Friday, December 30, 2011

CryoBUG breaks the -150°C barrier

Yes you heard that right, CryoBUG not only made it to my target temperature (-150°C), but it actually surpassed it by a small margin.

-151°C Evap-In, -148°C Evap-Out (not shown)

22 psig Suction, 235 psig Discharge

Evaporator Inlet (final cap tube feed)

Now to be honest, my temperatures and pressures were not completely steady in all cases (although some temperatures such as the evaporator outlet were rock solid at -148°C).

Here's the specs:
Compressor Suction Pressure: 17-23 psig
Compressor Discharge Pressure: 235-240 psig
Compressor Current Draw: 2.65-2.80 amps
Evap-In Temperature: -149 to -154°C (average -151.5°C)
Evap-Out Temperature: -148°C

The compressor discharge temperature was pretty steady at +57°C, and the Liquid Line equaled +20°C with an ambient room temperature of 19.5°C.

Now how did I do this?

Well I made a few hardware modifications to the HX Stack. First I chopped off about half of the Auxiliary Condenser's length...

Then I re-established the liquid and suction connections that connect to the condensing unit...

And then I took the coils I chopped off, trimmed off a small amount, and added this to the end of the Cascade Condenser's SLHX section, thereby creating an even longer final sub cooling stage..

Extending the Cascade Condenser's SLHX

And finally I brazed everything up, and got it ready to go back into it's orange Home Depot bucket.

Modifications Completed

Boy this thing is getting pretty ugly with all the cutting, hacking, and re-brazing going on. But it is a bread-board prototype, and that just happens to be it's fate in life.

Now back to the test results...

Well I started off with basically the same charge as I used before this particular set of modifications. But after seeing it take longer to pull-down, and also seeing some jerky jumps in temperature at the Phase Separator, I assumed that it was the result of an insufficient supply of refrigerant condensate. So based on this conclusion, I decided to up the R-134A by another 30 Grams. This really made the difference, and the temps quickly dropped, bringing the evaporator down into the -143 to -144°C range. So seeing things starting to level off, I gave it another 20 Grams of R-134A. This helped a tad bit more, but then I started seeing a bit more fluctuation in the pressures and at some temperature points including the evaporator inlet. I then went on to add a tiny bit more R-23 without any noticeable benefits.

My final addition was to add about 5 psi more Argon, and although this did jack up the compressor discharge pressure, it really delivered and pushed me past -150°C.

Aside from the continued fluctuations, I think I'm getting pretty close to a final design. And most likely with the extra mass of a machined copper evaporator fed by the final cap tube, the temperatures will tend to even out. I also think that I will add back to the Auxiliary Condenser half of what I previously removed.

Thursday, December 22, 2011

CryoBUG hits -140°C

Well hey I did it today, I passed another milestone, that being the -140°C mark!

Evap-Out (left), Evap-In (right)

So here's the charge that I used:

100 Grams R-134a (50 psig vapor pressure)
50 Grams R-23 (85 psi vapor pressure)
50 Grams R-14 (75 psi vapor pressure)
Added 35 psi of Argon (not possible to weigh it in)
Static Balance Pressure = 245 psig

Start-up initially hit a discharge pressure of 344 psig, with a discharge peak of 360 psig within 10 minutes, after which it began to drop. Below I have shown a graph of the Evaporator In & Out for temperature versus elapsed time in minutes to get a better idea of performance with this new charge and hardware configuration (previous cap tube changes).

As can be seen, the first 40 minutes yields an extremely fast pull-down in evaporator temperature, and then for the next hour or so the pace slows down quite a bit. And when we get to the 2 hour mark, things have pretty much leveled off.

Final running pressures at the end were 11 psig suction and 187 psig discharge, with a compressor discharge temperature of about 55°C. Current draw was actually quite low, at 2.5 Amps (compressor RLA = 4.3 Amps).

This thing is running beautifully, and well within spec for this R-410a rated compressor. I am continually amazed at how well this Fuderer hybrid performs. And I can't wait to see what it does with a properly designed Cold Head.

Stay tuned for more upcoming tests and data.

Wednesday, December 21, 2011

CryoBUG getting ready for re-test

Late yesterday I let the compressor suck in 60 grams (approximately 2 oz)  of PVE-FVC32D oil, and put the unit back on a vacuum pump for final evacuation overnight. I did this to make up the oil that was lost during my previous HXC flushing operation, and concerns about what may have been left behind in the original AC unit this compressor came out of.

The PVE oil is rated for a lower temperature then the POE that was in the compressor from the factory. Since the two oils are compatible with each other, they will blend and yield a mixture with a lower freezing point then the POE alone. PVE is also a superior oil to POE, and completely miscible with all HFC refrigerants, as well as oils used in older R-22 systems.

Free Sample Bottle I received from IDEMITSU

In a discussion going on at XtremeSystems Phase Change Forum, it was suggested that perhaps my earlier diagram of CryoBUG might be a bit confusing. So I thought about it and came up with what I think is a more easily understood drawing of what CryoBUG really looks like (see below).

CryoBUG Piping Diagram w/Thermocouple Locations

I hope this helps people to better understand what it is that I have built. And due to some of the proprietary aspects of the design, I have not spec'ed out the sizes on the heat exchangers or the cap tubes. Although I'm sure this won't stop someone from figuring it out on their own. But it will slow them down a bit none the less.

You'll see that I show a high pressure cut-out for the compressor. This is good design practice, and something that I strongly urge be implemented. However in my current prototype I have not installed this safety feature at this time. Of course I am also watching the unit like a hawk whenever it's running, and wouldn't think to leave it unattended. Later when things start looking more like they are in their final stages with the charge properly tuned, I'll definitely be installing a high pressure switch.

In a few more hours I'll be charging the unit. This will be the new initial charge spec:

100 Grams R-134a
50 Grams R-23
50 Grams R-14

I'll add the Argon later once I establish what the unit does without it. Besides it doesn't take much to push you over the top, so I got to be careful.

Not sure if I'll be re-testing the unit today, but for sure tomorrow.

Tuesday, December 20, 2011

CryoBUG gets a cap tube adjustment

After my last two tests, I really wasn't happy about the temperature I was seeing at the phase separator. Normally when the amount of flow from CT #1, and the amount of R-134a is sufficient,  I should be getting down to at least -20°C (if not colder depending upon how much of the other refrigerants have dissolved into the sub cooled condensate). So what was I seeing? No matter how much refrigerant I had added, I could never get it to go below zero. In fact it tended to always be up in the positive numbers most of the time.

So what to do? Well I decided I needed more flow from CT #1. The only problem is that CT #1 is stuffed inside my Cascade HX, so it's not really replaceable at this point. I decided the next best thing was to add another cap tube in parallel, and insert this from the out side by drilling a hole into the HX, stick the new cap tube through the hole, and braze it in place.

The other problem was how to sub cool this additional cap tube like I had done for the original CT #1. So I needed to get the cap tube and the outer tubing of the HX in intimate contact with each other. Brazing it was the only reliable method I could think of to do this. So although this was going to be a tedious process, that is what I decided to do.

Just me starting to braze the new cap tube in place

Yep this is not easy

CryoBUG HX Stack with 2nd parallel CT #1 brazed to outer tubing

Close-Up View of brazed cap tube

It took me about 3 hours to complete, and although not very pretty, it should definitely do the trick.

Another problem cropped up while I was brazing the cap tube to the outside of the HX. Although I was under constant nitrogen purge, I began to notice a lot of smoke coming out of the loose end of my additional cap tube, and this was also accompanied by a constant drip of oil. This started to concern me that whatever oil still remained coating the inner walls of the HX, might be under going a bad chemical breakdown because of all the heat. So after I finished the brazing, and cooled the stack off with water, I decided I needed to flush out any oil that remained.

The flushing process consisted of evacuating an empty cylinder, sucking it full of Acetone, pressurizing this with nitrogen, and then forcing it through the cap tubes via the discharge line. I captured what came out the suction line into a bucket, which initially looked to be the color of oil (pale yellow). After repeating the flushing process 2 more times, until everything came out as clean as it originally went in. At this point I blasted everything out that I could with high pressure nitrogen and then put the HX stack on a vacuum pump for about an hour to evaporate anything that was left.

So I took the cleaned out stack and re-brazed it back to the condensing unit, and then put it on overnight pump-down. The next day It was sitting at 35 microns. All was once again well in my world.

Opps almost forgot! I also changed the final cap tube in order to reduce it's flow to about half of what it used to be. I figured since I was doubling the flow in CT #1, that I needed to reduce flow elsewhere if I was ever going to maintain the suction pressure where I desired it to be (looking to be around 10-15 psig).

Next on the agenda: Re-charge and test.

Sunday, December 18, 2011

CryoBUG 2nd Test Run

After thinking about the previous day's test run, I decided to not only try a new ratio for the R-134a and R-14, but while running, I also added in some R-23 and Argon. Since I was doing a little bit of adding gases on the fly (not always weighing it in), I'm guessing that the final charge consisted of:

130 Grams R-134a
50 Grams  R-14
40 Grams R-23
5-10 Grams Argon (the weight on this is probably pretty bogus -- Argon's too light for my scale)

As I tweaked back and forth with the various gases, the temperatures just kept getting better (see pressures and temps below).

Green = Evap-Out, Blue = Evap-In

Compressor Running Pressures (17 psig Suction and 235 psig Discharge)

Initially I had started out with only R-134a and R-14, with the R-14 at 40 Grams instead of the 100 Grams I had utilized in the first test. To make up the balance, and to see what would happen if I filled the nearly 100 degree gap between the two refrigerants. I added R-23. This reduced the gap to approximately 50 degrees between each refrigerant's boiling point. The end result was not quite as cold as with more R-14, but pretty close and with lower running pressures. Next began the Argon additions.

Every time I would add Argon the suction would drop a bit, followed by the discharge rising. And if the suction started to wander back and forth (hunting), adding a bit of R-23 and/or R-14 would usually set it straight. And with these additions the temperature kept creeping downwards.

The only thing that began to bother me was the fact that the phase separator temperature was no where near as cold as what I had anticipated. Only making it down to +2°C, instead of -30 to -35°C that had been expected. And even though I was using an extra 30 Grams of R-134a, nothing had changed in this regard. This could only mean one thing, and that was that CT #1 wasn't flowing enough. So I think to remedy this, I'll need to add another cap tube in parallel running along the outside of the cascade condenser's tubing, and drill a small hole in the side for insertion near the termination point of the original. In order to maintain the sub cooling aspect, I'll need to spot braze this additional cap tube to the outside of the cascade's tubing to maintain heat exchange.

Other then this small setback, the unit is working remarkably well, and doing so without over working the compressor (current draw = 2.6 Amps, Compressor Discharge Temp = +55°C).

Still got a ways to go if I'm to meet my goal of -150°C, but I think the journey to do so will be quite fascinating and fun.

Saturday, December 17, 2011

CryoBUG's first test run

Yep it's time to take this puppy from concept to reality. Luckily I didn't run into too much trouble assembling all the pieces, and in the next series of pictures you can see how it turned out (please keep in mind that this is essentially a bread board prototype).

Original AC Evaporator being used as Air-Cooled Condenser

Insulation of HX Stack Nearing Completion

Insulation Completed and Vapor Barrier Sealed

Side View showing Fred Flintstone style fan shroud

For the initial charge I decided to mimic what Fuderer had done with R-12 and R-14, only in my case I used R-134a in place of the R-12 (gotta keep it environmentally friendly). The proportions were 50/50 at approximately 100 grams per refrigerant. Static balance pressure = 205 psig. Start-up peak pressures = 14/307 psig (suction/discharge).

Air cooling of the condenser is being handled by two fans I had in my junk box, so I'm not exactly sure of the CFM, but I do believe they are close to 100 CFM each.

In about an hour the unit had achieved -100°C average temperature across the evaporator. After two hours the unit seemed to be fairly stable and was roughly about -109°C, with pressures of 25/223 psig.

Stabilized Evap-Out (green) and Evap-In (blue) after 2 hours (both readings are in celsius -- label incorrect)

The compressor appeared to be quite comfortable, with a discharge temperature of +57°C, and although a bit warm, I could rest my hand on top without too much screaming (actually no screaming). Current draw for 115 VAC operation = 2.5 amps (this includes the fans).

Based on some of the other inter-stage temperatures I was seeing, I do believe that I would do better with more R-134a, and less R-14. Perhaps a 60/40, or even 70/30 ratio might work out better. This will have to be investigated in additional tests before I start getting into adding any Argon to the charge. but no matter what, it really is quite impressive that such a simple system with only two refrigerants can produce these results. Fuderer really knew his stuff back in 1965.

Friday, December 16, 2011

CryoBUG gets a new Compressor

I was originally going to use a Rechi 39R131E R-22 based rotary compressor that I had removed from a used 5200 btu window AC unit (pictured in first CryoBUG post). But I have now decided to go out and purchase a new LG LW5011 5000 BTU R-410a based window AC as my parts source. My reason being that number one I was not sure of the condition of the used compressor, and number two that the R-410a compressor will already come with an HFC compatible lubricant oil. Being R-410a rated, also means that it will be guaranteed to work at high pressures. Which will be a big plus in my application, especially if I would like to eliminate the need for an expansion tank.

So I found a good deal on the LG LW5011 unit at my local Home Depot store for $100. Now all I need to do is fork over the cash and go pick it up.

Not sure yet on the model number for the compressor being used, but it is a rotary, and should only require 500 watts of energy to run it (probably even less in my application). I'll be posting specs on the compressor once I tear this window AC unit apart, and get a model number off of it.

For the air-cooled condenser in my system, I'll be reusing the original window AC's evaporator, fashioning a fan shroud out of plywood, and utilizing two 4" muffin fans for air flow. Keep in mind that this first version is strictly a proof of concept prototype build, and certainly nothing fancy. Assuming that I see good results, a more refined version will be designed and built later on.

I wasn't able to get the LG LW5011 because unfortunately I'm shopping out-of-season for a window air conditioner (go figure). So luckily there was one customer returned unit that they found way up on top of a shelf (and believe me when I say "way up", those pallet shelves are not something I'd want to be next to in an earthquake). So what did I get? A Zenith ZW5010 5000 BTU window AC, which looked very similar to the LG unit I had wanted.

It has a made in China rotary R-410a rated compressor with POE oil. Manufacturer: Qingan Refrigeration, Model: YZG-A048Y2D2, Rated BTU/Hr: 4810, Rated Watts: 480, RLA: 4.30 amps, LRA: 27 amps, Displacement: 4.8 cc3/rev, Capacitor: 35uf 250V

Here is a pretty good list of R-410a compressors they make for the US market:


Thursday, December 8, 2011

CryoBUG Theory

So what is this Fuderer patent I talked about in my last post, and how does it relate to the CryoBUG chiller I'm building now?

Well let's begin with what Fuderer had developed in 1965. In this first diagram contained in his patent #3203194 he lays out the most basic version of his unique refrigeration system.

Identification of Main Components:
20 -- Compressor
30 -- Condenser (water cooled)
40 -- Flash Chamber (Liquid/Vapor Phase Separator)
50 -- Evaporator/Condenser (Cascade Condenser HXC)
60 -- Lower Boiling Point Refrigerant Super Cooler (Sub Cooler #2)
70 -- Higher Boiling Point Refrigerant Super Cooler (Sub Cooler #1)
80 -- Evaporator
90 -- 1st Expansion Valve
100 -- 2nd Expansion Valve
110 -- Auxiliary Condenser HX
120 -- Cooling Water Supply/Return

Although the concept is sound, in it's portrayed form it still seemed a bit complicated. So I decided to see if I could simplify things and came up with this idea...

CryoBug HX Stack / Cold Head Diagram

This diagram has a reverse orientation from Fuderer's, showing the coldest part at the right (the Cold Head), and the warmer stages at the left. I'm just used to drawing things this way, and as the saying goes, old habits die hard.

As can be seen, the basics of Fuderer's design have been retained (Aux Condenser, Cascade Condenser, Sub Cooler #1, Sub Cooler #2). Although the expansion valves have been replaced with cap tubes (shown as dotted lines), and are directly sub cooled either within the cascade condenser (for CT #1) or within the Cold Head and it's interconnecting return line (for CT #2). I have also implemented additional sub cooling within the colder portion of the cascade condenser.

To elaborate...
What is shown as Sub Cooler #1 in the Fuderer system (item 70), is actually being performed by sub cooling CT #1 in the first half of my cascade condenser.  And what is shown as Sub Cooler #2  (item 60), is actually the 2nd half of the cascade condenser in my system. Further sub cooling is also provided by the returning gases after evaporation in the Cold Head, as they continue to travel past CT #2. This additional aspect is not covered in the Fuderer patent, but should still provide some added improvement in ultimate temperature.

The Phase Separator, which is based on a Temprite 340 coalescing oil separator, should provide superior liquid/vapor separation, and hopefully keep the high boiling portion of the refrigerant mixture from passing on to the cascade condenser.

In concept, the idea is to separate the high boiling component (probably Freon R-134a) via the phase separator and then use this to condense the low boilers, which in my case will be Freon R-14 with a little Argon dissolved into it. the better the phase separation, the more pure will be the R-14 being evaporated into the Cold Head, and thus a lower temperature will be the result. The dissolved Argon just makes it boil even colder, and the left over gaseous Argon creates a partial pressure effect that simulates an even lower evaporation pressure, and thus makes things even colder still.

Unlike Fuderer's example, CryoBUG in it's first iteration will not be water cooled, and instead will employ an air cooled condenser as shown in the diagram below. Cooling will be active, utilizing fans to move air through the condenser's fins.

CryoBUG Condensing Unit Diagram

The prototype will start out with no expansion tank, and probably no safeties (these will come later in the more refined design -- I'll just have to keep an eye on things in order to prevent possible damage to the compressor). And my charging manifold will suffice for the gauges. I'll also be doing the first test runs with a simple coiled tubing evaporator, no sub cooled CT #2, and no fancy machined Cold Head.

Next chapter, complete assembly...

Friday, December 2, 2011

CryoBUG starts a new life...

It's been about a year (maybe more) since I worked on this project. And I had almost resigned myself to leaving this path unfinished. But with Winter approaching and my pond project nearing completion, I found myself having a renewed interest in making this mini AutoC a reality. So with that in mind, here are some pics of my newest Heat Exchanger (HX) Stack design...

5200 Btu/Hr Rotary Compressor and CryoBUG HXC Stack
HXC Stack (Bottom View)

This new design is based upon the work of Andrija Fuderer which was patented in the later part of 1965 (Patent #3203194). His approach was extremely simple as compared to ultra low temperature autocascade refrigeration systems in use today. However over the years I've really begun to embrace the idea of "simpler is better". And unlike Fuderer's time, I can take full advantage of technology that did not exist in 1965. Thus making this simple design into something more reliable and that can realistically work without the inherent oil freeze-up problems that could have plagued his system.

HXC Stack (Top View)
So as can be seen when looking into the inside of the HX stack, a coalescing oil separator is being used as the refrigerant phase separator. It will also still serve it's original function, that being to separate oil from the refrigerant stream, and then return it back to the compressor. However unlike the normal position, that being inline with the hot discharge gases leaving the compressor, this one will be seeing much colder temperatures (-35°C).

The oil I'll be using has a pour point of -46°C, but due to the influence of the refrigerants that it'll be mixed with, the actual temperature should be lower.

The small coil on top is a dummy evaporator, being made from about 6 feet of 1/4" OD copper tubing. It's mass is probably fairly close to the cold head I eventually want to chill, so it'll be a good first test to see what happens. My target temperature is -150°C, and I plan to do this with an all HFC refrigerant charge plus a little Argon.

Stay tuned for more info as the project continues...