Morphing Air Conditioner into Autocascade System

[n00b 0f l337]

And the temp at the first strainer is really low! Even loaded! Never expected to see that...

Theres the magic

And one other thing I noticed... about your HX, you're condensing upwards in your HX. Doesn't seem logical to me

Is he? Never looked, though I've done mine condensing down.

Gom, I think he would probably have to shorten captubes a bit more then. As to the high pressures, do you have any room? How about hollowing out the center of the HX foam a bit and adding another one of those expansion tanks.

[mytekcontrols]

But why the need for a bigger expansiontank? Having high static pressure?

Yes!!! I need more gas to start with then can be comfortably accommodated in the tank I am using when it warms back up to room temperature. If it gets too high the compressor will have a hard time of starting up, and the motor will draw too many amps, thereby exceeding it's rating.

And one other thing I noticed... about your HX, you're condensing upwards in your HX. Doesn't seem logical to me

The velocity of the discharge stream relative to the expansion (suction) stream is much much higher. So there is hardly any chance of liquids being hung up on the up hill discharge circuit, but there would certainly be a potential problem on the suction circuit if it were opposing gravity, such as poor oil return.

When you do your tests would it be possible to load it up to 200-300w? I imagine the temps will drop considerably but would it stabilize with more R-170 mixing into the higher stages?

No, unfortunately I think anything much past 175 watts would be too much for the little compressor to take. Also I would expect to see the return temperature suddenly heating up to the point where the heat would not be pulled out fast enough before the heater (resistor) would fail. See chart below; which shows the unit before adding extra refrigerant to enable it to take on 125 watt heat load.

[n00b 0f l337]

Well very good Mytek.
Would it be possible to get -90C or so for 200W? With that compressor.

[[XC] gomeler]

Very cool. Thanks for that Mytek, can't wait till you add the larger expansion tanks.

[n00b 0f l337]

Also what is your static now?

[n00b 0f l337]

The velocity of the discharge stream relative to the expansion (suction) stream is much much higher. So there is hardly any chance of liquids being hung up on the up hill discharge circuit, but there would certainly be a potential problem on the suction circuit if it were opposing gravity, such as poor oil return.

Oh really? Ah well crap, I'm plumbed that way thinking then liquid won't flood back as easily.

[yngndrw]

Fantastic work.

Just a couple of questions:

Firstly how does the efficiency of this autocascade compare to a well build standard cascade, at the same loading / temperature tuning ? (Heat moved Vs Compressor power.)

Secondly, looking at a chart the boiling point for R-600 is -0.4C. Sorry if I've missed this but it's a long thread with a lot of detail - How does this help the autocascade ? Would R-600a not be better, seems as it has the same molecular mass yet has a boiling point of -11.8C ?

-Andrew.

[n00b 0f l337]

It helps lower pressures and a bunch of other functions. Search for auxillary condenser.

[yngndrw]

The closest explanation I could find was: http://www.xtremesystems.org/forums/...er#post2110016

I understand that, but how does lower the condenser's temperature ? The R-600 isn't evaporating in the condenser ??

Cheers.

[n00b 0f l337]

Nah, it condenses very easily and in quantity, this helps give off heat in the condenser, and serves as a big sponge for other refrigerants to seep into.

[yngndrw]

Sorry if this is going off topic, but seeming as I've done Physics at A Level and I'm still clueless as to why it has this effect, I feelquite dumb at the moment. :P

What I basically don't get, is how it condensing helps remove heat from the rest of the mixture ?

[n00b 0f l337]

Because it ends up helping move heat down the stages as far as I can see.

[mytekcontrols]

Also what is your static now?

I hooked up a standard recovery tank to it before I it shut-down yesterday, and today the static BP = 55 psig.

Edit: My target static BP would probably be to keep it at 125 psig or lower (need to test maximum safe start-up abilities of compressor). So I'll need to calculate the tank requirements based on the present volume (inclusive of the added recovery tank), allowing for a BP at least 2.25 x the current pressure.

Oh really? Ah well crap, I'm plumbed that way thinking then liquid won't flood back as easily.

I'm not saying that it wont work upside down, but it depends to a large degree on the volume of your suction side, as well as the size of your compressor. However in an ideal design you do want to pay attention to this, especially if you are using your 1st phase separator to do most of your oil return (as I am).

I understand that, but how does lower the condenser's temperature ? The R-600 isn't evaporating in the condenser ??

Just like Nol said, it acts like a sponge for other refrigerants to dissolve into. Predominately this would be the R-170 in the system. When this is later evaporated, it'll do so at a temperature somewhere between R-600 and R-170. Think of what happens when you mix refrigerants to create something with a unique boiling point, same as what is done for R-410A a substitute for R-22. It also seems to have the added affect of enhancing the heat exchange in the air cooled condenser, think of it as a liquid heat sink in this case.

Edit: It also evaporates within the compressor, pulling heat out as it travels to the condenser.

Firstly how does the efficiency of this autocascade compare to a well build standard cascade, at the same loading / temperature tuning ? (Heat moved Vs Compressor power.)

I think the best way to answer this would be to have some of the standard cascade builders chime in, on how many total compressor amps it would take to duplicate the performance I am seeing with only 3.7 amps. Also I only have 1 fan, whereas a cascade would probably have at least 2 fans (or a single larger one being shared by multiple condensers). the additional or larger fan would also contribute to the amount of amps required to run the system.

Would it be possible to get -90C or so for 200W? With that compressor.

Yes, if you only use 2 cascade condensers (I'm using 3), increase the captube flows slightly, and charge with less R-14 and more R-170.

[{.bLanK} GoD]

Very nice Mytek. That man deserves a DB.

For comparison, My new cascade atm will do,
100w@-93°C and draw 3.35 Amps (230V = 770w)
150w@-87°C and draw 3.4 Amps (230v = 782w)

[mytekcontrols]

Very nice Mytek. That man deserves a DB
For comparison, My new cascade atm will do,
100w@-93°C and draw 3.35 Amps (230V = 770w)
150w@-87°C and draw 3.4 Amps (230v = 782w)

Thanks for the info, and I guess I should list mine as:
125w@-102C and draw 3.7 Amps (115V = 425w)
To give a better comparison.

BTW, what is a DB

[n00b 0f l337]

So lets see, if we take an average of pointblanks, call it 125w @ -90C, we can see that Mytek's does -12C better, and draws a bit more then half the energy. Autocascade really is quote amazing isn't it. Whats the max load you can hold though blank? And I'm surprised you don't have much better results coming out of that cascade.

[mytekcontrols]

Yep, about 1/2 the energy. I think this gets even better when designing for colder temperatures (-130C to -150C).

I should mention that there was a change in the final captube sizing I used on the AC-2 unit. I made a last minute decision, and increased the flows on the last 3 captubes (decreased length). Here is a chart with what I used:

[mytekcontrols]

And here is the heat exchanger info, as well as specs on the 115V LG QA090C rotary compressor I used (notice the rated load current of 5.2 Amps --- I still have some breathing room):

Edit: I gave an average spec on the length of the HXC's. In reality AUX and Cascade #1 were slightly longer, and Cascade #2 and #3 were slightly shorter. Increasing the length of all the heat exchangers would probably yield even better loaded temperatures. Keep in mind that the pressure drop, which is not beneficial, will also increase with added length.

[n00b 0f l337]

What do you think about plate heat exchangers?

[mytekcontrols]

I think they are great, and have better heat transfer then a tube-in-tube design for a given amount of area

And I also think that the AC-2 project would have been significantly easier to fabricate utilizing plate heat exchangers.

Of course Nol you know that already

[n00b 0f l337]

Hahah yes, any idea to sizing plate wise? I think I need to get some more cash sometime to pickup a window AC, I really like those condensers.

[yngndrw]

Thanks for explaining about the R-600, I think I get it now.

I'm pretty supprised to see how efficient this is - I was expecting to see that cascade compressors would not be strained as hard, hence would draw less.

I'm wondering, why are autocascades so much more efficient ? I'm sure that heat exchanger inefficencies can't loose about 400Watts of energy, so where is it being lost ? (Based upon the comparion statistics that PointBlank God posted.)

Autocascades look like they could replace Single Stages for 24/7 use.

Wish we could pick up window Air Con over here for as cheap as you guys can. :c

[[XC] gomeler]

Well a large portion of an autocascades high-side is operating at sub-zero temps so the pressures the compressor is fighting against are much lower. The suction pressures are also much higher than traditional single-stage/cascade loops, so the compressor is running in a very happy state. That'll partially explain why it's got such great efficiency. On the other end, you are using the saturation of gases within liquids to help with the temperatures, from my understanding you need specific blends to do this, certain gases just won't get along as well. Oh, did I also mention, hydrocarbons rock?

[n00b 0f l337]

They can replace single stages, most definitly, but costs of parts and gases are higher. Now why are they better off? Often time the compressor is cooled much better, operating pressures can be much lower, and your not trying to cool the next stage down from ambient

EDIT:

From my understanding you need specific blends to do this, certain gases just won't get along as well. Oh, did I also mention, hydrocarbons rock?

Now not necesarily. From what I've seen almost all the failing autocascades here at XS are only missing the "warm" boiling point refrigerant basically, and often the HX sizes.

[mytekcontrols]

New thought...

I have been pondering why it took so much more refrigerant to get this unit to work than I originally thought. I have been staring at the data collected, and what stands out to me is that that strainer#1 (coming out of the first phase separator) is running colder then expected, and most unusual is that Strainer#2 and #3 tracked pretty close together until I began flooding the final stage. I think I have a plugged captube. And it's either #2 or #3

So I'll have to dig into the foam insulation to expose just the captubes, run the unit from a warm start-up, and feel each captube to see if it is flowing (does it freeze water vapor).

If one of them is plugged, then who knows, I might be able to get even more out of the unit.