you think about the COP gains and the capacity gains of using a SLHX with R290 as opposed to R22?

And in terms of condensor sizing, if you buy a Window AC may as well as you said Adam go with the condensor it comes with.

[mytekcontrols]

if your looking at a 200W heatload, how much condenser do you need. Your window AC has like a 1.7kw condenser!

Exactly! That is why if you really want to shrink the package down you can achieve this quite easily by reducing the fan and condenser size. But if the original AC unit size suits your needs, why bother?

Michael what did you think about the COP gains and the capacity gains of using a SLHX with R290 as opposed to R22?

What can I say, other then "Most Impressive", and I think it speaks for itself as to the performance increases that can be achieved with HC refrigerants.

[n00b 0f l337]

Exactly my thoughts

[mytekcontrols]

Well I got the actual cold probe, or what can also be called the final evaporator configuration for the AC-2 unit attached a few days ago, and have been testing the performance. Presently it is not in a vacuum (although the final application will be), and it is instead sitting in a box of fiberglass insulation.

On the initial warm start-up; the probe temp pulled down fairly good during the 1st hour, hitting -117C. But it was soon apparent that the rate of decent was beginning to level out, so I tweaked the charge just a tad bit with some more R170 and about another ounce of the R123. This got things moving again, and within the next 1 1/2 to 2 hours it got down to about -130C (this was with an 82 F or 27.8C ambient).

I've kept it running for about 4 days now, and it's gotten down to -138.5C on the probe when the ambient conditions drop to 68 F (20C) in the late evening.

Presently as I write this post I am seeing the following:

Comp Discharge: 77.0C
Comp Suction: 9.0C
Ambient Air: 26.4C
Middle of Probe: -134.2C
Comp + Fan Amps: 3.4
Low-Side Press: 30 psi
High-Side Press: 175 psi

I expect to see even better temps when the probe is in a vacuum.

Here's some pics (sorry for the lack of a temperature meter):

[n00b 0f l337]

Really incredibly impressive, every shot, every picture.

Did you make a different top?

[loonym]

That's a fantastic piece of work and amazing temps.

[n00b 0f l337]

Probably saved your shop quite a bit of money there too.

[mytekcontrols]

Did you make a different top?

Nope, stock AC sheet metal. Just added the 1/2" aluminum angle to create a place for the front panel to mount to (Having Front Panel Express make me a nice custom one).

That's a fantastic piece of work and amazing temps.

Thanks loonym Although to be clear, I didn't fabricate the actual cold probe and vacuum flange. These were stolen off of a faulty Polycold P-75 unit (see image below).

...Probably saved your shop quite a bit of money there too.

Yes I did save them quite a bit, considering that my little bitty AC-2 unit does everything and more then a properly working Polycold P-75, which carries a price tag of $10,000 new. However I will be getting reimbursed for the materials (approximately $850), and about a weeks contracting fees to convert the AC-2 into its new HMS configuration.

I know this doesn't sound like much in comparison, but it was never my intention to make money on this project, so anything is appreciated.

Polycold P-75 Unit:

[mytekcontrols]

The front panel as I said earlier is being custom made by Front Panel Express. I have attached a captured image from the program (free version has no export filters), and the actual construction file (you'll need to download the free program from Front Panel Express to view it).

Don't let the quality of the image fool you, the finished product from these guys is top notch.

The Mytek Logo is actually going to be a stick-on PC style case badge I had made by a company called Scotgold. Their pricing is pretty reasonable, and they'll do quantities as few as 10 pieces.

At this point in time, I'm just waiting for the front panel to arrive (ships 8/22/08), and then the unit can be buttoned up.

[gosmeyer]

Sweet, Real nice!!

[n00b 0f l337]

Very nice indeed! Not going to show off the discharge pressure though? Too bad, I like watching it with a buffer valve
Again great project, with the Polycold costing that much, have you considered making a few more and selling near half that?

[mytekcontrols]

Very nice indeed! Not going to show off the discharge pressure though?

No room for 2 gauges, so I had to compromise on the one. I picked suction, because it allows for a better resolution, and I think shows the unit's overall pumping condition better than the high side alone. Yes it would have been cool to see the discharge gauge dip each time the system buffers, something it will do at least 10 times on a warm start-up.

...with the Polycold costing that much, have you considered making a few more and selling near half that?

No I've had my fun, and hopefully helped others learn more about autocascades, that's all I really wanted. Of course the info on how to do it is now available if someone else wants to step up to the plate .

Polycold no longer makes or sells the P-75 model, and instead concentrated on it's 1hp brother (same basic design, and dimensions, just a bit more mass flow --- 1hp vs. 3/4hp). So I'm sure there is ample opportunity for someone else to enter the market, especially considering the AC-2's physical size reduction, and about half the power consumption of the P-75. I'll be sure to post a picture later on of the AC-2 sitting next to a P-75 for comparison.

[[XC] gomeler]

This thread has been amazing to say the least. The work you have done here will be immensely useful for years to come. If it hasn't already been added to the stickies, I'd like to see it added as it's the first 100% successful autocascade I've seen on these forums built by a professional.

[mytekcontrols]

Thanks Gom

And to just let you know (and everyone else as well), I am hoping to create a nicely formatted pdf that covers this entire build, so as to make it easier for someone to duplicate my results. It'll take a while, but I am hopeful that it'll happen by the end of this year.

Edit: (10/10/2009): Sorry, but it looks like I'll never find the time to do the pdf.

[n00b 0f l337]

Ya know what,
I'm going to put mine together, got an evap to spare on the way from Teyber I think,
Have an HX stack all brazed up and pressure tested, have spare condensers, just none exactly what I'd like, and total dimensions aren't as beautiful as I'd like. But why not.
Mytek expect a thread soon after I finish and ship the Mini Window AC Cascade and one other build, hopefully will have this then done before I ship away on the Sept 1st.

Expect:
Dual condenser (incase one isn't enough), of 8.5 x 10.
SC18CLX.2
HX Stack
Teyber stepper
24-30" flex line (have to order some Flex from Ron, will do that now).
Buffer valve and such as well.


*Yep ordered Braid from Ron, two gauges, some more captube, some more its and bits, an expansion tank of decent size empty volume bout 9" tall 4" diameter rated very highly on pressure. So on it's way, we'll see how this cooks out , I'm thinking a "tower" as before*

[mytekcontrols]

I'm going to put mine together... Mytek expect a thread soon after I finish and ship the Mini Window AC Cascade and one other build

I'll be looking forward to it

[n00b 0f l337]

Ya know whats odd, Frosty Freeze, an ebay seller, has cans of R114, for like $60 a can, its like an 8-16ounce can. Though you couldn't even sell that anymore!

[s0lid]

U have not created an a machine! That is a true monster dude!

Btw when u add cpu evap to this autocascade monster ?