CryoBUG is Back

[mytekcontrols]

Quiet a nice pattern, simple and functional

It reminds me my wiring, series of normally-closed switch embedded into a loop with relay's coil

Thank you

Yes I was thinking of doing a latching relay circuit with the normally-closed safety switches breaking the coil connection, thereby unlatching the relay. But then I also wanted a momentary push On, push OFF action through a single push button switch, and some sort of fault memory/display. So although this could have been done by adding additional relays, I though it best to do some of the logic electronically instead. And of course CMOS chips are extremely cheap compared to relays.

Another method would have been to use a PIC or Atmel microprocessor as the logic, which would have cut down on some of the components, but not enough to have really changed the cost. And I also didn't want to take the chance of creating any EMF/RF emission problems that might have resulted (although I could have run the clock speed very slow to counteract this possibility). And I was also afraid of me being tempted to use the programmability and built-in peripheral support to add all sorts of bells and whistles with all the inherent bugs that might come with it (creeping featurism).

So yes I decided to bite the bullet and keep it simple.

[Sk_rmouche]

I can tell that PIC's aren't very reliable in wild conditions, my first speed control was based on a 16F type, didn't live for a while
Finally I went with a NE555, limited but amazingly hefty ( I've one running for over a year @ > 70°C in a home-made lamp ).

Edit : back to your AC, I read somewhere that pentane is sometimes used to fight clogging issues in valves or even cap tube, what do you think ? Otherwise, it's a liquid @ ambiant pressure/T°, I don't really understand how it couldn't damage the compressor with continuous liquid pentane at suction :o

[mytekcontrols]

Never used Pentane, but with a boiling point of 37c I would tend to agree with you. And yes I would also think that it could damage the compressor at any appreciable amount. I guess you could heat exchange directly with the compressor discharge (before the air condenser) to flash it off to a vapor before entering the compressor's suction.

Although AutoC's have been successfully charged with fairly large quantities of HFE-7000 which boils at 34c. But I think the difference is that the HFE has absolutely no miscibility with the oil, so it can't act like a solvent, and wash the oil off of the cylinder walls like Pentane surely could. However in both cases, you still need be careful with how much is in the refrigerant stream as a liquid entering the compression stage, so as not to cause physical breakage.

It is also felt that using such a high boiling point liquid in the AutoC blend, yields a low boiling point blend in the 1st phase separator (assuming an auxiliary condenser precedes it) by acting as a SubCooled "sponge" for absorption of lower boiling gases. And since it has such a warm boiling point in itself, does not get past the 1st phase separator to contaminate later stages. (Edit: Only a relatively tiny amount would get by)

And it is also believed that using such a warm boiling refrigerant in the blend can aid heat transfer in the hot discharge stream as it passes into the air cooled condenser.

Of course these are all theory's, but there does seem to be suporting evidence for this in the effects being seen in working systems.

[n00b 0f l337]

I'd almost be worried that such a refrigerant in larger systems that use oil separators could actually liquidfy before the oil separator. Then again you could just move them before the condenser and damn the consequences (aka upscale there size).
As to prevent liquid return to the compressor, would certain kinds of compressors be rather immune to this? I'm thinking of those that enter the suction to the shell and oil. Might really help compressors temperatures.

[ultralo1]

Pentane is still used in several cascade systems. Sanyo ULTs use 2oz in both stages. First stage refrigerant is 407d, second stage is suva 95. I dont have the charge particulars with me. I believe the pentane mixes with the oil to keep its viscosity up at low temps. I do not think that it is used as a refrigerant in these systems. Its more of an oil additive in these applications.

[n00b 0f l337]

Interesting, didn't think of the fact that oil is passing through the compressor piston/rotary as well. Duh. Guess it wouldn't cause an issue.

[Sk_rmouche]

I guess you could heat exchange directly with the compressor discharge (before the air condenser) to flash it off to a vapor before entering the compressor's suction.

I think it could work, hope it doesn't affect too much the efficiency, because it's basically a SLHX, right ?
Maybe a suction bottle could do the job, but you have to get back the pentane into the circuit anyway... Piston type compressor could be used, since most of them aren't very sensitive to floodback ( vapours actually runs all over the windings, heat of compressor is maybe sufficient to turn pentane into vapours ... )

Clearly, liquid solvent in the cylinder is certainly the worst combination ever

[n00b 0f l337]

Not so sure it would strip oil from the walls. It's also a hydrocarbon, would just form a composite blend with the oil, as the propane and such does as well. Infact, most of the refrigerants are non-polar or minimally polar. Would all have that effect. Maybe more so if it's a liquid, but it's travelling quite quickly and isn't settling and such. Plus theres more oil then pentane.

[Sk_rmouche]

Yes, it's quite not an easy phenomenon, I want to buy some pentane to test, but it's expensive and not easy to find.
Being a pure hydrocarbon is maybe the reason why it's used, halogenated HC washes oil easily though... :/

[n00b 0f l337]

I don't see it "washing oil" off the inside of the compressor. It will be mixed and indulated with the oil surely, so it will either leave behind oil, or leave behind oil mixed with pentane. It won't just scrub the thing shiny clean as it comes thru as they'res more oil and pentane right behind that oil and pentane blob. Same with any hydrocarbon. Same with any refrigerant.

I don't see that as an issue or even what might happen.

[ultralo1]

Pentane does not cause washing of the cylinder walls or piston. It will be a week or so before I am in the shop again, when i get back I will get the exact oil charge and pentane charge for you. It simply thins the oil, all of them mineral, AB, POE.

Pentane can be pruchased from any scientific supply. I use Fisher here in the states. Its about $50 a liter for GC grade n-pentane. The hard part is putting liquid into a system that is under vacuum without introducing air into the system.

[Sk_rmouche]

But there isn't only pentane that could be used to additive for oil. Here in France I know someone using gaseous additives for a R23 cascade, I have to ask him what was exactly the blend of HFC/HC he used on that

[mytekcontrols]

Hmm... Very interesting discussion. Although I don't plan on using any of this in the CryoBUG. For one thing it's freezing point concerns me, and as most of you are aware, I don't have a lot of phase separation going on. My ultimate goal is to have a charge that is entirely composed of refrigerants all with freezing points =< -160c. With R600a, R170, R14, and Argon I'll achieve that. Also I want to keep the flammable portion below 57 grams, which will certainly be friendly to the new HC rules. I also don't think that will be a problem (I'm currently at 50 grams).

On a separate subject, I accidentally left out a resistor on my CryoBUG Controller schematic, and I discovered a problem with the workings of the remote On/Off circuit (very minor). So I'll be posting an update to the schematic in a day or two (probably just replace the one I have already attached in the early post with the new one, I'll give it the rev 1.1).

Hey Britt good to see you back.

[Sk_rmouche]

50 grams with all HC's !? I've over 1 kg of R404a and almost 500 gr of R23, it's just enough, hope it doesn't take too much with R14

With AC's pattern, what would you do to get T° < -160°C ? I know that kind of T° isn't reached with classic cascade pattern because of the ridiculous efficiency, and T° under -190°C, LN2 is commonly used, what about that -170°C range ? Here in France there's a company named Findus, they uses a very special tunnel @ -165°C to freeze food, it's running with methane as the last refrigerant, but I haven't been able to get the idea, if it's classic cascade or AC

[mytekcontrols]

50 grams with all HC's !? I've over 1 kg of R404a and almost 500 gr of R23, it's just enough, hope it doesn't take too much with R14

With AC's pattern, what would you do to get T° < -160°C ? I know that kind of T° isn't reached with classic cascade pattern because of the ridiculous efficiency, and T° under -190°C, LN2 is commonly used, what about that -170°C range ? Here in France there's a company named Findus, they uses a very special tunnel @ -165°C to freeze food, it's running with methane as the last refrigerant, but I haven't been able to get the idea, if it's classic cascade or AC

Less than -160°C with an AutoC would most likely require additional stages and gases. I don't think you could reliably achieve it with only a single cascade as I am doing on The CryoBUG. Polycold makes what is called the CryoTiger and now the PCC which are not AutoC's (more like a JT system) that gets pretty close to LN2 temperatures, but it does so at the sacrifice of the compressor's normal lifespan. It gets so hot that they had to put extended heat surface around the body in the form of heat fins, and even then it still gets pretty toasty.

Using methane as a refrigerant would certainly help, but you would still need something else a bit lower in boiling point to get any appreciable mass flow at those kind of temps.

[Sk_rmouche]

Okay, thanks it's pretty interesting

I had a thought this morning, is it interesting to run a classic 4 stager cascade with R50 at the end, but to that last stage, making an AC ? You get the idea ? Mixing the two ways to get some cryogenic T° with classical compression technique. You know something like R404a/R23/R14 and then an AC with R50 and some other gases, like Argon for instance ? It could be nice, even if for some load, a huge compressor is a requirement ...

[Kevin Hotton]

Hi Sk_rmouche

I have explored an AutoCascade 3rd stage for a three stage cascade using all HC's with methane in the final blend. Evaporator was at -130C. See my post on 3-stage cascade with 4-compressors. Yes if R-14 is available (the perfect 3rd stage cascade refrigerant) then a modified blend for a 4th stage AutoCascade would likely reach below -150C.

Kevin

[Sk_rmouche]

Thanks to both of you

R14 is hard to find, but I'm not giving up on that, even if I have to use in the worst case, some ethylene, I'll have my cooling chamber functional soon or later

[ultralo1]

Micheal,
Have you looked at R245FA yet? Its taken the place of r142B in some commercial units. They are getting much better results out of the units. Units that used to struggle to make -154 are now exceeding -160.

[mytekcontrols]

Micheal,
Have you looked at R245FA yet? Its taken the place of r142B in some commercial units. They are getting much better results out of the units. Units that used to struggle to make -154 are now exceeding -160.

The freezing point (-80c) is too warm for my application. And I doubt if it forms as tight of a bond with the oil as butane does. Remember I only have a single phase separator in my design.

[n00b 0f l337]

Could it be that the much warmer refrigerants are easier to separate and aren't making it into the final stages? But I also assume those systems have to have multiple stages of separation.

[mytekcontrols]

Could it be that the much warmer refrigerants are easier to separate and aren't making it into the final stages? But I also assume those systems have to have multiple stages of separation.

Yes it is less likely that any appreciable amount will escape separation when using warmer boiling components. So whatever is passing on to the next stage is more likely to be of benefit and not simply going along for the ride. And yes the systems that Britt was talking about most certainly do have several stages of phase separation, and in this aspect have very little in common with the CryoBUG other than being AutoC's. However I think the basic mechanisms behind how the refrigerants interact within the blend, at different pressures and temperatures is very much the same. So a lot of the theory we have already discussed about warmer boiling refrigerants acting as a "sponge" for the colder ones when in a subcooled region still very much apply. I have also seen compressor cooling improve with the use of warmer boiling liquids such as R114, R123, R600, and HFE-7000 (better rejection of heat). And as an end result, overall system performance improved. So I would think that the same would hold true for the use of R245.

[mytekcontrols]

I've finalized (locked in) the circuit design at Rev 1.2, and have layed out a Printed Circuit Board based on this design.

Changes:
Added fault indication ID. Essentially this consists of another memory latch circuit with LED indication to let the operator know what type of fault has occurred (High Liquid Temperature or Under/Over Pressure).

Implementation required adding one more CD4013 CMOS Dual R/S Flip-Flop, and some miscellaneous discrete support components.

Rev 1.2 Schematic (page 1 of 2)...
CryoBUG-Controller_Rev1_2-Page1.png

Rev 1.2 Schematic (page 2 of 2)...
CryoBUG-Controller_Rev1_2-Page2.png

Rev 1.2 PCB Layout...
CryoBUG-Controller-PCB_Rev1_2.png

Board size = 3.8"x2.5"

[mytekcontrols]

Well I ended up having to make the board a bit bigger due to the reality of fitting a slightly larger transformer. So this knocked me out of the running to get the mini-board deal through ExpressPCB. And the pricing for having a larger board made was simply ridiculous. So I ended up going through Futurlec and got a very good deal on 10 boards with shipping at $134.50. Total time from ordering to arrival on my doorstep was just under 4 weeks (shipped from Hong Kong).

The quality is quite good, with even the fine detail on the silk screen being reproduced without a problem.

CryoBUG_PCB_top_(800x638).jpg

And the final assembly went off without a hitch.

CryoBug Controller Assy.jpg

And best part of all, it works

What's Next?

I'm going to be running some oil separator tests, with the goal of selecting a properly sized oil return cap tube that can be left continuously connected (no hand valve). To this end I have purchased a sight glass and some very small ID stainless steel capillary tubing.

[n00b 0f l337]

Any reason for SS over copper on the cap tube?