1/2+ hp Single Stage Build

[aenigma]

A larger efficient condenser can increase efficiency when the expansion device is set correctly for it, assuming the smaller condenser was increasing high side and getting warm. Less strain on the compressor from lower high side pressure is the reason for it. As for evaporator surface area I don't know what evap he is using, I used to just use a pipe cap and piece of pipe to make a hollow flooded evap that worked great, then I drilled pitting holes in a thick bottom plate with a drill press and both worked well. Finally I used a huge stepper evap I bought with a lot of surface area and it kind of sucked since there was too much mass and not enough surface area and volume. You can get a lot of surface area in a small evap, not a big issue really.

I know what an accumulator and receiver is, I just don't think he had an accumulator since he doesn't need it. He used to have a receiver but I told him that was not needed for capillary tube and he got rid of it and it increased the performance. Neither should be in a small single stage unit like this. I think a language barrier might be a problem here.

I haven't tried to cool a cpu in a awhile since it doesn't seem to be worth it anymore and I am not into extreme overclocking like I used to be. I am more into refrigeration and making autocascades occasionally, that's why I hang around in this forum sometimes. I use my autocascade for testing and cooling a beer super fast haha. An autocascade can cool a beer nicely.

[Obijuan83]

My Unit is built that way, I din't use the copper 'Dessuperheater coil' , I kept the Sight Glass for decorative purpose and I used a 1'' diameter 20cm copper tube as accumulator before compressor.

The condenser unit at summer ~ 35C-40C temps was good for the Single Stage but at the momment with 20C room temperature it is big enouh so either I had to stop the fan or to add more R404a. I finally add some R404a and added a Fan regulator.

I think that the main benefit from a large condenser is that you can adjust the load by regulating fan speed and not by adding/removing refridgerant or adjust the cap tube lenght. If you have a fast spinning fan at a given load and want to increase load capacity if you slow down the fan the high side pressure increases so more refridgerant moves towards the evaporator. Correct me if I understand it wrong.Of course all this adjustment inside limits.

As regards my compressor from the site of Embraco , NEK2150GK has a cooling capacity of 303W at -40C at 25C ambient temp so maybe its real.

[nachtfalke]

well , we speak 2 different language ( sorry for my english) ... but the "refrigeration language" it is or should be the same from US to EU and Asia and all over world , every condensing unit is calculated from the evaporator to compresor and condenser, this is one of the main issue with those " oc condensing " builders today ... they start first with condensor or compressor..... my questions is what kind of evaporator he using? for what load target is used? how much load can that evaporator holding?? how much refrigerant flow has/ restriction.. , how much cooling capacity and so on.... nobody named.

here is used an manufacture condensing units ...modified intro OC condensing unit. all are OK , but from my experience when you using a to large condensor , you will enter deep in the " condensing temperature" issue to low ...what mean this??? you will lose cooling capacity!... hight side pressure have nothing to me mentioned here , only if condenser is calculated wrong or to small / XV wrong tunned etc.....but if he is dimensioned /calculated corectly here will be no issue with the hight sided pressure ..., but if the condensor is to large then you will be happy with your lower hight sided pressure ...when you will loosing cooling capacity on the evaporator ..
COP will be incresed only by XV ....have nothing to do with condensor!
in my life i builded "some" SS , chillers and cascades , besided this i was an power hw points in the hwbot forums ...today i?m inactive on the hwbot. , my units was optimized well for thoise kind of loads .

"As regards my compressor from the site of Embraco , NEK2150GK has a cooling capacity of 303W at -40C at 25C ambient temp so maybe its real."
my dear it is not real because theyr test is to much different from yours ....what subcooling you have ? but superheat? ...what XV you use ? ...what cooling capacity has your evaporator ?
those number from above regarding Aspera test is done with a large evaporator for freezer vitrine ...in your case those number will be no more applied ...a 12ccm comprerssor will do -25 to -30 , and this is your gas return temp ( outleet evaporator ) and at those temp add a +60-70 degree Delta T ...and your cpu cores will be always at positive temp .
i build some time ago a large SS unit with rotary compressor 42cc , evaporator was hand made be my ( 1,5kg masively cooper) with 2 expansions valve , and 3,3kw condenser ...and i was able to hold a 7920x 12 core cpu in heavy load at +10 degree ( internal cores temp) in R15 benchmarkt with 1.55v core.
the main problems of the OC condensing units it?s that actually evaporators surface with the heat is to small ... we can?t do nothing here . yust some add of the evaporators design with will increase Delta T with fews lower degree.
regards all!

[aenigma]

His measurement is only evap temp not CPU temp so you're right, the cpu is probably much higher depending on the evap design. Most CPU evaps you buy from guys that machine them have a lot of mass but not much volume and surface area so they don't seem very efficient. Best ones I used were simple ones I made that were just hollow with a lot of surface area on the base plate, but you have to lay the computer down so the refrigerant pools in the evap, I just had the motherboard laying on a table.

Keep in mind this was quite some time ago and I don't think the CPU's back then got as hot even with a 100% overclock. I had one SS made from a 5000 BTU window a/c that held -40c actual CPU temps under load easily and that was a big window a/c condenser. But as you say, it all depends on cap tube length/width or TXV size and adjustment, but that is regardless of the condenser. In the case of capillary tube the charge is critical and charging with superheat and subcooling (mostly superheat) is a good way to get close then fine tune from there. I have had no problems with oversized condensers, you just have to know how to charge it correctly and use the correct sized expansion device which can take a lot of trial and error.

[Obijuan83]

Evaporator is from Asetec Vapochill LS

I used 9.5 feet of 0.31" cap tube from the cap tube size rule for 200-210W load

The fact is (and I beleieve you can't understand) that if you have a large condenser you can adjust the cooling capacity by adjusting the condensing temperature.Also it more easier to have cooling capacity regardless of summer or winter when in an fully optimized SS with a specified condenser for the load in summer you will suffer from high ambient temps.

[aenigma]

Yeah precisely, fan controllers are essential for big ambient temperature swings. That is why most RTU's (roof top units) have fan controllers and/or head master valves for winter/summer operation. Either way he is right about the evap most likely not being very efficient and your CPU being much warmer than the evap itself. Your evap is very cold, but the CPU is probably quite a bit warmer. Just the nature of the beast with such a small surface area on the CPU cores. For one thing you're dealing with the heatplate/heat spreader on the cpu to have a good thermal compound for low temperatures between the CPU die and the heat spreader but also how small the cores are nowadays. It truly is an uphill battle.

[Obijuan83]

I disagree a bit about evap.

For example the following cpus with the same heatspread surface

A Core 2 Duo E8600 @ 5400 @ 1.8v the evap temps are @ -35C to -40C @ Load (~210Watt)

A Pentium D 820 @ 4500 @1.675 the evap suffers with temps to -15C to - 17C (~260Watt)

This fact tells me that the SS with this evap can handle 200 to 225Watt ( the cap tube is set for 215Watt) but not the 260Watt from the Pentium D series.Perhaps with a modified captube that load could be handled from the SS.

The main benefit with the large Condenser is that you can charge it to hold for example -40C at load @ 215W @ standard fan speed but if you want -45 to - 47C at 125-150W you can have high fan speed , so lower high side pressure which leads to lower low side pressure with leads to lower evap temps.

In simple words you can adjust evap temps at load for different heatloads by adjustin condensing temperature.

If you want -47C for a Spi1M (lower heatload @ 125W for example) you can have high fan speed and if you want -40C for a Wprime1024 you can lower the fan speed to reach the optimized balance of R404a charge - Captube lengh heatload (215W)

[aenigma]

Wait, how do you know your exact CPU temperature? Evap temp isn't meaningful, actual CPU core temps are.

[nachtfalke]

the fact is , that you wont calculated your unit (and that you can t understand) and you wont charge it with load tester

a fan controller will be mandatory only when you have very low ambiental temperature( specially winter time with external unit mount outside) so in your case , using it on the desk at normal ambient temp you will not add to much at evaporator performance , in fact you will modified superheat ...but stay did you calculate your superheat and subcooling? the fact is that you don?t understand this thing!
i look over your subs on hwbot ...that tell me your SS can?t do much over a good water loop!

[Obijuan83]

i look over your subs on hwbot ...that tell me your SS can?t do much over a good water loop!

Are you serious? You are telling me that my SS can't do better than a good water loop?

Sorry I thought that I have a conversation with adults...

I cannot understand you attack...what is your problem?I thought that forum are for discussion and to became better...

You came here and judge my effort with the worst way...Maybe the problem is that a new builder came to surface?

Tell me that you can achieve that with custom watercool...

[Charles Wirth]

A few things,

CaptainCascade dropped in and didn't say anything!
Capacity issues, reduce cap tube length. You lose in temp and you gain in capacity. Do not pull into vacuum.
I have an industrial water chiller with EK block, it cannot hang with a proper SS.
Lastly, do what you enjoy and what is in your reach, nothing wrong with experimentation within your means. You have lots of experts here to tell you how to do it by the book but go at your own pace and learn from improvements on your setup. Keep asking questions, that's how I got all sorts of stuff built from my dumb ideas.

[Obijuan83]

A few things,

CaptainCascade dropped in and didn't say anything!
Capacity issues, reduce cap tube length. You lose in temp and you gain in capacity. Do not pull into vacuum.
I have an industrial water chiller with EK block, it cannot hang with a proper SS.
Lastly, do what you enjoy and what is in your reach, nothing wrong with experimentation within your means. You have lots of experts here to tell you how to do it by the book but go at your own pace and learn from improvements on your setup. Keep asking questions, that's how I got all sorts of stuff built from my dumb ideas.

Thanks that is the purpose of the forum to exchange experience and ideas