First build single stage (many images)

[makaka]

let's start with a bit of pre history , i got my first Vapochill ls (from ebay ) back in 2005 year or so , it was badly broken after shipping , i took it to repair shop , they fixed a broken pipe and recharged with R404 . After getting home the unit worked but temp were poor and compressor would shut down randomly . then completely refused to start , took it back to repair , they said that the control unit was bad , i ordered a new one , after about a month i got it plugged to the system , and again same issue as before compressor shut down randomly and heating up a lot . i took it back to repair , this time they refused to accept it for repair . since i was a student at the time , i was almost out of cash trying to repair the unit , so i forgot about .
Now 15 y later i found the unit in my garage , so i said what the hell why not to fix it , i spent about 2 week in forums reading about how all this work , went ahead bought all necessary tool and started the repair .

I figured out that the unit still had a leak on it , this caused the compressor to overheat . This little bad boy was danfoss nf5.5clx i was able to start the unit and get it working normally but this S*** was very noisy . so i decided to swap it with a NL11F ( Thanks to Charles Wirth for double checking with me the model) .



The compressor was used it cost about 25$


since i am planing to get load in the 200-300w range , i updated the condenser


about 20$ , according to spec 15c delta T @ 700w

i noticed that the old accumulator has some debris inside that wan't be removed , so i decided to make a new one using an old filter drier .




Also installed 2 king valve to later adjust capillary tube length easily (Thanks to aenigma for idea)


Used remaining copper tube to make a small coil on the compressor outlet . Was really hard doing this by hand , require some huge grip strength .


finally my new SS tube arrived


Silver brazing to cooper

Now putting everything inside the small case





Now finally i am allowed inside the house (wife allowed ) to do remaining work . It was really cold out in the garage there was no heating :down:.

capilary tube install and deep vaccum , i used 3.2m of 1 mm .(so far maybe will tweak later .)





i also used a hair dryer over the compressor and tubing to boil any remaining moisture when pulling vacuum .
system did hold the vacuum for 24h now we can start charging the R507 .



first frost

insulation ,



charging until 6c super heat .

and finally :





now finishing insulation :



testing with p4 945 @ 5.2 ghz 1.6v , according to asus monitoring it is 220w (according to estimation from power meter 260w , and finally according to OuterVision? Power Supply Calculator 254w ) . i have a D820 i will test it later since it should be hotter .




342w is the compressor power usage

high side pressure under load

lowside pressure under load


All those under load (200-250w):
decharge temp : 61c
condensing temp : 36c
HP : 260 psi
ambient temp : 26c
low side pressure : 0.7bar
evaporator temp : -28c
temp 6" from compressor : -14c
compressor temp : 40c (delta fan doing great job )

It was really fun time building the unit , some time it was a pain when i had some micro leak and then cap tube blockage , but with patience and learning everything's is possible .

 

[ATMINSIDE]

Do you mind uploading the images to the forum instead of ibb? A lot of users have issues seeing images on that hosting site, myself included.

 

[makaka]

Do you mind uploading the images to the forum instead of ibb?
Your wishes are order .

Done

Now my first 6 ghz , it was first try . it can do more but i am more interested on testing the max load capacity of the unit .




Not sure how much power it is , this is a D820 @ 4.4 Ghz with 1.76v (really crappy overclocker) , it should be in the 240-280w range .




Cpu temp @ -13c after 20 min prime 95 load .





Evaporator temp -30c , section line temp (7.2c) sound like i almost reached the max cooling capacity with this capillary tube length .





High side pressure almost 250 psi .

Will get my hand on QX6700 and do some testing , this cpu should reach 300w easily with some overclock .

update : system is even 1h + prime stable same temp remains

 

[ATMINSIDE]

Nice build!!

 

[hokiealumnus]

That looks like a lot of fun. Great job rebuilding that one to be a very effective cooler!

 

[makaka]

Nice build!!

Thanks

That looks like a lot of fun. Great job rebuilding that one to be a very effective cooler!

Thanks

Tomorrow already the QX6700 will arrive . and i will get finally a good idea about the cooling capacity . Also this s* is getting addictive , maybe i will work on adding another compressor on parallel , or make a small 250w cascade .
i tested today few E8400 , some crappy celeron's and a E5800 , result was under -25c cpu loaded (evaporator -4x C ) .

 

[makaka]

Finally got the QX6700





well i can see why this processor is considered one of the hostest cpu , cpu temp already in the positive side but still less than 10c , delta T evaporator to cpu is huge , maybe i have to delid this processor later .
Evaporator temp rock solid so far -33c (6" from compressor temp already +8c), we should be in the 260w+ range (280-300w maybe).


I will post more image later after some more tests

 

[makaka]

Well after few day playing with unit , i ended up with a leak , it was my fault since i decided to not braze the flexible extension line , i like the look of it clean without solder


This was responsible for leak

Now since the wife is not home i can enjoy freedom of brazing in my room




Now i run out of Azote the there was no way to test pressure . since this section is completely isolated from the system due to the two valves i used air and my bike air pump to pressurize this part





and it hold 100 psi fine for about 4-5 hour




now fast cleaning the mess before the wife get home


Success mission accomplished

 

[ATMINSIDE]

IT LIVES!

 

[MaddMutt]

Thank YOU
I have been looking at a SS and have a question. You can tune a SS to a specific heat load/watts??? What about the newer processors = same heat load?? -> The AMD FX-9xxx was rated @ 220w~250w at 5GHz BUT IIRC would go well over 300w when OCed. Does the contact plate on the IHS need to be larger for the current processors??

 

[EarthDog]

Wow, great job on this!!!

 

[eatmc7]

I really had fun reaing your journey, thanks

 

[makaka]

Thank YOU
I have been looking at a SS and have a question. You can tune a SS to a specific heat load/watts??? What about the newer processors = same heat load?? -> The AMD FX-9xxx was rated @ 220w~250w at 5GHz BUT IIRC would go well over 300w when OCed. Does the contact plate on the IHS need to be larger for the current processors??

You can tune a SS to a specific heat load/watts???
Yes you can , with captube or cpev ,
What about the newer processors = same heat load??
Newer processor in general require less power . but if we talk 9900ks then with heavy overclock you can get in 300-350w range .

I can't talk about AMD TR , but for all other processor vapochill ls evaporator is quite large , it need to just cover the processor die .

 

[makaka]

Some retro s478 overclock

I had about dozen p4 to play with , most cap out in the 3 ghz range except few one .
Evaporator temp (load) -40c , processor temp -33c .

strongly enough this cpu can do super pi32 mb stable but adding 50 mhz or more and system freeze i think there is some limitation in the mobo northbridge



I have to find another s478 motherboard this one burned out


after full checking the was one dead mosfet in 12v line and 2 capacitor , i swapped them , the short cicuit went away , but now the motherboard start then stop instantly , so it require more trouble shooting or to the bin .

 

[makaka]

more update , now the 2600k.
this cpu was at first under water with 4.8 ghz and 1.47v and about 70c load temp (stable) , overtime it degraded to 4.6 ghz and 1.47v and about 70c load temp .


evaporator temp @ -34c

i did not notice a huge boost from subzero temp , at the same voltage 1.47v cpu was @ 4800 mhz stable but not more , but lower temperature allow more voltage so i boosted the voltage to 1.56v and was able to get stability at 5100 mhz .
the processor is poor overclocker moreover degraded and was not able to hit 5.1 ghz even when brand new , so i consider this a good result.

cpu temperature remain under 0 during stress test .

 

[MaddMutt]

Is there any thing I can do to make my Single Stage cooler better? (I'm not BASF )
I will have to test some more but it looks like it's best with a 2c/4t CPU. The temp probe on the SS unit has ~ -32c but CB-15 with 5c/10t @5.0GHz will hit ~ 10c. The coldest/hottest core is 5c - 14c and this is with the i7-8700k de-lidded.

Thank You for your advice





Here is a screen shot of the i7-8700k running CB-15 with all cores/threads. The temp probe from the head unit read -34c. I had a 2nd temp probe placed under the cold plate/on top of the CPU's PCB that read ~ 0.3c - 1.5c. As you can see in the picture, I have one core hitting 29c @ 5.0GHz and 35c @ 5.3GHz :-(
I have a EVGA z390 Dark that stops reading the CPU temp after it goes below 0c, so the CPU could be -10c or -100c and I would not know it

Is this damaging the SS Unit running/benching it like this?