Everything I have read says to size the accumulator to contain the entire charge. I can't tell you how to size the suction line HX only that the 42mbtu unit I am making now has one about 8" long. It is a 3/4" pipe inside a 7/8" pipe, IIRC. I got some of the parts from an old westinghouse unit which was working well after 30 years but the case rotted out.

Mikesolar,

Yours looks to be the type of HX i plan on using. I'm with AC Hacker on this one, it seems MASSIVE! How is the heat transfer on that behemoth? It's half the size of your compressor... I understand about the accumulator, but are your coils that mismatched? Or is it sized for max heat transfer? Or do you pump down every cycle?

If I could get another 8 deg c out of my unit, that would get me to -13 c (8 deg F). This would cover my heating needs in Kentucky for 363 days a year!

I have been reading up on the research surrounding these units, which claims that a 10-15% gain in capacity and efficiency is possible, as well as increased system stability. Also, they seem to perform the best at high temperature spreads. Combined with a TXV to keep the evaporator near saturation, it would seem to be a winning combination.

Here's another paper, which looks very similar to Mike's rig:

http://www.annex32.net/pdf/presentat...008_Stene2.pdf

The unit I envision will be functionally plumbed like this:

Take cover off of 1982 Window shaker A/C, chop compressor suction line. Add 2 meters of suction line, coiling excess suction line tightly along length of liquid line in counter-flow. Braze. Recharge, replace cover. Run unit with performance gain.

In actuality, it will most likely be much more complicated.

The difference between that one and mine is that the old westinghouse unit is that the TXV actually meters the liquid line to the condenser NOT the evap so is called a subcooling valve. The evap becomes a flooded Evap which took me some time to get my head around but my mentor has been making them this way for 25 years and they work very well.

Other than that, the idea of the the SLHX is a good one and should be combined with a bit larger Evap and condenser but sizing those, especially when they are air coils is a real art. Even the manufacturers do more trial and error than you would expect and they don't like to give away their design ideas. I've tried to get it out of them without success.

Jeff, how long of a HX would you put in? I don't think there is much advantage to anything over 6-8".

Aha, so that's why you're running such a large accumulator. It's also a flash tank, or a "slobber box". It took me a while to decipher your plumbing, the reversing valve is the same color as the compressor shell! That and your full-wave bridge rectifier check valve setup threw me off for a minute. I'm somewhat confused with your TXV description, however. The condensor is not fed liquid.

The metering device in flooded systems is usually a float in the flash tank. When the refrigerant in the flash tank drops below a certain level, the valve opens to maintain that level. When the flash tank fills past this level, the valve closes. Kind of like a toilet. The flash tank sits above the evaporator, and gravity keeps fluid in the evaporator. This is obviously not what you have.

I found info about the westinghouse valves here:
http://sporlanonline.com/literature/10/210-60.pdf
http://www.refrigeration-engineer.co...bcooling-valve

A subcooling valve is the opposite of a TXV. Instead of measuring superheat against suction pressure, it measures subcooling against head pressure. Instead of regulating the amount of liquid in the evaporator, it regulates the amount of liquid in the condensor. Since you have a flooded evaporator, this makes perfect sense.

Your system is plumbed closely like what I am planning. The main differences being:

I will have only one SLHX, the big can in yours.
TXV bulb will be on the suction line just prior to entering the big can.
Two metering devices, two check valves.

Yup, I made a booboo. It is metering liquid which goes to the evap and the bulb is measuring the temp on the discharge side of the condenser before the two HXs. The desired subcooling is around 15C or even more. superheat is only a couple of deg.

I think I see what you are doing. In this way, you have a different TXV for each direction.

New Regenerative Cycle for Vapor Compression Refrigeration
 

Just came across this juicy paper, courtesy of Uncle Sam:

New Regenerative Cycle for Vapor Compression Refrigeration

-AC

Yes, I will have separate metering devices for heating vs. cooling. I may use two txv's in this one, or I might keep the cap tube for cooling and for a method of easily charging the unit. The unit will essentially function as a larger R22 type split heat pump with the Hx in the suction accumulator.

This guy calls it a liquid suction heat interchanger, why I don't know.
Liquid-suction heat interchanger

The main purpose of the exchanger will be as an accumulator. With the unit I have now, I can only push the superheat down to about 15 deg F before the evap starts burping foam. It seems that oil dissolves really well in R290. With nearly no distance between the evap and the cxr, the foam was going straight to the pot.

The pros all told me that this always happens with room ac units if the superheat at the cxr inlet goes below 30 deg F. BTW, I find that the forum experts will tell you anything you want to know if you stay IN THE BOX. This includes making up a story about a generic something you are working on somewhere. It also means talking in their language. In your case, I would reference the westinghouse unit at a brokedown palace. Homeowner Joe Dirt inherited the place and can't afford a new unit. Only then will the experts bare their souls on your problem.

Jeff, what was the original refrigerant? If it was R410 then the oil needs to be changed to mineral oil from POE. The foaming is one symptom, moisture in the system is another. POE is very hydroscopic.

Are you referring to that US based HVAC talk forum? 30F is a lot of superheat, 20F being what Copeland uses in their specs and typical is way less.

Do you know of another one?

-AC