Thursday, June 19, 2014

CryoSPRITE Sooo Close to -100°C

After doing some more charge tuning and a cap tube adjustment CryoSPRITE is now verging on the negative century mark.

This required pushing the RIGID compressor's speed higher than I would like to see it sustained at indefinitely. But with that said, current draw was not unreasonable, and compressor discharge temperature was still comfortable, well within the manufacturer's spec.

These test results were done with the evaporator coil wrapped in about 1-2" of loose fiberglass insulation, stuck inside a plastic bag to act as a vapor barrier. Nothing very sophisticated, and judging by a few frost spots I was seeing on the outside, far from perfect.

Test Data (Click on Image to Enlarge)

Based on these preliminary tests, I would think it safe to say that a -85°C bio-storage freezer could be built utilizing this technology, while still allowing a good margin of extra temperature capability.

It wouldn't be a very big bio-freezer, but it would certainly be the smallest ever built for this temperature range, and definitely the smallest AutoC version in existence. Of course if someone would like to challenge this assertion, I would be thrilled to hear from you, and see what you have created and/or discovered. Now I just need to get to work and create one myself based on the CryoSPRITE technology used in this first prototype.

CryoSPRITE "Un-Wrapped"

Tuesday, June 10, 2014

CryoSPRITE Testing Resumed with New Compressor

I received my new compressor from Rigid Auto Parts about 2 weeks ago, installed it into the CryoSPRITE prototype, and got some test data over the weekend with a fully insulated evaporator coil (simulated Cold Head).

Test Data (click on image to enlarge)

For this initial test, I left out the Argon, and went with a mixture of 4 hydrocarbon refrigerants. The cap tubes are presently optimized for a -80°C system, and as can be seen, under very little load this was exceeded.

Rigid ACXZ14DCH Compressor and Controller
The compressor model number ACXZ14DCH has a displacement of a mere 1.5cc per revolution, but unlike most conventional larger rotary compressors, can spin up to 6,500 RPM. This is nearly twice the speed of the conventional, thus giving it a mass flow much better than it's seemingly small size suggests.

The reason for this increase in speed, can be attributed to it's DC Inverter motor and drive electronics. Unlike a conventional constant speed AC motor, this one can be varied in speed by changing the frequency of the 3 phase output of the Inverter, taking it from 1,800 up to 6,500 RPM.

What's also cool about the Rigid compressor's controller, is that it's output looks more like a sine wave than a square wave as normally used. This makes for quieter low vibration performance. I also like the fact that you are supplied with a multi-turn pot for speed adjustment, and a pre-wired plug for the compressor. Basically this thing is ready to run out of the box. The only thing missing in my opinion is a mount for the compressor, but they do supply you with a silicone gasket-like material (to wrap around the body of the compressor underneath a user supplied clamp) and a set of rubber feet. I think the idea is to buy an off-the-shelf capacitor style clamp for mounting purposes. The controller also automatically controls the ramp-up speed, which prevents you from accidentally slamming the compressor at high speed upon start-up (which is what happened to me with the Aspen compressor I was previously testing).

CryoSPRITE running without insulation
Just for grins, I removed the fiberglass insulation surrounding the test evaporator coil near the end of my test, just to see how it would handle a bit more heat load.

As can be seen in the photo, it still maintained -80°C while frosting up. And it was able to do so without excessive running pressures or drawing too much amperage. And while I used a 240 watt 24VDC power supply to run this thing, it never exceeded what a 200 watt supply would comfortably handle.

Even after several hours running in this uninsulated state, it still stayed fairly close to an average temperature of -80°C, as shown in the test results shown below.

Non-Insulated Test Data (click on image to enlarge)

What's next? When I get some spare time from my 'real' job, I'll be trying some different refrigerant charge blends, as well as cap tube changes to see what I can get this little guy to do. Most likely we are looking at a -100°system in its final configuration. To ask more of it, would probably yield very little heat load capacity, making it fairly useless.

Monday, June 9, 2014

Aspen 1.9cc Compressor Dissected

The Aspen compressors are miniature rolling piston rotary DC Inverter refrigeration compressors (see image below courtesy of Thermaltake for how the internal stuff works). This is also the same basic principle and implementation used in the RIGID Tiny compressor I purchased as a replacement for the Aspen that met an early grave at my hands.

I figured since I blew up my Aspen (p/n 19-24-1101), I may as well take it apart and see what makes it tick (click image to enlarge).
Top shell removed

Motor windings (stator) removed

Permanent magnet rotor removed

Inside view of rolling piston compressor

compressor top plate/discharge valve removed

Under side view of top plate

Eccentric crank removed

Rolling piston removed

Vane removed

Pretty cool design!

Stay tuned for exciting news about my tests with the RIGID compressor.

Monday, May 26, 2014

CryoBUG Licensing Suspended

By Michael St. Pierre
Mytek Controls
May 26, 2014

Santa Rosa, CA -- I Michael St. Pierre (owner Mytek Controls) have begun negotiations with an overseas manufacturing company, that for the time being wishes to remain incognito. We are presently investigating applications where the technology demonstrated in the CryoBUG demo unit may prove useful, and ultimately lead into a future product release. Pending the outcome of this new relationship, licensing of the CryoBUG technology has been put on indefinite hold.

I look forward to seeing this technology take its place in real world applications.


To learn more about this, please use the 'contact me' button below.

Tuesday, April 15, 2014

CryoSPRITE HX Stack Test

Click on Image to Enlarge
A few weeks ago I had a very expensive mistake occur, and that was the loss of a brand new miniature Aspen compressor.

I rushed into the first test with far too much optimism, and not enough caution, thinking that this was going to be a walk in the park.

However the problem was that I didn't know as much about this new smaller HX Stack configuration, as I thought I did. So that combined with starting with too high of an Argon charge, caused such a huge up surge in discharge pressure, that the compressor motor current draw opened up one of the motor windings before the controller could safely shut it down.

End result: Dead Compressor.

So having no compressor, and not quite ready to shell out another $450, I needed a new strategy. My solution was to use my Appion G5 Twin recovery pump as a temporary compressor. This way I could better figure out what was going on without killing yet another mini-compressor.

If you look at the picture in the beginning of this post you will see the results of an actual test using the recovery pump as a compressor, and getting some very decent performance (the blue TC Meter is connected to the evaporator coil mid point, which is surrounded by pink fiberglass insulation). In order to get to this point, I had to first make some radical adjustments in cap tube flow. Basically I increased flow.

When I was designing the CryoSPRITE HX Stack, I forgot to take into account just how small the discharge flow path was getting to be. This started to have the effect of the HX tubing acting like a cap tube. So combine this with the actual flow regulating cap tubes, which I had down sized as well, and the end result was a huge restriction in flow. So much so, that even with a moderately charged system, the discharge peak was enormous. And with the recovery pump I could see this before anything broke. After several adjustments, I was able to get everything in proper proportion, and then good results soon followed.

Best results to date using the recovery pump/compressor...

Suction Pressure: 32 psi
Discharge Pressure: 250 psi
Evaporator Temperature: -131°C

Charge: R600a, R1150, R50, Argon

Aspen Customer Service not so good after the sell.
Before I actually placed an order for the Aspen compressor, the President/Owner was very helpful at answering any of my questions. But later the first thing that slipped was the delivery date. This got pushed out by more than a month beyond the month I had already been waiting. So I compromised and suggested that they send me the older design instead, which they did (including a small refund for the price difference). However when my compressor blew up, and I contacted Aspen explaining that something had gone wrong (even admitted that it was most likely my fault), help at figuring out what was truly wrong never came. So I had to diagnose it myself. However I will give them the benefit of the doubt, and assume they just didn't know how to respond due to the unique nature of what I am using the compressor for (definitely not an R134a system). But it would have been nice to get some troubleshooting tips.

ACXZ14DCH 1.5cc/rev Sine Wave DC Inverter Compressor
Being kinda spooked by what happened, I have decided to use an alternative compressor from an alternative source. That source being Rigid Auto via Ali-Express (Store No: 331602).

The compressor model number is: ACXZ14DCH, and it is a 24 VDC Inverter Drive Rotary Compressor. Being a sine wave based system, it promises to be quieter and less vibration prone then its predecessor (ACXZ14DC-S). It also promises to be cheaper then the Aspen, coming in under $380 including FedEx shipping.

The people over at Rigid Auto are very nice to communicate with, and kinda go the extra mile at trying to answer any questions you may have, and/or provide data sheets on what they sell. My plan is to order a sample in about 1-2 weeks for use in my CryoSPRITE prototype unit.

Wednesday, February 26, 2014

CryoBUG Life Testing Stopped

Well as predicted, on the second round of life testing the gradual accumulation of oil getting up to the final cap tube, caused enough blockage to reduce the flow of refrigerant feeding the Cold Head. The end result: gradual reduction in temperature under moderate load conditions over the first 2-3 weeks of 24/7 operation, followed by a rapid warm-up of at least 20°C when flow could no longer keep up with load conditions.

Vapor-Liquid Phase Separator
The oil blockage occurs when it gets into an area having cold enough temperatures to freeze it. The usual method of trying to avoid this is to use a coalescing oil separator, something that would be difficult to do in CryoBUG, and even more so in the upcoming CryoSPRITE due to size constraints. So what I'll be doing instead, is to try to improve my refrigerant phase separation design, of which I only have one in the system.

The image to the right shows a fairly sophisticated vapor-liquid phase separator (courtesy of Wikipedia). I will be trying to implement some of the techniques shown by this example into a small enough package to work in the CryoSPRITE, and eventually into CryoBUG. Since I have very little space available, I will need to be creative with my approach, and will likely try to combine some of what is independently shown in the diagram into a commonly shared aspect.

It is a pity that I have to do this even with an all hydrocarbon charge, something that I thought would insure good oil return back to the compressor before it had a chance to freeze. But this is the way things often are when doing product development. It is also most likely a result of my one phase separator autocascade design, which is really pushing the envelope at -150°C.

Most autocascades designed for ultra-low temperature operation use up to 3 refrigerant phase separators, as well as 3 cascade HX's. Having the extra phase separation normally gives you ample opportunity to separate out the oil along with the condensed refrigerants, well before it gets into a cold enough stage to freeze.

But I have no intention of giving up so easily on my simple one phase separator design. So we'll see how things go when I get into testing CryoSPRITE, which should be fairly soon, since I now have all the parts required for assembly.

Sunday, February 23, 2014

CryoBUG requires regeneration after 21 days

Well it appears that a very small amount of oil gets past CryoBUG's phase separator, and over time accumulates in the final cap tube feeding the Cold Head, ultimately freezing and blocking passage of refrigerant. The effect is gradual, taking nearly 3 weeks to cause a cooling problem.

Presently I have been running a 2nd life test going onto 18 days now, with an adjusted refrigerant charge. Thus far things are okay, but signs have already begun to appear that are similar to what was seen in the first test run. These signs signal a possible problem to come. I'll know for sure in about another week of 24/7 operation.

The problem isn't catastrophic, and the unit will recover by allowing it to warm up overnight and then restarting. However it would be better if a solution can be developed. And just such a solution is being implemented in the new CryoSPRITE prototype, which if successful can be incorporated into CryoBUG.

So life testing will continue.

Stay tuned for more news to come.

Thursday, January 23, 2014

CryoSPRITE is about to be born

Because of the success I've had with the CryoBUG platform, and seeing a need for something even smaller, I've begun work on a new development unit code named CryoSPRITE. Currently the rough dimensions for this unit look to be a 7" x 7" footprint x 12" tall. It is hoped that it'll be able to do a solid -140°C at 20 watts of heat load, and be light enough to easily carry around by one person. The application is the same as CryoBUG, doing water vapor cryopumping for small vacuum systems, or imaging detector cooling. But the idea behind this project is to make it a better fit for very tight spaces, which will help in many existing applications.

Aspen Original Series Compressor
(Shown with its DC Controller)
The first order of business was to find an extremely small compressor. And that was done thanks to Aspen Compressor, utilizing one of their Q-Series Rotaries. The new Q-Series takes the place of their original compressor line (shown in the image on the left). And because this compressor series is so new, they didn't have one available for shipment until the end of this month. So now here I am anxiously awaiting its arrival, like a kid waiting for Christmas to finally get here. Link to: Q-Series Compressor Dimensional Drawing
Photo Courtesy "Honey I Shrunk The Kids" © Disney Studios

The next thing on my list, was a very tiny air cooled condenser to go with the small compressor I had ordered. This was not something I could readily find for sale off the shelf, or at least not in a quantity one buy. So I needed to take something larger and shrink it down to size.

Now unfortunately I don't have access to a Shrink Ray Gun, other wise this would have been the perfect solution, and a done deal. In fact this would have solved everything by simply placing my existing CryoBUG unit under the ray, and reducing it down to just under 1/2 its original size. Hmmm... I wonder if I could rent one of these?

Custom Air Cooled Condenser with Fan
Anyway I had to do it the hard way and chop up a bigger air cooled condenser (actually it was a left over window AC evaporator from one of my earlier projects).  Although it was a bit of a challenge, I was able to fabricate a fairly decent miniature condenser, that was just slightly larger then the 120 mm muffin fan that I would use to cool it.

With the fan and the compressor both running off of 24 VDC, and combining this with a switching power supply having a wide voltage range, a more universal power input should now be possible (90-250 VAC 50/60 Hz).

CryoSPRITE Prototype HX Stack
Continuing to go down the list of parts needed for my miniature chiller, the next item was a autocascade Heat Exchanger (HX) stack. And as you probably guessed, this also needed to be much smaller then anything I've ever built before (still no Shrink Ray Gun in sight). So I had to design and build something from scratch that was of an appropriate size to match up with the Aspen compressor's throughput.

This actually was easier for me to do then modifying the air cooled condenser. Which had involved some very tricky bending of the copper tubing return loops.

What's left?

I'll be fabricating a small combination evaporator/heater assembly to run heat load tests. And of course I'll need to put all these pieces together, leak check, charge refrigerants, and then power it all up for the first time.

Stay tuned for more to come.

Wednesday, January 22, 2014

CryoBUG Life Testing 14 days and counting

Its been 2 weeks now of continuous operation for CryoBUG, and everything appears to be optimum, with a Cold Head temperature today of -154°C.

Having run for this long without a problem, I don't foresee any failure from this point forward, But to be on the safe side, I plan on continuing the life test for at least 2 more weeks just to see how things pan out.

Now that I know that this design and refrigerant charge is looking good, I have already started project planning on an even smaller version of CryoBUG. The new project code name will be CryoSPRITE, and it'll be about 1/3 the size and just under 1/2 the capacity of its bigger brother. My ultimate goal being to develop a -140°C or colder unit that can fit into a very tight space, and be light enough to hang off of the vacuum flange that connects its Cold Head to the coating or imaging detector chamber.

Aspen Q-Series Compressor
To make this drastic size reduction possible, I will be using a miniature 1.9 cc/rev rotary compressor made by a company called Aspen Compressor based in the United States.

It is a DC variable speed design, that packs an amazing amount of punch for its size. And compared to the compressor in the CryoBUG, comes in at 40% as much displacement, but with an overall size and weight reduction of nearly 70%.

Having the variable speed aspect should help with the start-up, quick cool-down, and energy savings under low load situations. And being a low voltage DC input, will allow for a universal switching power supply to feed it, thereby being able to run on either 120 VAC or 240 VAC 50/60Hz power systems without any reconfiguration necessary.

I look forward to working on this project, and expect to see some of the first prototype tests being run by the end of February 2014.

Friday, January 17, 2014

CryoBUG vs Polycold's PCC Chiller

Polycold® also makes a compact chiller which they call the PCC™ (newer version of Polycold's CryoTiger™). The major differences between what they make and a CryoBUG, is that their unit uses room temperature flex lines to interconnect the Cold Head with the Compressor Unit. And they also offer a charge blend that can do liquid Argon temperatures (-186°C) at 6 watts of heat load.

Peeking under the hood, the differences become even greater. CryoBUG uses what is referred to as autocascade refrigeration to achieve its very cold temperatures, whereas the PCC is primarily based on the Joules-Thompson Effect with no inter-stage phase separation. Utilizing only the JT method tends to be rather tough on the compressor, causing it to have high differential pressures, resulting in poor cooling which can considerably shorten its life. However with an autocascade system like CryoBUG, compression ratio is low and compressor cooling is excellent, thereby promoting a longer lifespan. Also an autocascade system is generally more efficient, requiring less horsepower for a given temperature and heat load.

And if you look at the following chart you will see that CryoBUG is actually more power efficient while also taking on more load (CryoBUG has a smaller compressor then the PCC).
CryoBUG vs Polycold PCC Comparision Chart

Presently Polycold® offers 4 different gas blends for the PCC, and two different Cold Heads (Standard and High-Performance). I picked the two gas blends that came closest to CryoBUG's -150 blend for my comparison (other CryoBUG refrigerant blends are possible that would yield even higher capacity results, albeit at a warmer temperature).

The PT-16 although a great performer, tops out at 24 watts according to Polycold's literature. The PT-30 looks to be the closer match for the CryoBUG, but it is limited to 32 watts, and quickly warms up above 20 watts of heat load.

CryoBUG Demo Unit
Another noticeable difference between CryoBUG and the PCC, is that CryoBUG is equipped with a very nice tech package, including a temperature display, dual compressor pressure gauges, and a remote control/monitoring capability. Also the safety system is much more robust, having sensors for low and high pressure, and high liquid line conditions (the liquid line is the refrigerant line exiting the air cooled condenser, if it gets too hot then it's probably time to clean the condenser). All of the faults will produce a visually mapped front panel display that tells you specifically which one has occurred.

This stuff isn't just a bunch of unnecessary bells and whistles. It all serves to let the user know how well the unit is performing its job, and gives them a heads up if something is beginning to go wrong, thus allowing them time to schedule a service. And probably one of the most important additions, is the inclusion of the built-in temperature display. Kinda nice to know what temperature your chiller is producing, after all it is a refrigeration system.

The PCC unit does have an hour meter, which CryoBUG does not, but this was primarily done because the PCC has an oil absorber cartridge that needs changing after so many hours of operation (CryoBUG doesn't need this, since it has excellent oil management not reliant upon parts that need to be changed).

I think the biggest advantage that the PCC has over CryoBUG, is the flexibility of being able to place the Cold Head independently from the Compressor Unit. This could come in handy for many situations. And although it is certainly possible to provide a similar feature on the CryoBUG, it would never be quite as good as the PCC in this regard. This is  primarily due to the colder line temperature, and not being able to offer self-sealing couplings as is the case with the PCC. Also there would likely be considerable performance degradation if very long lines were utilized. However with all that said, CryoBUG's line-less, coupling-less implementation does insure trouble free operation.

I think that pretty much sums it up.

Click here for more information on CryoBUG

Wednesday, January 15, 2014

CryoBUG Life Testing 7 days and counting

CryoBug has been in continuous operation for just over 7 days now, and seems to be doing fine. Although there has been an ever so slight decrease in suction pressure, but the discharge remains unchanged. This morning it was a bit colder then it has been at 62°F. With this lower ambient condition, the Cold Head temperature dipped down just a few tenths of a degree away from -155°C.

Basically what the life testing will determine, is if I need to specify a certain Regen time period.
Regen =  regeneration cycle, a shut-down and thawing out of the unit, mainly for purposes of releasing accumulated trapped water vapor on the Cold Head, but can also restore system balance. So for CryoBUG this could free up any frozen oil that migrated up into the heat exchangers over time, and thus allow it to return back to the compressor during the first part of the next run before everything gets super cold again.
Hey even the Borg need to regenerate from time to time!

So I'm pretty excited and happy about how things are going with my little BUG. But not one to sit back and just be satisfied, I am already making plans for the Next Generation (pun intended).

By the way, I added a widget to my side bar that is keeping track of how long CryoBUG has been in Life Testing mode. If I was more adept at programming and setting up web servers, it would have been so cool to have a real-time display of the Cold Head temperature as well. But unfortunately my brain couldn't take the strain of having to figure that out, so you'll have to use your imagination instead.

That's it for this post :-)

Friday, January 10, 2014

CryoBUG Life Testing in Process

CryoBUG Life Test
(Click on Image to Enlarge)
About 2 days ago I began 24/7 life testing of CryoBUG with what looks to be its final charge, based on Hydrocarbons and Argon.

For the test, I have it set up in a small insulated room in the back of my shop, which sees a room temperature variance of as low as 64°F at night, to a high of 74°F during the day. The room is only being heated by the energy feeding the CryoBUG demo unit (about 350 watts).

Over that 10°F (5.5°C) testing room ambient temperature range, the Cold Head has only varied by 2°C, ranging from -151 to -153°C.

I also don't see any unusual changes in compressor running pressures, which if there were, would likely be indicative of something beginning to freeze up. So far so good.

Using yesterday mornings readings, I entered the data on a piping diagram pictured below.

These readings correspond to a static heat load of roughly 10-15 watts. This is due to the Cold Head being poorly insulated by a barely adequate vacuum in a 2" vacuum vessel, which is not actively being pumped on. There is also no radiant shielding in use.

My plans are to run this unit non-stop for 4 weeks, or until it shows some signs of trouble. I'll try to report in weekly as the life test progresses.

Wish me luck :-)

BTW; here is the typical energy cost to run this thing 24/7 broken down into days, months and years.

Monday, January 6, 2014

CryoBUG: New Controller Board Assembled

I got the new controller PCB's late last week, and assembled one over the weekend. Best part was that when I bench tested it, everything appeared to function perfectly (what a relief).

CryoBUG Controller with optional OLED Display connected

Today I installed this into CryoBUG and took it for a test run, and was pleased to see that nothing out of the ordinary happened when controlling the actual compressor. One of my concerns was that the switching currents might upset the microcontroller, causing it to randomly reset. So for that reason I made sure to place a couple of power decoupling capacitors real close by.

The only unexpected thing that I noticed when probing the board with a volt meter, was that the 12 VDC supply drops down to about 8 volts when the compressor relay is powered up. Doesn't seem to be affecting anything though, and since the voltage is about 14+ volts before kicking on the relay, it shouldn't pose a problem (the datasheet shows a must operate at 8 volts). Probably wouldn't be as bad if I had implemented full wave rectification, and I might look into this later just for grins (this might mean a possible Rev 2.2).

Since the 5 VDC power draw is very small, I could most likely eliminate the heat sink on the regulator.

I really do like the OLED display. It looks very bright and sharp. But unfortunately no place to put it on this build.

Update 1/8/2014...

I assembled a second board and decided to post a couple of close up shots.

Top View
(click on Image to Enlarge)

Bottom View
(Click on Image to Enlarge)

And as I suspected earlier, the heat sink is not needed, so it was left off of this build. As can be seen, this is a pretty simple and small controller board. And yet for all its simplicity, it has a multitude of possibilities. All that's required is some creative programming of the on-board microcontroller.