CryoBUG is Back

[n00b 0f l337]

Kang, he said it was just a test charge to make sure the system was functional.

[mytekcontrols]

Kang, he said it was just a test charge to make sure the system was functional.

Yep that is correct, just a test charge to be sure nothing was blocked. I have had it down around -142C since that time, but this was with a poor vacuum serving as the cold head insulation. And I am definitely seeing the effect of having considerably more mass to contend with then my simple 1/4" tubing test coil as used on the breadboard prototype. Because of this, I am once again tuning the charge blend to better optimize it in its new configuration. Although I probably won't get -156C this time around.

A better Cold Head vacuum feed-thru design would help with this, since through conduction, the Cold Head is picking up more load then it should. But it will suffice for my Demo unit.

[mytekcontrols]

Hi everyone

I know it has been a long time since I last posted on these forums, and I am truly sorry for that. But around 10 months ago I got a job offer I couldn't refuse, which promises to last for at least the next 5 years. It really has been challenging though, and didn't allow me much stress-free time, or put me in the mood for sharing (and there were, and still are things I can not talk about with this new venture). Anyway I'm not building weapons of mass destruction, or doing anything illegal, just very busy.

So what's up with CryoBUG?

I'm glad to say it is alive and very much kicking

Here is a pic showing off it's latest skin (this got redesigned, and now allows for easier side panel removal).



Link to informational web page: www.cryobug.com

You'll also notice that there are a couple of fans added to the side. I discovered with the old panels installed, it was getting a bit too hot inside, so I added a pair of 40 CFM low profile fans to help pull the air through. This actually works quite well. The front panel also got redesigned, sporting a cleaner look.

I also switched over to an all HC charge (+Argon), thus making it totally 'Green' and ECO friendly. Here are a couple of charts showing the capacity (Heat Load vs Temperature), and the cool down curve.





If you look at the first chart, it will be apparent that performance above 50 watts drops off quickly. And I know 50 watts may not seem like a lot compared to many of the CPU chillers being built by the members here, but it is perfectly suitable for the intended application which is probably around 25-30 watts. Also it is quite impressive that a single 4.8 cc3/rev compressor can achieve these temperatures.

I was very happy to be able to get rid of the R14, which was the only non natural refrigerant in use. Going all HC actually ended up improving my cool down time, and upped the heat load capacity slightly, while still keeping compressor pressures reasonable. And the temperature stability is excellent, no more hunting as it was prone to do before. Also power requirements have dropped. Formally it had a full load amps approaching 4.5, but now it is just a bit below 4 (@120 VAC). So all in all, CryoBUG looks to be ready for show time.

Well almost. Still need to conduct some extended run time experiments to be sure that it doesn't become unstable.

[Buckeye]

Hey man good to hear from you again !!
Happy New Year to you and your family !

Good deal on the new job, sounds pretty exciting

That box is looking very sweet and professional, nice work !

Keep us updated when you can

[Patrickclouds]

very nice results. congrats.
so you only use a 4,8ccm mini rotary and achieve -140?C at 30 Watt?
very impressive

my 6,4ccm Rotary with R404a achieves only -38?C to -40?C at that load. http://www.xtremesystems.org/forums/...78#post5191978
so maybe I should also try with some mixed gas Joule Thomson process?

Happy new year.
I'm still following your blog: http://www.mytekcontrols.com/

[mytekcontrols]

very nice results. congrats.
so you only use a 4,8ccm mini rotary and achieve -140?C at 30 Watt?
very impressive

my 6,4ccm Rotary with R404a achieves only -38?C to -40?C at that load. http://www.xtremesystems.org/forums/...78#post5191978
so maybe I should also try with some mixed gas Joule Thomson process?

Happy new year.
I'm still following your blog: http://www.mytekcontrols.com/

Thanks

Yeah when you need to go below -40 C, autocascading has some distinct advantages. The biggest problem with a single stage at or below this temperature, is that your mass flow starts to suffer because of the need to run a low suction pressure. CryoBUG runs 50 psi suction and 275 psi discharge for -150 C at 15-20 watts heat load. With those kind of pressures, mass flow is very good.

Glad to hear you are still following my blog, sorry that it was rather boring for the last 10 months.

Happy New Year to you as well. Lets hope that it is a good one

[mytekcontrols]

Hey man good to hear from you again !!
Happy New Year to you and your family !

Good deal on the new job, sounds pretty exciting

That box is looking very sweet and professional, nice work !

Keep us updated when you can

I'll try, but it looks like next year things will be getting crazy once again (start work on next secret project for client).

Still have plans to market the CryoBUG design for licensing, but even that has to take 2nd seat to my new job.

I wish you and yours a Happy New Year as well

[Patrickclouds]

is there no oilseperator necessary in your final build?

[n00b 0f l337]

I don't believe there is one on this one Patrick.
Looking good Mytek, was just thinking of you; you popped up in my email the other day incidentally. Hope all is well and 2014 will be good.

[mytekcontrols]

is there no oilseperator necessary in your final build?

That would be a no on the oil separator. With the predominately HC charge, I haven't seen any problems thus far, even on several day runs. With the phase separator running between -30 to -65 C (dependant on load), it appears to do a good job of returning oil back to the compressor. Of course I'll need to really give it a shake down, so in about a week I'll be starting it up, and leaving it running for several weeks 24/7. Hopefully it'll do well.

Hi Adam

I was thinking about you as well over the last few months. Still doing the college thing? Any projects?

[n00b 0f l337]

Med school the last few years, about to start my residency.
I saw an update on your LinkedIn actually, what are you working on these days? That blog is a bit empty!

[mytekcontrols]

Med school the last few years, about to start my residency.
I saw an update on your LinkedIn actually, what are you working on these days? That blog is a bit empty!

Almost a year ago I got approached by a third party to help develop some refrigeration products for them. Anyway long story short, I accepted their offer and have been very busy ever since. I now have my own 1000 sqft shop complete with all the tools and gases I need to do autocascade design. And I just completed the first of several product designs for them. I'm under an NDA so I can't say too much more about what I am working on or even who I am doing it for.

However CryoBUG is my creation and was not part of this agreement, so I still have plans to see this marketed. I also intend to develop an even smaller version which could work well for the detector cooling arena. That one will go by the name of CryoSPRITE, and be less than half the size of its bigger brother.

So it sounds like you have also been, and will be busy for quite some time with your medical pursuits. I imagine not much spare time for you either.

[samsarulz]

Sorry if its an inappropiate question but how much did you invest in the CryoBug Prohect (more or less)?. It really looks promising. Im still digesting most of the information displayed.

Have you already tested it with some CPU (or GPU) and how it performs. Well Im very satisfied with the external aesthetic of the case. Very professional.

Hope to see more projects!

Jose Luis

[Buckeye]

I don't think it's intended use is for CPU's, it does not have the load handling for that. Rather I believe it's intended use is in manufacturing of things like HDD Read/Write heads etc.

[mytekcontrols]

Sorry if its an inappropiate question but how much did you invest in the CryoBug Prohect (more or less)?. It really looks promising. Im still digesting most of the information displayed.

Have you already tested it with some CPU (or GPU) and how it performs. Well Im very satisfied with the external aesthetic of the case. Very professional.

Hope to see more projects!

Jose Luis

Hi Jose,

As Buckeye pointed out this is not for CPU cooling. It just doesn't have enough heat load capacity for that application, although with a different charge it might be possible (R600 and R1150 would do nicely).

The applications I'm shooting for are mainly small vacuum coating systems for things like depositing the magnetic media on the disk within a hard drive. So in this application CryoBUG would be used to freeze the water vapor floating around in the coating chamber, something that is referred to as water-vapor-cryopumping.

The custom enclosure design was quite expensive, even more so when I decided to change the front and side panels which became a do over. With all things considerdd I probably have $2.5k invested counting refrigerants that were used in development. But keep in mind that this is a one off Demo unit and it tends to cost more to achieve this look when only making one of something. In a production version the enclosure would be done completely different so as to bring the cost down.

[samsarulz]

I thought that was more than that jaja considering the details of the case. Lets hope to see this unit by production, but first (as you said) you have to test a different charge. BTW I didnt knew about that kind of stuff regarding disk read/write heads jajaja.

[ultralo1]

Nice work Micheal. Glad to see your still working and sharing projects. Glad to here the new business venture is going well. Hope to see the Cryobug and Cryosprite in production sometime soon.

[mytekcontrols]

Nice work Micheal. Glad to see your still working and sharing projects. Glad to here the new business venture is going well. Hope to see the Cryobug and Cryosprite in production sometime soon.

Hi Ultralo,

Now I'm good with handles, but horrible with remembering names. Would Brent be correct? It's been a long time, and my mind has gotten a little mushy.

Yeah things are chugging along, although it took way longer then I anticipated to get CryoBUG to this point. But at least I'm amongst other people that share my same problem (gotta get that cloning machine working).

So what's been up with you? Still working on sub zero equipment?

[ultralo1]

Hi Ultralo,

Now I'm good with handles, but horrible with remembering names. Would Brent be correct? It's been a long time, and my mind has gotten a little mushy.

Yeah things are chugging along, although it took way longer then I anticipated to get CryoBUG to this point. But at least I'm amongst other people that share my same problem (gotta get that cloning machine working).

So what's been up with you? Still working on sub zero equipment?

Your mind is anything but Mushy Name was close (Britt) and still working on the same stuff at the same place. Nothing much has changed accept the kids are taking more of my time (as it should be) so not much time for hobbies. I stop by here and check on things once and a while. I cant add much to the discussions but I do like reading up on things. I cant wait for your NDAs to expire so we can see the neat stuff youve been up to. Your knowledge and expeirience is amazing and your willingness to help is unparalleled.

[mytekcontrols]

Your mind is anything but Mushy Name was close (Britt) and still working on the same stuff at the same place. Nothing much has changed accept the kids are taking more of my time (as it should be) so not much time for hobbies. I stop by here and check on things once and a while. I cant add much to the discussions but I do like reading up on things. I cant wait for your NDAs to expire so we can see the neat stuff youve been up to. Your knowledge and expeirience is amazing and your willingness to help is unparalleled.

Hi Britt,
Yeah I'm pretty much the same as you with having time for the hobbies, although in my case the kids just bought their own house late last year (I think I'm a bit older than you).

I wish I had more time (as in all the time) to dedicate to pure research, because for me that really is the fun stuff. However with my current job I find myself doing a ton of documentation and training, in order to transfer the design details to my benefactors. So when there is an active project, only a small fraction is actually about experimentation.

I do miss doing the tutorials here on XS, but it kinda seems like the interest has been dying out, and my time is limited.

Have fun with your kids Britt, because they'll be grown up before you know it


And now to get back on topic, here are some observations that I have recently made with CryoBUG related to the refrigerant charge.

The CryoBUG project has been very educational for me, and also heavily influenced by others here on XS. When I first got on this forum I had never delved into using HC refrigerants, being kinda fearful at what might happen. But after seeing what others were doing, realizing the cost savings, and obtaining a much better understanding of the dangers and how to avoid them, I incrementally switched over to HC refrigerants. In my final formulation I was able to fully realize an all HC/Argon blend based on 4 components (R600a, R1150, R50, Argon). However it does have its quirks, and requires a good balance in order to avoid some of the pitfalls of not having a -128C refrigerant (R14) in the mix to better bridge the gap. One such pitfall being a 'self-refrigeration' effect (as I call it).

This 'self-refrigeration' effect was actually something that Dale Missimer (Polycold's Creator) had used to describe a problem in his early autocascade experiments when he was working for Conrad/Missimer in the early 70's, where after initial cool down the evaporator would begin to warm up, and the warmer stages would start getting much much colder. And dependent upon the choice of lower boiling point refrigerants, you could produce a situation where for instance in the case of CryoBUG; the R1150 that was needed in a blend with R50 and Argon to produce a pseudo refrigerant that allowed condensation at moderate pressure (=< 250 psi) and temperature (=> -90C), gets completely taken away by a lower stage's phase separation, thus leaving only the R50 and Argon which can not form a condensate under those conditions. So the evaporator warms up, the lower stages get even colder because of the disconnection from the evaporator load, and the problem gets worse and worse as the colder phase separator temperatures pull more and more of the R1150 out of the gas stream.

Note: at 250 psi (approximately 17 bar), R50 condenses at or near -112C (See: R50 Properties)

I've seen this effect occur with R1150 as the R14 substitute, and accelerated this by swapping R170 in for the R1150, which produced an even more exaggerated self-refrigeration effect, where the evaporator which had dropped down to -145C, suddenly turned around and warmed up to -115C within a matter of 60 minutes. So basically the R170 having an even warmer boiling point than the R1150, dissolved much easier into the R600a, and once completely taken out of the gas stream predominately left only gas cooling for the evaporator without any real phase change.

So to make this work, R1150 is better than R170, and a careful balance of this and the R600a is required in order to produce phase separator temperatures that still allow for some R1150 to pass out of the separator into the remaining gas stream. Of course this is an over simplification, because many other factors such as compressor through put, cap tube sizing, and the influence of other refrigerants all come into play. But you get the basic idea.

[Patrickclouds]

I've often thought about that problem that the liquification defelects in the earlier heat exchangers the colder the hx stack gets.
Should it be possible to control this effect by selecting the captube a bit longer that the temperature spread over the hx is wider? Because the evaporating refrigerant component gets more superheated and causes lower cooling capacity because of the lower mass flow.

I'm planning on getting some experiance by building an autocascade. I have acces on r600a, ethene, methane and argon so exactly the refrigerants that you used.
I want to achieve something about 200 watts cooling capacity, so I'm not 100% sure of the design. I plan on using two phase seps and maybe use two cap tubes at the end of the hx stack. One for evap and one for the hx stack itself. I think one capetube could cause a very large superheat after evaporator with the 200 watt goal and so maybe I loose temperature for the liquification in the hx stack (the last heat exchanger).
I'm still on research and open a topic when I start building.

[mytekcontrols]

I've often thought about that problem that the liquification defelects in the earlier heat exchangers the colder the hx stack gets.
Should it be possible to control this effect by selecting the captube a bit longer that the temperature spread over the hx is wider? Because the evaporating refrigerant component gets more superheated and causes lower cooling capacity because of the lower mass flow.

I'm planning on getting some experiance by building an autocascade. I have acces on r600a, ethene, methane and argon so exactly the refrigerants that you used.
I want to achieve something about 200 watts cooling capacity, so I'm not 100% sure of the design. I plan on using two phase seps and maybe use two cap tubes at the end of the hx stack. One for evap and one for the hx stack itself. I think one capetube could cause a very large superheat after evaporator with the 200 watt goal and so maybe I loose temperature for the liquification in the hx stack (the last heat exchanger).
I'm still on research and open a topic when I start building.

That's an interesting idea at the end point, dividing the condensate to feed back on the final cascade. Never tried something exactly like that, although I have done this with a sub-cooler HX to yield colder evaporator temperatures.

On the idea of using a longer cap tube (less flow) to better control the liquid deflection as you call it; at moderate to low load conditions this might very well have the opposite effect you are proposing, and cause the phase separators to run colder, thus deflecting more liquid. Another idea would be to down size the HX prior to a given phase separator.

In my experiments I am finding that using less R600a tends to give me what I am after, since this both influences the phase separator temperature (warms it up), but also gives less sub-cooled liquid volume for the R1150 to soak into. Of course I am only using one phase separator, and having two might complicate matters.

Anyway there are a lot of variables that can be played with to make this work better, with the object being to not steal all of the R1150 before getting to the final stage.

Good luck on your AutoC project

[n00b 0f l337]

One of the thoughts there with two final caps is something we saw from cryo-tek I believe. He forked the suction at the end, one to the evap and one back to the us stack. In his design tho the final hx was only fed by that line, and thus received no flow back from the true evaporator. His explanation was such a drastic subcooler allowed for much lower temperatures with higher argon loads (exactly what you use argon for), but required a much larger system as this final subcooler could easily become a third or more of the massflow.

[mytekcontrols]

One of the thoughts there with two final caps is something we saw from cryo-tek I believe. He forked the suction at the end, one to the evap and one back to the us stack. In his design tho the final hx was only fed by that line, and thus received no flow back from the true evaporator. His explanation was such a drastic subcooler allowed for much lower temperatures with higher argon loads (exactly what you use argon for), but required a much larger system as this final subcooler could easily become a third or more of the massflow.

Actually what you just described is a standard part of most all of the Polycold units, and is also something I implemented near the end of the AC2 project. We call it an 'active' sub-cooler for lack of a better term (CryoBUG uses a passive sub-cooler). In the active sub-cooler implementation about 20-30% of the available condensate coming out the end is diverted back to feed the suction side of this same HX. Meanwhile the evaporator return bypasses the sub-cooler entirely as you described, and feeds back through the final cascade HX (whatever that may be depending on the number of stages).

Since the amount of heat that can be removed using this method is relatively small, so is the required size of the sub-cooler HX. The main benefit of using an active sub-cooler vs a passive one, is that the quality of the condensate feeding the evaporator remains good even when taking on high loads. The disadvatage is that you are stealing condensate to feed the sub-cooler instead of using it to directly cool the evaporator, so you may lose some capacity. This is kinda a catch 22, and needs a carefully balanced approach to reap the rewards.

[n00b 0f l337]

Yeap that is 110% how it was originally explained to me. And I do think I remember you implementing something similar in your AC2. Saccing load for temp or temp quality... That's always been the name of the game.