Christmas (Benching) Box

[Gray Mole]

Well it started out as a little test of captube lengths and ended up a 'theme' unit again. I seem to get a thought in my head and can't really leave it alone, but that's the OCD kicking in, as always.

So the unit specs are...

Matsus/hita Rotary - 49LRA/3/4hp
Chilly1 spiral evap
9000btu condensor
200cfm barrel fan
30" flex line
R290 gas
7.5' .031" captube
Max capacity 360w
Square black vented case on wheels


So first, it was a little exercise on captube tuning. I started with 9' and worked my way down to this length. 9' was just too restrictive, so I tried 8' and it was close, but far too slow to cool down, so I ended up at 7.5'.

Edit - oh, I changed the plug wire just a bit, didn't get a pic of that, when I put the switch-plug on. It exits down instead of to the side now,and I found a real clamp for the wire. That was 'ok' but ugly, and the new one is better and safer. That was just a strap I insulated with rubber. The new one is a proper clamp, and I lined the hole with the same edge beading I used on the suction line exit hole.

The condensor and fan are a match, but the shroud wasn't sealed, so I just cut up some sheet metal to seal it up. Works fine, though not pretty like it could be. It's in a case so I was more worried about it being solid than pretty inside. I also added that flap on top to block the air. The fan blows upwards, and the case was a few inches higher than the fan/condensor, so that directs the air out the one side vent




Just used the 'plug switch' setup like the other SS bencher. Works pretty good.
That's a bit longder that I'd been using, but it seems to work well on this one. The static pressure is around 110psi, which isn't terribly high but a little higher, normally around 80-90psi.




The results are good though, it's been great for R290, from 0w right up to 360w. Pretty linear temperature curve, and no drastic drop from 0w to any of the load ranges, though 0 to 50w was about 10 degrees. Funny how it pulls so cold unloaded. Unusual for high load tuning, and I think the long captube does that. Makes it so you can 'show off' but at the same time, still have the high load unit.

These are all out of order, sorry


250w

360w

100w

0w

175w

50w

I tried to get the full 400w max of the new tester but it wouldn't hold, so the max is around 360w or so, which should hold even the 6 core cpu's at full load
I suspect that at the slightly higher head pressure, the refrigerant coming out of the captube is 'jetting' more so, maybe creating a finer mist. Since it's normal to see a mixture of liquid and vapor even right at the end of the captube (according to refrigerant theory) then the behaviour of the exiting refrigerant may change with the pressure behind it being slightly higher or lower. If it's fully liquid, then the amount of pressure should have less effect, but more and more I think it's still in transition, at least partially.

I woudn't say I knew that for sure, just guessing from the results I get, trying to extrapolate why it's working how it is.








So that's the project though, more or less ready to use

Only things I'd like to do is make up a set of plates that I know for sure is ok for i5/i7 use, and possibly get a display and switch in, though it's not needed. Not everyone has a K temp meter though, so a basic display is nice too. I have one here that's busted I was going to use for PT100, and one that's PTC1000 that I have no sensor for but I might grab one for it. Just depends who ends up with it and if they care about that.

The red line with the braided stainless K probe wire kinda made it look 'festive' in a way, so the 'Christmas Box' idea came to mind with the season and all that

I do like the way that part worked though, The wire is just a wrap, so can be pulled, but I made up a red extension wire for the sensor too, can be seen there, so the wire could stay if a person liked it.

And in keeping with the season, one last parting shot to say happy holidays to everyone here Merry Xmas guys.




Gray

[runmc]

Awesome job Gray The cooler looks very nice. It will make some one a nice Christmas present.

[Gray Mole]

Thanks Ron

Means a lot from you. I know my brazing will never be quite as pretty, but it worked out nice. Thankfully I don't need to make the 'guts' quite as pretty as you do with the case on it

Merry ho ho!

[sdumper]

Its all about the performance baby! Nice work!

[Fhqwhgads6680]

Hey gray, nice unit man, awesome to see you back on here building again!

I've been away from phase for quite a while now... is 7.5' of .031 cap tube about what people are using to cope with the beefy processors of today? Those temps with 360W and R290 seem very impressive to me!

[Gray Mole]

From what I've seen, it's a bit longer. Ron used larger (42) captube at 8', and I've seen other guys using similar lengths of 40 and 42 for high load, but I had a feeling that the 31 would still do it, just needed to play a bit with it.

I DO have a higher static pressure at 110psi. I don't mind though, it's not astronomically high, just higher than the 80 to 90psi I usually see.

But the way it's working makes me thing I've got the combo of max length vs max. charge for the length of tube and gas used. Different compressor or gas may want a slight change, but it's what I was looking for in terms of maximum use of the gas and tuning if that makes any sense.

But I usually see 6 - 6.5 feet of 31 used for high (300w) tuning.

Gray

[Fhqwhgads6680]

ahh dang.. It has been awhile... last I was around .031 was large... HOLY CRAP!

I will be sending my last evap to Ron to be Brazed up to a Flex Line.. I may go with the larger cap tube then. Thanks Gray for the info, and sharing a nice xmas themed unit!

[Gray Mole]

I'd be inclined to go with the recommendations Ron gives you

I had to play a fair bit to get this working nice, and I think having a relatively oversized condensor helps a lot with it, since the pressure increase on the high side from having to push so hard to get it through may actually need the extra capacity there to make it shine.

I'm still condensing at ambient, but when I measure the temp through the coils, it's fairly warm, right close to the end of the condensor at 360w, so it all seems to match up.

Just depends how much playing around you want to do, but this was a lot more involved than when I used the shorter captube length.

I was able to pull over 500w out of 6.5' of 31 captube though, which is why I wanted to try this and I'm glad I did

And your welcome, sir. Merry Xmas

[sdumper]

Personally I prefer .040 at 10.6 feet but everybody has variations.

[Gray Mole]

Oh yeah, and the size of the tube (as long as it's big enough to stay over 5' but under 15' I've found for stable, consistent results) isn't really that critical as far as that goes. 31, 40, 42, they all work well.

Just that after some playing I'd found that the condensor sizing was tricking me into thinking that the captube needed to be a lot shorter to work. That's why I wanted to do this one, mainly to see how long I could keep the captube.

I'm thinking about 40 next time, and do the same kind of testing, maybe start with 14' of 40, and chop down slowly until I find what will just barely hold with the other factors I've got, and see if it get's a marginally better result.

Bit obsessive with some things. Might have noticed that. I think I'd trade that for some social skills if I could though.

I get the impression that if you go with the absolute longest captube you can for the max load you want, that you end up 'stretching' the gas just a bit more throughout the load range. It's not a huge difference but this one netted probably 2-3c colder load results (for R290) that any other of the 3/4hp benchers I'd made. The fact that it goes to like -60c unloaded is pointless, but it kind of underpoints that theory. If you're using a shorter captube than you need, and not tuning it to it's full potential, it seems like you may be losing a small degree of efficiency. Just a theory though, but it seems like a good enough reason to waste a bunch of hours playing with captube, when it'd save a day or 2 just to go with what works

I'm trying to find variations of whichever captubes that work to an absolute max of around 350w, since that's what I think the 6 core cpu's will max out at. I don't know for sure, would need to see one pushing this or another one hard to be sure, but I think so. So I'm seeing if this will work to get the best I can out of it, though really for the sake of a degree or 2, it's not what I would consider 'worth it' in the grand scheme of things. Just something to occupy my head and lately I'm finding I need that a lot more.


Gray

[Splave]

really great finish work

[Fhqwhgads6680]

Oh yeah, and the size of the tube (as long as it's big enough to stay over 5' but under 15' I've found for stable, consistent results) isn't really that critical as far as that goes. 31, 40, 42, they all work well.

Just that after some playing I'd found that the condensor sizing was tricking me into thinking that the captube needed to be a lot shorter to work. That's why I wanted to do this one, mainly to see how long I could keep the captube.

I'm thinking about 40 next time, and do the same kind of testing, maybe start with 14' of 40, and chop down slowly until I find what will just barely hold with the other factors I've got, and see if it get's a marginally better result.

Bit obsessive with some things. Might have noticed that. I think I'd trade that for some social skills if I could though.

I get the impression that if you go with the absolute longest captube you can for the max load you want, that you end up 'stretching' the gas just a bit more throughout the load range. It's not a huge difference but this one netted probably 2-3c colder load results (for R290) that any other of the 3/4hp benchers I'd made. The fact that it goes to like -60c unloaded is pointless, but it kind of underpoints that theory. If you're using a shorter captube than you need, and not tuning it to it's full potential, it seems like you may be losing a small degree of efficiency. Just a theory though, but it seems like a good enough reason to waste a bunch of hours playing with captube, when it'd save a day or 2 just to go with what works

I'm trying to find variations of whichever captubes that work to an absolute max of around 350w, since that's what I think the 6 core cpu's will max out at. I don't know for sure, would need to see one pushing this or another one hard to be sure, but I think so. So I'm seeing if this will work to get the best I can out of it, though really for the sake of a degree or 2, it's not what I would consider 'worth it' in the grand scheme of things. Just something to occupy my head and lately I'm finding I need that a lot more.


Gray

Hey Gray, This helps alot! I need to do some reading to get but up speed its been so long... I Didn't meant to hijack your thread and turn this into a captube discussion BTW

[Gray Mole]

To be honest I like it that you did

I'm more interested in the science behind it than the unit really, though the Xmas theme made that fun. Captube tuning's been such a 'hokey' principle, with no definity rules there at all. Just some guidelines and recommendations.

I'm working to see where the limits are for myself, but I wouldn't really recommend a length outside of what people know work because it's a huge commitment to getting it there there way I've done it here.

I've used the idea that 100-120psi static is my limit for pressure with the thing off. No higher or the captube just has to be shorter.

Then, if the working high pressure is too far out of spec, the captube is just a bit too long, since there's no point in pushing so hard that the compressor will die sooner.

The condensor, if giving you ambient, but warm throughout the whole condensor, or at least very close to ambient at the filter, is a good indication that you are at the limit of length for the captube size you're using.

That's a fair bit of work in real terms for ensuring it's safe and ok to use, when trying to extend the captube to it's longest possible variation. Especially considering that it's only giving a couple of degrees of improvement. I've done that to 350w as well as the high limit of load, so if a person wanted more or less load as a max, they'd have to rework that whole thing accordingly.

But if all of those factors are satisfied, then the longest of possible captube (of whichever size and for whichever load) seems to be the most efficient (in cooling terms, not power) way to tune it.

At least, that's what I'm finding. For the sake of a very marginal improvement, you do have a couple of consequenses.

Like, being that very slight high side increase, there could be a slight reduction in the life of the compressor, though this one sits at around 70c compressor body temp at load so I don't see that being much of a risk, it's always there.

Also, the cooldown is longer initially, seems like 3min 'first time' or after 2 days to a week of non use, instead of some that cool down fast no matter how long, and still more like 1-2 min. after a couple hours to get to full cold.

So there are tradeoffs for it, but I like it and the way it performs.

I'm going to keep going with 350w as the max I'll tune for now, since it does seem like the max that should be for 6core load, but I hope I can see it on a phase unit to confirm that. Mine or anyone's, tuned for it, and what temps it gives or how the load temps vary with the amount of cores used etc to see if that holds true.

There, now I hijacked my own thread


Gray

[Fhqwhgads6680]

Lol, since you've gone and done the hijacking yourself Gray, Would you apply this form of tuning to any build? I remember (whoa look at me going and remembering stuff! lol) that r290 is propane, and a very lightweight cool running gas. Which I would guess means you can have the cap tube a bit longer as high side temps and pressure would still be resonable. What if you were using R22 or R507? I am guessing Captube length would need to be a bit short, albeit not by much.

Also if this were more of a 24/7 build, I.E. Smaller Condenser, and Recip compress in the ballpark of 1/2 HP. I know Rotary's are built to handle higher temps. Would this play into your limits?

Just curious on your captube sizing methodology as judging from this unit you're very good at it!

Thanks as always gray!

[Gray Mole]

Yeah, any variable you change will affect the captube's potential limit, whether the compressor being smaller or larger, the condensor being smaller especially, or really anything.

Different gasses do take different lengths, and while you could just setting on a length that works for all gasses more or less, this kind of fiddling will find it's best length.

I find that the stronger the compressor, the longer (if only slightly) you can get the captube and still hold the load.

I find that the softer the gas (weight) the shorter you end up needing the tube, if only by a small amount.

I find most importantly, the better the condensor, the longer you can get away with on captube, following the guideline of ambient (or close) temp at the filter after the condensor. I mean, you can have a higher condensing temp, and in a way it can be beneficial to operation, like faster cooldown and a some other things that a higher condensing temp (within reason) can give. Just that when you're pushing the limit of captube lengths, you've got to consider these (and other) aspects of what's going to give you grief if you let any one of them go too far out of spec.

Curiousity is what keeps driving me to play well beyond any kind of necessity as far as captube tuning goes but finding a kind of 'meta-rule' for this kind of tuning just means I don't end up relying on a single length for all purposes, but have a way to determine what will give a better result for a variety of applications. Just means a bit of extra work every time, but I don't mind that much.

I wouldn't use this method on a cascade though, especially on the second or subsequent stages if captube is used. That slight increase on the high side pressure is easy to account and compensate for in a single stage, but any pressure spike or even marginal increase has often drastic and detrimental effects on the whole of the operation. I think that going shorter and just tuning it is better for that. Having a 500w max captube length, and tuning it to 350w, gives a more general good result, and allows for a lot lower head pressure. The base/first stage's temps will be more consistent that way from idle to load, which should make the tuning overall a lot easier, and the HX temp more consistently low.

The mole weight on propane is around 45 from memory, not light like ethylene but still very light compared to the 100ish of the other SS gasses. In actual use it doesn't play as much a part in what temps you get, but it does play a huge part in defining the best captube tuning, since it's going to require a different enough length to be significant. Using the method I am is netting me load temps very similar to what I would see with 507 and the benefits of the gas being lighter and somewhat easier to use make R290 a lot more tempting to use in systems. If I can get more or less the same temp with a gas I know can give a lighter (overall) head pressure, and be able to carry any gas (and better than any other gas for the most part) then I like and will probably continue to use it.

I guess being completely obsessed with it kind of helps a little too :p


Gray