Sunday, December 30, 2012

Refrigerant Blending/Charging Cart

On my continuing quest to get repeatable results when charging the mixed refrigerants into my CryoBUG unit, I decided to put together a refrigerant charging cart (see below).

Blending/Charging Cart (click to enlarge)
This essentially combines 3 elements...
  1. Liquid Refrigerant Charging Glass
  2. High Precision Mirror Gauge
  3. Refrigerant Mixing Manifold
The charging glass is a modified version of a standard R-22 Charge-Check made by Thermal Engineering (see blog post: Creating an R-600a Charging Glass). This allows for precise measurement of the liquid part of the charge (in my case that would be R-600a).

The next part of the apparatus  required a precision method of measuring in the refrigerant gasses (R-23 or R-170, R-14, Argon). This was accomplished with a high accuracy pressure gauge (Marshal Town 0-400 psi 4.5" Mirror Gauge).

And to bring it all together, a refrigerant mixing manifold needed to be fabricated, which allows the individual refrigerant components to be metered into the CryoBUG unit without having to disconnect and reconnect charging hoses. This greatly simplifies the charging process, and minimizes any errors and/or contaminates from getting into the final mixture.

Charging Cart Diagram

Mixing Manifold and Pressure Gauge Assembly (click to enlarge)
I'm very happy with how the refrigerant mixing manifold turned out. Luckily for me I have access to a Bridgeport Mill, which made things so much easier when drilling out the brass block.

Afterwards, I installed five brass 1/8" NPT x 1/4" flare fittings to complete the assembly, and prepare it for refrigerant hose attachment.

For the base of the charging cart, I used one of those moving dollies that I picked up at Harbor Freight for about $17 (sale price). I then added a 3/4" piece of MDF to create a solid top, and a 4x4 post to allow for mounting the charging glass, pressure gauge, and mixing manifold.

Mixing Manifold and Gauge prior to Installation (click to enlarge)

After everything was mounted. I threw in a couple of short refrigerant hoses to connect the refrigerant gas cylinders and charging glass. Total cost on this project not including the gas cylinders was around $200.

Now I really think I am ready to resume my CryoBUG testing (yes, I know I keep saying this, but this time I mean it).

Thursday, December 13, 2012

Creating an R-600a Charging Glass

Charge-Check Model #7003
I got my Ebay purchase a couple of weeks ago, thinking that I was getting a Charge-Check 7001 (16 ounce) charging glass, but instead I ended up with the 7003 model which holds 2-1/2 lbs (40 ounces). Although for $25 I couldn't complain too much.

However this was quite a bit more volume than I needed, and my concern was that the resolution was going to be very poor as a result of this. With my CryoBUG unit requiring less than 50 grams of R-600a, I knew that it would be very difficult to measure this out accurately. It seemed like my best option would be to displace some of that extra volume by the use of an internal displacement device. So the first order of business was to disassemble the Charge-Check.

7003GS Gasket Set
Very quickly after pulling things apart, I realized that the seals were not reusable. After Googling around for a bit, I came across a gasket set being sold by a company called FieryChill for $33. After adding in shipping for both items, I was now up to $78. Still not bad considering that a brand new Charge-Check would have cost me around $178-$200.

So while I waited for the gasket set to arrive, I went about creating a suitable displacement vessel to mount inside the original Charge-Check cylinder. I had some 1-5/8" caps and tubing left over from an earlier autocascade project (used for making refrigerant liquid/vapor phase separators). And this seemed to give me the ratio of open v.s. closed area that I was looking for.

I cut the 1-5/8" tubing to fit inside with a small amount of vertical space to spare. Next I silver soldered a small threaded stud to the bottom of one of the caps, after which I brazed the caps to the tubing. The stud would be used to secure this displacement vessel to the bottom end cap of the Charge-Check, after a hole was drilled and tapped to fit the stud.

I then bead blasted the original cylinder to remove the R-22 measurement sticker, gave it a coat of paint, and reassembled the Charge-Check with the newly made displacement vessel inside.

After insuring that everything was leak tight, I filled the Charge-Check with two 5.5 ounce cans of the Wall Lenk iso-butane fuel, having previously measured the empty weight of the Charge-Check with my electronic scale.

Subtracting the filled weight from the empty weight, yielded the actual net weight of the iso-butane now residing in the Charge-Check. This ended up being fairly close to what was suppose to be contained in the Wall Lenk butane fuel canisters that I used.  Final weight was 10.5 ounces. With a liquid column height of 13 inches this converted to 1 ounce for every 1.25 inches of height. Using DeltaCad, I created a new measuring scale based on this information.

Here is a look at the finished R-600a Charge-Check produced by all the modifications (click on the image to enlarge). I got the measuring scale printed on card stock and laminated at my local Kinko's. This ended up adding another $5, bringing the total price of my customized Charge-Check to $83.

As can be seen, I also included a kilogram scale as well, having increment marks spaced at 10 grams each.

Now I just need to get back to my original project, that being charging, testing, and tuning of my latest CryoBUG Demo unit.

Tuesday, October 30, 2012

CryoBUG new source of R600a

In my last post I mentioned that my Ebay source for the R600a (100% pure Iso-Butane) had unfortunately succumbed to a several month back order. They were gracious enough to quickly refund my purchase price, and also apologized for the lack of inventory. However this temporarily left me with  no way to recharge my CryoBUG demo unit. And on another front, I was still trying to figure out a much better way to charge this into my unit, with repeatable results.

So I started Googling and came across a site called IQ Dental Supply which appeared to have the Wall Lenk 65F Iso-Butane fuel I had been using as my R600a alternative (the MSDS verified it as 100% pure Iso-Butane). And they also sold the burner accessory as well, which I had previously used to fashion a 1/4" refrigerant flare fitting adapter from (no need to make another, but it was good to know that they carried this item as well). The only problem was that, I didn't know if they would sell this to a non-dental person like myself, having run into problems like this before.

Well I decided to place an order for a quantity of six WLC-65F (5.5 oz Wall Lenk Butane Refill), so as to get their quantity discount pricing. Within a few seconds the order was processed, and I received a confirmation email (so far so good). And then about 2-1/2 hours later I received yet another email verifying that the order had been fulfilled and shipped (Woohoo!!!). And yesterday my case of butane refills arrived. The only glitch in ordering from IQ Dental was that the email confirmations did not reflect the additional S&H and HazMat charges that were added to the order, the later of which adds a pretty hefty toll ($25.00 HazMat fee). So in the future I would be wise to double my order in order to help offset this extra cost. On the plus side, their product pricing was the lowest I've seen thus far.

Now that I have my R600a once again, I'm almost ready to get back to my demo unit testing. The only other problem left to resolve is the repeatability issue when charging by weight. As I had mentioned in my last post. Due to the very small quantity of refrigerant to be used, and the near impossibility of negating the influence of the interlinking refrigerant hose connection, the variance in quantity charged vs. what was desired is unacceptable. I needed a better more accurate way of doing this. And that way would be to use what is called a Charging Glass, which is a pressure cylinder having a graduated column. Problem is these can be quite expensive, and are normally of a much higher volume than what I need (hence not as accurate for small refrigerant quantities).

Charge-Check 7001
Enter Thermal Engineering's Charge-Check 7001, which can hold up to 16 ounces of refrigerant. Although at $150+ it was still a little bit more then what I wanted to spend. And by the nature of the beast, it is not something that becomes inaccurate with age, so buying one on the used market seemed like a good option.

Checking Ebay revealed several models made by Thermal, but only a couple were the smallest model 7001 that I was seeking. Out of these, most were being sold for near retail pricing, which always makes me ask "why would I buy this used for the same price as new including a warranty?". Anyway the good news is, that there was one going for $25.00, and I grabbed it.

The Charge-Check I'm getting is most likely graduated for R12 and/or R22 since it appears to be an older model. For my purpose this really doesn't matter, since I doubt that anyone makes one for R600a. So what I'll need to do is fill it with a known weight of R600a (10 ounces), using my electronic scale to verify the before and after weight of the charging glass. Then I'll create my own graduated scale by dividing the liquid column height by 10, and further subdividing this down to show 1/2 ounces. Or if I'm lucky, I'll find that it already matches up with the existing graduated scale (however I'm rarely ever that lucky).

The great thing about using the charging glass method of measuring in my liquid refrigerant, is that it is entirely immune to any outside influence other than a slight variance due to ambient temperature changes (less than 2% error).

All my gaseous refrigerants will be charged in after the liquid using additive pressure. And by weighing the CryoBUG before and after charging is completed, I'll get the total charge weight of all the blended refrigerants. This should work well, and give me repeatable results.

Saturday, October 20, 2012

CryoBUG in Standby

It's been a little while since my last post, so I thought it would be good to bring people up to date on what has been happening.
I charged the new demo unit with what I thought was the same charge as used in the breadboard prototype. But after doing a test run, I realized that there was something very wrong. Apparently my method of weighing in the various refrigerants was just not producing repeatable results. A lot of this can be blamed on the influence of the hose attachment between the refrigerant cylinders and the unit. Try as I might, it is virtually impossible to keep the weight and drag of the hose from offsetting the scale readings.
So I'm left with adding all the high boiling gases based on pressure. And for the R600a, which is charged in as a liquid, I will need to use a self sealing fitting in order to disconnect and weigh several times during the charging process for verification.
I also ran out of R600a, so I am patiently waiting for some more to arrive. Had to find another source, since my eBay connection ran out, and had a 2 month back order. As it turned out my new source is even cheaper.
My preliminary tests suggest that my ultimate temperature will not be as cold as achieved with the breadboard. Although I am still seeing fast cool downs into the -100 C range (18-19 minutes from warm start). However with the extra mass of the cold head, and some conduction into the vacuum flange, temperatures will most likely be in the mid -140's.
If all goes well, I should be testing again next week.

Monday, September 24, 2012

CryoBUG Demo Unit Assembly Completed

I've been pretty busy over the last couple of weeks taking all the individual pieces that make up a CryoBUG, and bringing it all together into a complete package. A lot of planning had to be done in process, in order to adapt the conceptualized drawings into reality. But I think in the end it all went very well, and stayed fairly close to what I had originally envisioned.

The HX Stack stayed pretty true to the original design as used in the final prototype shown here. So at least this didn't require a lot of rethinking, but it did get housed in an entirely different way (see overhead view below).

CryoBUG (Overhead View) Almost ready to be foam insulated
A full complement of thermocouples were also installed, so as to get a good comparison to the prototype test data. This may very well reveal differences brought about by the new Cold Head being used in this demo unit. And I will also be doing a full capacity curve by applying heat directly to the face of the Cold Head via an attached power resistor. So it will be nice to see how all the different stages handle the extra heat.

Since I thought many of you would be interested in seeing the expanding foam insulation being applied to the HX Stack, I have also included a video showing the entire event. Due to me being the cameraman and having to deal with the actual foaming process, sometimes the framing is less then desirable, often times cutting off my head.

Note: In this video I mention 8 minutes elapsing, in reality it was 8 seconds

After the foam insulation was completed, and all the excess was trimmed away, the next step required installation of the air cooled condenser, which resides in the upper most section of the enclosure.

HX Stack foamed, now ready for installation of condenser
Unfortunately I had a slight mishap when I drilled through the condenser's fins to allow the lift handle support post to pass through. I nicked the copper tubing causing a small leak. It didn't look bad at first, but after applying some heat and brazing rod, it kinda got a bit ugly. Luckily it all gets covered up by the top grill and handle, so it shouldn't be too noticeable.

The main problem that caused this, is that the threaded support post really needs to be reduced in diameter. At it's present size, it's just too close for comfort to fit between the two copper tubes passing on either side of it. Something I'll have to remember if I ever build another one of these again.

-- With the air cooled condenser in place and plumbed, the main assembly is done --

Next I secured the top panel, put the unit on a vaccum pump, and when it reached a pressure of 78 microns, I put in a very simple preliminary test charge made up from some reclaimed HCFC refrigerants, including Ethane (no R-14 or Argon). Basically I just wanted to confirm that everything worked. And for this go around, I wasn't shooting for anywhere near -150°C, which is the ultimate target temperature for CryoBUG.

It was nice to see the white snowy stuff appear soon after start-up. Also it looked like the liquid line bonding inside the 2" SST tubing was doing a good job of keeping water from condensing on the outside, which was the intended purpose.

Red = Cold Head Temperature, Green = Set Point Temperature for Relay Activation
With the initial test charge, the unit got down to -77°C pretty quickly, but soon after, it leveled off around -80°C. Which is about as good as it gets without any R-14 or Argon. The main thing, is that I was able to verify that all systems are a go and ready to be recharged with the real hydrocarbon juice.

The custom controller board also appeared to be working correctly, and even detected a low pressure fault soon after start-up when my initial refrigerant charge was a bit weak. This resulted in a system shutdown, followed by a red LED lighting up in the compressor symbol and the On-Off switch's LED illumination flashing red (see video below).

Here is a view of the controller board mounted behind the front panel. The long green terminal block is where all the safeties are plugged in (LPCO, HPCO, HTCO).

Controller Board glowing blue, a sign that it is active
And in case you noticed, I still need to clamp the bottom of that discharge vibration eliminator.

On the news front, my brother has been working on a new version of the test software that I use for logging multiple temperature points from thermocouples. I can hardly wait to set everything up and see how it all goes with a "real" refrigerant charge in use. It should be exciting.

Saturday, September 8, 2012

CryoBUG Enclosure Arrived

Front Panel Express came through with flying colors. Everything looked great and was packaged very well for shipment. They offer a very complete enclosure service, including all the hardware (screws, nuts, brackets) required to complete the assembly. And by using their free software (Windows, Mac, Linux) it is relatively easy to create whatever you want.

And here it is assembled with one side panel removed (this was just a test assembly to make sure everything lined up properly). By the way that really slick looking handle on top, I got from Halder. It's made by a German company called Rohde, and boy do they know how to make a robust, yet beautiful flip-up handle.

I'll have more pictures later when I start doing the final assembly with all the refrigeration components.

In the meantime I was able to get the Cold Head welded to the mounting base, and polyurethane foam insulate the interior space. A special thanks goes out to Randy at Randy's Design & Machine for cleaning up the messy cut I made when I modified the original piece. And also to Mike from M&T Systems for doing such an excellent job on welding it all together.

I would also like to acknowledge the donation of the original unmodified Cold Head from back in 2009, although they will remain nameless since I'm not sure if they wish to be known.

It's good to have gotten to this point, because now I feel like a kid with a really cool Erector Set. You know the deluxe one with all the motors and gears. So basically I just need to bolt all the pieces together, wire it up, charge it, and press GO!

Wednesday, September 5, 2012

CryoBUG Demo Unit Design in Process

Well sometime later today I hope to receive my enclosure from Front Panel Express in Washington. I am having mixed feelings of excitement and anxiety in anticipation of seeing what I have been working on come into reality. I just hope that I didn't make any major blunders in my enclosure layout.

Here's a look at some of my virtual piping design work...
Top View of basic piping (click on image to enlarge)

In the meantime there has been good progress made on fabrication of the Cold Head assembly that will be used on this initial pre-production CryoBUG demonstration unit.

Below can be seen the connection side of the actual "Cold Head" itself. The final (evaporator) cap tube is tightly coiled around the suction return line, and although it can not be seen in this image, enters on the left side, and then continues down into the copper head below. the stainless steel tubing is the thermowell used for measuring the temperature of the copper head, through use of a thermocouple probe.

To mount the Cold Head to the CryoBug, I am using a conflat half nipple which will be welded on to the Cold Head back side pictured above. Because of concerns about possible thermal conduction from the cold feed and return connections on the inside to the outer 2" stainless steel tubing of the half nipple, a copper tube was bonded to the inside which will eventually be connected to the air cooled condenser liquid line. It is hoped that this will maintain close to ambient temperature on the 2" tubing, extending back into the conflat mounting flange, and thereby avoid condensation or ice up problems on these parts.

The Inside of space will be insulated with an expanding polyurethane foam, same as to be used for insulating the HX Stack.

Here is a shot of the complete Cold Head assembly prior to be welded together...

For future versions, I have some redesign ideas that may alleviate the thermal conduction problems, as well as simplify the overall design. This particular Cold Head that I am using was designed for a specific application by a 3rd party, and has been greatly modified to suit my purposes. The price was right (as in "free"), so I really can't complain too much.

In my next post I will be showing images of my enclosure panels when they arrive, and an initial pre-fitting to render the complete enclosure.

Thursday, July 26, 2012

CryoBUG Accelerated Cool-Down (Turbo Charged)

Well it isn't actually Turbo Charged, but what I did is revert back to a higher flow in the final cap tube, and got a pretty phenomenal speedup in the cooling rate. Of course there was a small compromise, and that was a slight sacrifice in ultimate temperature, as well as a bit more work for the compressor (still quite comfortable for a R-410a compressor).

In my last test I was able to achieve a stable -156°C, but it took nearly 2-1/2 hours to get there (not exactly a speed demon). Anyway after going back over my notes on previous tests, I saw a rapid cool down in a test I blogged about on April 4th 2012. In this test I achieved nearly the same temperature, but in only 1-1/2 hours. The problem was that soon after it achieved this very cold evaporator temperature, it started to warm up. This was due to the R-600 (N-Butane) freezing in the final cap tube. At the time I only suspected this as a possible reason behind the warm up, and now I know this with absolute certainty.

So what was different? The final cap tube was much shorter, giving more flow to the evaporator. So yesterday I modified the cap tubes, making them both shorter, CT#1 only slightly, and CT#2 less than half the original length. I also added a bit more Argon to make up for what I knew would be a higher suction pressure, and thus a warmer evaporating environment (the extra Argon decreases the boiling point). And of course I was once again using R-600a (Iso-Butane) to avoid the freezing problem.

So here is a look at how the unit tested in this new configuration...
CryoBUG 7/25/2012 Test Chart
Note: If you forgot what the TC locations were, refer to this diagram: CryoBUG V5 Piping & TC Locations

Final compressor running pressures were 38/227 psig, and the Evap-In was -153°C, with an Evap-Out of -151.5°C. As can be seen, the cool down time was amazing! Hitting -100°C in about 19 minutes, and -150°C in barely over 50 minutes. This is nearly 3 times faster than the 7/24/2012 test run, and even faster than the April test using shorter cap tubes with R-600. Basically a "Guinness Book of World Records" moment for such a small AutoC.

All I can add at the moment, is that I am extremely happy with the progress I'm seeing in this prototype unit.

Saturday, July 21, 2012

CryoBUG Totally Stable at -156°C

Today I retested the CryoBUG unit with the following refrigerants in the charge.

R-600a (Iso-Butane)
R-170 (Ethane)
R-14 (Tetrafluoromethane)

Note: This would be the first time utilizing Iso-Butane instead of N-Butane.

The results were better then expected, allowing the unit to get down to -156°C, while still exhibiting stable temperatures and pressures.

TC Meter connected to Evaporator Feed

Test Data

In order to get a better feel for what the unit was doing at various points across the system, I have included a diagram that hopefully makes this a bit easier to see.

As can be seen, there is a very small temperature spread across the evaporator, being on the order of 4 degrees. The estimated evaporator load should be approximately 10 watts, which is primarily due to insulation losses.

The proposed oil control system shown may not be necessary, since I haven't seen any signs of oil freeze-out problems, which is most likely the result of utilizing two hydrocarbon based refrigerants in the charge, thus having better oil solubility than freons.

Switching to the Iso-Butane has also apparently cured my refrigerant freeze-out problem as I had expected.

During this test I took the opportunity to incorporate smaller cooling fans, going from two 100 CFM models down to a couple of smaller 41 CFM types. This allowed me to better gauge what will be needed for air flow when I build the more finalized version of this unit. As can be seen by the liquid line and discharge temperatures, the smaller fans still did the job (ambient air temperature in the room was about 22°C).

Technology Update

My experiments have shown that a very simple single-stage autocascade refrigeration system can reliably achieve temperatures in the cryogenic range. And furthermore that this can be accomplished with the use of open technology, no longer under the restrictions of patent protection or patent pending. Although obviously there are some trade secrets involved, such as HX and cap tube sizing, and of course the exact formula for the mixed refrigerant charge. Trade secrets that are currently held by Mytek Controls (aka; Michael St. Pierre). But none the less, this does not present any legal obstacle for anyone else wishing to exploit this technology on their own. And I encourage others to do so in the spirit of competition.

I say all this to discourage what I will call the practice of filing BS Patents. These would be patents that should have never been issued, which are nothing more then taking what would be common sense derivatives of previous inventions, and trying to monopolize on their use.

I would also like to acknowledge Andrija Fuderer for his great contribution and inspiration, as well as for providing the foundation that CryoBUG is based upon.

Posted by: Leonard Barden, 

"Andrija Fuderer, who died aged 80 last month, was an eminent chemical engineer and inventor with more than 50 patents to his name."

Link to Full Article: 
Andrija Fuderer, a gifted player and all-round talent, has died aged 80...

Friday, July 20, 2012

CryoBUG Iso-Butane charge experiments about to begin

I got my inexpensive R-600a (Iso-Butane) last week. Thanks to a friend who steered me in a direction I would have never thought of, I was finally able to get my hands on enough of this refrigerant to charge my CryoBUG well over 20 times. The source turned out to be an Ebay store called Med-Supply-Online who stocks a 5.5oz disposable cylinder of Iso-Butane used for fueling a Dental sterilizing burner. As you can imagine this has to be very pure, and according to the MSDS it is.

So who makes this stuff? A company called Wall Lenk does. And when checking online it varies from about $9.00 up to $12.00 per cylinder. Unfortunately most of the sources I found wont sell to individuals unless you are a dentist. But luckily this isn't a condition at Med-Supply-Online.

So here's a look at most of my stash...

After getting the cylinders as pictured, I discovered that I had no way to get the butane out of them. The fitting at the top is actually much more then a piercing type as used on R-134a cans. This one has a depression valve incorporated, that will shut-off if the burner head is removed. So I had to also order a burner head as well, and then modify it for use with a standard 1/4" refrigeration hose (I also got one more cylinder of butane fuel with the purchase of the burner head).

Then to make things easier to deal with, I transferred all 5 butane cylinders into a refrigerant recovery tank.

So I now have 27 ounces (765 grams) of pure Iso-Butane, and have already gone ahead and recharged CryoBUG for a test tomorrow.

The interesting thing I noticed when charging my unit, was that the R-170 (ethane) appears to have a much better affinity for the "Iso" versus the "N" type butane. Or in other words it soaks in more, yielding a lower vapor pressure. This could be very helpful.

Tuesday, June 12, 2012

CryoBUG Controller Assembled

I received my controller boards from Futurlec yesterday, and I got to say they turned out very nicely. And although I'd been hearing rumors about it taking 6 or more weeks, in my case it was just under 4 weeks from order to delivery via Standard Post. Not bad considering they came from Hong Kong.

Here's a look at the un-stuffed PCB (scanned image)...

CryoBUG Controller PCB Top-Side (Click to Enlarge)
CryoBUG Controller PCB Bottom-Side (Click to Enlarge)

All the component holes were correct, although the header connectors were very snug (my fault, need to increase drill size a tiny bit). Quality of the boards was excellent, and the silk screen was very legible (even the Mytek Controls logo detail was faithfully reproduced).

Here's a look with all the components stuffed and soldered...

CryoBUG Controller Assembled - Top View (Click to Enlarge)
CryoBUG Controller Assembled - Bottom View (Click to Enlarge)

I ran a functional check, and I am glad to say everything works as it's suppose to. Although I still need to check the Remote, but this will take making up a cable with some switches and indicators.

The green headers from On Shore Technology accept pluggable screw terminal blocks, making it easy to build the wire harness separately, and then plug it into the board. It also means that swapping out a board in the field becomes a very simple task that most anyone could do, including the customer if need be.

This sure beats the heck out of having to solder up or crimp terminals like on some of the Molex style connectors.

CryoBUG Refrigeration System Status...

Well I'm still trying to decide where to buy my R-600a from. Prices are pretty ridiculous (as in too darn expensive), unless I want to buy a big cylinder, which I don't. I'll keep looking but pretty soon I'll probably just have to bite the bullet and pay the price.

In the mean time, I'm looking at running some oil separation tests to zero in on a proper cap tube size for the oil return back to the compressor. To help with this, I picked up a sight glass, hand valve, and some very small ID capillary tubing. My goal is to spec out a cap tube that is just big enough to do the job, and nothing more, since I wish to have this be a continuous flow situation not dependent upon a valve.

Well I think that wraps it up for now.

Friday, May 11, 2012

CryoBUG PCB Layout Revisited

Without fail there is always something that goes wrong at the last minute, or more specifically something that I overlooked. Well in this case it was the transformer's footprint and pin-out that I totally misrepresented. Apparently I need to do a better job at reading data sheets, because I mistakenly confused the transformer I wanted to use with one of a much lower VA rating, hence a much smaller size. Luckily I held off on having printed circuit boards made until I had all the parts in hand, and had checked them for physical fit against my layout.

So I needed to make room for a bigger transformer. This ended up taking my original 2.5" width to 3", and at the same time it took me out of the running for getting the ExpressPCB Miniboard deal. So here's what the new board layout looks like...

Rev 1.3 PCB Layout  -- Click Image to Enlarge --
I also took the opportunity to clean up some of the trace paths, and to add the Mytek Controls logo to the silkscreen layer. So the final board size is now 3.8" x 3.0". So just for grins, I used the compute board cost feature in the ExpressPCB software and here is what it came up with for production quality board cost.

Qty 2 Boards  = $284.22
Qty 10 boards = $321.94

Wow! That was quite expensive. So next I went to Futurlec's PCB manufacturing page and entered my data. Here is what they showed when calculating the price.

Qty 2 boards   = $  75.70
Qty 10 boards = $134.50
Shipping = $9-14 by Standard Post

So it looks like Futurlec will be making my boards.

Since there were some changes in the transformer pin-outs, and I made some adjustments in the LED drive resistors, including adding a resistor for the external On/Off switch's internal LED back lighting, I updated the schematic to Rev 1.3 as well.

Rev 1.3 Schematic (page 1 of 2)  -- Click Image to Enlarge --

Rev 1.3 Schematic (page 2 of 2)  -- Click Image to Enlarge --

Originally I thought I could purchase the switch I had in mind with a built-in resistor for 12V operation, but as it turns out this would have been a special order, so I opted for the one without the built-in limiting resistors. To accommodate this, a resistor (R19) was added to the board.

I should be ordering boards within the next couple of days.

R600a news

Well my source in England turned out to be a bust. They basically credited my PayPal account with no explanation given as to why the order got aborted. So I am once again searching about for a cheap source of iso-Butane.

Friday, May 4, 2012

CryoBUG Controller PCB Layout

I made a few more changes to the circuit design, so this now brings it up to Rev 1.2 (the design is now locked in). The changes mainly consist of adding another CD4013 Dual RS Flip-Flop IC chip and some discrete support circuitry to enable "Fault Identification" LEDs.

Rev 1.2 Schematic (page 1 of 2)  -- Click Image to Enlarge --

Rev 1.2 Schematic (page 2 of 2)  -- Click Image to Enlarge --

Page 2 of the schematic shows the additional latch circuits that retain the state of the last fault, and display which safety (via an LED indicator) caused the actual fault condition. Was it a high liquid line temperature fault, or an under/over pressure cutout? By knowing which condition caused the fault, it allows for a better response to correct the problem. Typically a high liquid line temperature would be caused by a clogged or blocked air cooled condenser, or perhaps a fan motor malfunction. A low pressure cutout would primarily be caused by a loss of charge due to a leak. And a high pressure cutout could be indicative of an excessive load condition.

I lumped the pressure faults together because quite frankly it was more easily accommodated with only one extra IC chip. And secondly, pressure faults will rarely be a problem. It is far more likely that an overheat condition may occur due to a neglected condenser cleaning.

These two fault LEDs will be mounted so as to be visible from the CryoBUG's control panel, along side the remote connector.

I was able to squeeze all of this circuitry on one 3.8"x2.5" MiniBoard (ExpressPCB's 3 for $51 deal). Although I'll probably opt to go with their MiniBoardPro which is a full production version having solder mask and silkscreen.

Here is the completed PCB design, although I still need to double check that I got everything connected, and that all the components will fit properly.

Rev 1.2 PCB Layout  -- Click Image to Enlarge --

When I'm ready to place my order, it's as simple as making a few selections in the program, give em my credit card info, and about 2 days later they're suppose to arrive at my doorstep. Pretty awesome!

BTW this controller can be set-up for either 115V or 230V operation just by changing the jumpers (W1-W3).

Friday, April 27, 2012

CryoBUG Controller

Well I've been busy over this last week or so. Besides trying to track down a source of cheap R-600a (Isobutane), I have also been designing an electronic controller for the more finalized version of the CryoBUG.

Here is a list of some of the things I wanted to incorporate into this controller.
  1. Toggle compressor power ON and OFF by pressing momentary push button switch.
  2. Have a bi-color led back light in the switch for "ON" as well as "FAULT" indication.
  3. When experiencing a fault such as an Over Pressure condition, to shut-down the compressor and require user intervention in order to reactivate it.
  4. Have the fault indication retained even after the condition that caused the fault has cleared (fault memory).
  5. Provide redundancy in case the logic circuits fail, so that a fault can still shut-down the compressor.
  6. Provide a remote means of both monitoring and control of the CryoBUG functions.
  7. Utilize a short circuit proof control transformer and a sealed compressor relay for hydrocarbon refrigerant safety.
  8. Maintain low EMF and RF emissions by the use of low speed CMOS switching circuits and linear voltage regulation.
And here is the circuit design that incorporates all of these requirements...

Click on Image to Enlarge

And here is a small video showing the bread boarded prototype in action...

Just a little more background on the remote aspect. This design incorporates a fully isolated interface for remote monitoring and control of all the CryoBUG functions. And in the anticipated end user application, there will be several of these units providing water vapor cryopumping for small vacuum coating chambers on an automated process cycle. It is desirable to have all machinery, pumps, ect. controlled remotely by the central computer control system of this vacuum coating system.

Here are the functions provided by the CryoBUG remote interface:
  1. Control: Compressor ON/OFF (universal 6v-24v AC/DC optically isolated input)
  2. Status: Compressor Running (contact closure)
  3. Status: At set point temperature "Ready to Coat" (contact closure)
  4. Status: Unit Fault (contact closure)
(Edit 4/29/12 Rev 1.1: the remote control pin-out has changed as well as some functionality)
If it is desirable to disable the manual On/Off switch via the remote, then holding the remote in an active high state will do so. With this inverted logic, it would be necessary to momentarily remove the remote On/Off signal (bring it Low) in order to toggle power remotely, and then immediately restore the signal (bring it High). However If simultaneous manual and remote operation is desired, then use normal logic for remote power control (momentary High to toggle power).

As for temperature monitoring and set point, I decided that it would be best to utilize an independent process controller such as the Solo SL4824-RR 1/32 DIN from Automation Direct.

These PID controllers are getting to be very inexpensive for all that they are capable of doing. It just wouldn't be practical for me to try and duplicate this functionality into my CryoBUG controller. Although there are some cheap temperature displays available from China, but most of them are not able to measure much below -40°C.

So in my design I routed the relay contact of the process controller through my board, and make it available as a status contact on my remote connector. The user simply enters a set point on the process controller where they wish to be alerted when the CryoBUG has achieved that temperature. This "Ready" signal let's the vacuum coating system know that water vapor cryopumping is active, and that the coating process can begin.

At this stage, the controller design looks pretty solid, so the next step will be to lay out a printed circuit board and get a few fabricated. I'm using the ExpressPCB software to design my board, and will compare their cost to Futurlec's to see which fabricator I wish to use. As far as I know, these are the only two choices available when using that software package. On the plus side, it is extremely easy to use, so much so, that it's truly a pleasure to work with. Also the ExpressPCB software works great on my Linux machines under Wine.

As for other parts of my CryoBUG's development. I'm still waiting for some R-600a to arrive from England, so further charge related tests are temporarily on hold.

Wednesday, April 11, 2012

CryoBUG Stable -150°C Operation

Well it took quite a few different tests and refrigerant variations, as well as some minor hardware changes to finally get what I consider to be a stable -150°C unit. To qualify for this prestigious stability award, it must have virtually no fluctuation in evaporator temperature (+/- 0.5°C), and meet the target temperature requirement (-150°C or colder). Today we did just that.

Evaporator Temperature -152 C, Total Unit Current Draw 2.36 Amps

Compressor running pressures were equally amazing at 26/180 psig, which is also the reason why the current draw was so moderate at 283 VA (226 Watts, PF=0.8).

CryoBUG 4-11-2012 --- 4.5 hour Test Run
Other then the strange oscillation in the CC SUCT temperature, the Evap was about as Flat-Lined as one could hope for. Reference V5 Piping Diagram for TC locations.

Cool down was also very fast and steady...
-100 C @33 minutes
-110 C @36 minutes
-120 C @41 minutes
-130 C @64 minutes
-135 C @70 minutes
-140 C @77 minutes
-145 C @85 minutes
-150 C @98 minutes
-152 C @just under 2 hours (flat line)

So what did I do to achieve this? Basically one thing, and that was to dramatically reduce the flow in the final cap tube (CT#2). The charge was the same N-butane, Ethane, R-14, Argon blend as my first V5 hydrocarbon test run, and in the same proportions.

The funny thing was it took me in quite a circle trying to zero in on what exactly caused my initial freeze-out problems. I tried using Propane instead of N-butane, which didn't freeze, but also didn't yield temperatures nearly as cold. This was probably due to the higher suction and discharge pressures. So I tried reducing cap tube flow, but still couldn't get down to where I wanted to be while still using Propane in the charge. Next I went back to N-butane, but used less of it. Still I wasn't able to achieve my target temperature. So I added more R-14, and got a tad bit of improvement. And then added more only to see the improvement go away, as the suction pressure once again started climbing too high.

So finally just for grins, I decided to try the original N-butane charge at the same proportions that I had originally started out with, while hoping that just by some weird stroke of luck the resized final cap tube would do the trick. Well it sure did!

The only other thing worth mentioning, is that the compressor discharge temperature was very comfortable, settling in at 51°C. This combined with the good running pressures, should insure a long and happy life for the compressor.

What's next?

I would like to try Isobutane (R-600a) in place of the N-butane (R-600). My gut feeling tells me that this should yield similar test results. And with the Isobutane's lower freezing point of -160°C, I would feel a lot more comfortable in the long term (N-butane freezes at -138°C). Now I just need to get my hands on some of this.

Wednesday, April 4, 2012

CryoBUG at -155°C and dropping

Yesterday afternoon CryoBUG made it down to -155°C in 1 hour and 30 minutes after start-up.

Over the next 1/2 hour it continued to drop, getting down to -159°C before showing signs of something freezing out (not sure if it was the oil or the butane). Yeah you heard that right, I've switched over to hydrocarbons in this charge (butane and ethane). I also went back to a single-stage AutoC, because quite frankly 2-stages without an auxiliary condenser just doesn't work all that well.

The reason for all these changes, is that I just wasn't having much success at getting the "BUG" to work with HFCs only. HFCs in the area of interest, all have relatively warm freezing points. With the BUG's simple design and very few separation points, it just kept freezing up around -95 to -105°C, or showing general signs of instability evidenced by erratic pressure changes.

So in this latest iteration, the refrigerant charge consists of:

R-600 (N-Butane) Boiling Point  = -1°C, Freezing Point = -138°C
R-170 (Ethane) Boiling Point  = -88°C, Freezing Point = -183°C
R-14 (Tetrafluoromethane)  Boiling Point  = -129°C, Freezing Point = -184°C
Argon Boiling Point  = -186°C, Freezing Point = -189°C

Total amount = 90 grams (3.17 ounces), static balance pressure = 235 psig

As can be seen, the butane can still pose a freezing problem if it gets near the evaporator when going below -138°C. So this might be the cause of the freeze-out problem I saw occurring towards the end of my test yesterday. Or it could be the oil. Or it could be overcharged, with the warmer boiling refrigerants flooding out the Phase Separator and spilling over into the next stage. Not really sure at this point.

The latest design has gone back to Andrija Fuderer's single-stage AutoC (patent #3203194), which has shown excellent promise, assuming that the freeze-out problems can be solved. In this current revision (V5), I have reduced the tubing size of the heat exchangers, and gone with a very simple phase separator design. I also decided to keep all the cap tubes external to the HX, unlike what I did on previous versions where the #1 cap tube was slid inside the suction side of the Cascade Condenser. Keeping the cap tubes on the outside allowed me to down size the HX, while still having minimal pressure drop on the suction circuit.

Yesterday's test was without any kind of oil management system in place, so in other words, the compressor's discharge line goes directly to the air cooled condenser. The cascade condenser and the subcooler are actually formed as one tube-in-tube heat exchanger, with cap tube #1 entering through a small hole drilled into the outer (suction) tube about mid way.

And now for the moment you have all been patiently waiting for. I give you the test chart from yesterday's test run.

CryoBUG 4/3/2012 Test Chart (Hydrocarbon Charge)

Pull down was very fast.

28 minutes to -100°C
49 minutes to -130°C
57 minutes to -140°C
72 minutes to -150°C
90 minutes to -155°C

Running pressures were 24/176 psig when the evaporator hit -155°C, and compressor current draw was very reasonable at 2.4 amps.

Highest peak discharge pressure was 332 psig, and peak current draw equaled 3.4 amps (this was within 15 minutes of start-up). One thing to keep in mind is that this is all being done with a single 4.8cc/rev compressor rated for 4810 BTU/Hr.

I think for my next test I'm going to substitute R-290 (propane) for the R-600 (butane), since the propane has a colder freezing point of -186°C. This will let me determine if the oil is the cause of the freeze-out problem, by ruling out the refrigerants.

Monday, March 26, 2012

CryoBUG Version 4 assembly

Well I learned a few things over the past several weeks when working with the CryoBUG. Not the least of which was the fact that the BUG was growing, and would soon be too big for what I had in mind. So going back to the drawing board, and taking into consideration what I'd seen in the last few months of testing, I decided to shrink things down and squish that BUG so-to-speak.

Below will be seen the result of this new design effort.

CryoBUG New HX Stack (on the left) compared to previous version (on the right)

The upper most coil in both cases is the dummy evaporator being used for test purposes only.

CryoBUG Version 4 complete prototype assembly

So in this latest version you can see that the HX stack (minus evaporator) is no larger than the compressor. This particular design is a 2-stage AutoC with no auxiliary condenser proceeding the 1st phase separator (Temprite 340 oil separator). The 2nd phase separator is made up from a 3/8" copper bullet strainer (almost hidden behind right side), which in proportion to the stack, is of appropriate volume to do a very reasonable job of separating the liquid condensate from the vapor at this stage.

When I get a chance, I'll post an updated piping diagram of this new CryoBUG version.

If you look over towards the left side, you'll see that I have now incorporated a High Pressure Cut Out (HPCO) and used it's housing to also add a power ON-OFF switch. Since I have been steadily increasing the amount of refrigerant as well as the Argon percentage, I thought it would be a good idea to have this extra measure of safety.

Testing should resume in a few more days. Just need to get some thermocouples attached and insulate the HX stack.