What performance to expect of a PA160.1?
 

Hi,

I'm looking for a low noise watercooling solution for my Coolermaster ATC-110C miditower and I thought about the recently released Thermochill 160.1 wich will fit the lower front with some modifications (removal of 3.5" bays).

I just don't know what to expect of the watercooler. In my experience watercooling performance is more dependent of the amount of air thats blown through the radiator than the surface of it.
Therefor I'm skeptical since I won't tolerate louder fans than a Papst 120mm 4412F/2GL wich is rated 1600RPM/55CFM (26db). Now 55CFM is'nt much - a double or triple radiator is to prefer simply because of the doubled og tripled amount of air that's moved at same low noise, but a radiator of that size will be hard to fit inside a miditower without major and insane modifications.

So any ideas of the performance of a Thermochill PA160.1 at >55CFM using the original shorud? I'm planning to pair it up with a Laing DDC-1T pump.

The hardware to be cooled:
- Intel Pentium 4 630 (@4,2GHz / 1,4v)
- Ati Radeon X800XL - core: 450MHz
- Maybe a overclocked 955X chipset
- In all - above 200 watts of load?

If a single 120mm fan at low noise won't be suitable for cooling my hardware then would it be a good or bad idea to make a custom fanshroud and purchase a low noise 140-160mm fan instead?

Regards
Salkcin

Sorry of the rather long thread :p

I can't say what the exact performance is going to be, especially when you haven't said what blocks you are going to use. I couldn't even say if you told me what blocks (other variables intrude). However, I can tell you that a decent block and good routing with the above will probably work at stock CPU speeds with a single 120mm fan.

I don't overclock Intel CPUs, so I'm going to stay mum on that too. Sorry! It should work, though.

All that's to say that if you want QUIET, you should stick with 120mm fans, not 140mm or 192mm fans. The quietest fans are 120mm (as far as I have seen). Nexus 120mm fans are nice and quiet and work well with the PA160.1 line. You won't get minimum load temps with a single fan, but it probably won't affect overclocking that much if you design your loop and airflow correctly. Again, others can speak to overclocking P4's better than I.

Test results from Marci on a pelt setup shown 1 fan vs two fans on pa160 doesn't improve the performance that much. Simply, the pa160 is focused on silence and decent temps. You get a single panaflo at 7V on pa160~ with very good temps but not the best.

Hope this article helps

http://forums.procooling.com/vbb/sho...hlight=pa120.2

Thanks, ricecrispi. I was certainly not clear in my thinking or typing when I wrote that. I was trying to say that the 1 fan PA160.1 will have a higher C/W than a 2-fan PA120.2 setup. If you have a larger rad with more CFM (for the same design), you might be able to overclock more. That's all I was trying to get at.

For a single PA160.1, one fan is good. Again, thanks ricecrispi.

PA160.1 was designed for a single fan and designed for noise vs cooling.
The design thread here is a good read: basically it starts with a fan that is quiet (ie 7v) and then tries to make the best radiator for it that will fit inside "most" cases.

Sounds precisely what you are after.

A PA160.1 will cope with the 120W to 200W your rig is likely to put out. In fact, with a 200W load, the difference between a HE120.1 and HE120.2 is less than five degrees. Now five degees may get you tens of MHz more, or it might give you jack. The 2nd fan will definately add more noise, and the bigger rad will be harder to fit.

PA160.1 is a an excellant compromise between the two.

I'm sorry to hear. Stock Pentium 4's are deadly slow, but all of the CPU's in the new 600 series overclock easily at stock voltage above 4GHz wich suddenly gives a overclocked Athlon64 competition while being able to multi-thread :cool:

I'm actually going to buy a whole new system since the anti-corrision fluid I had in my old wasn't as good as the company claimed :mad:.
My considerations for the new system are Thermochill PA160.1 + 50CFM 120mm, Laing DDC-1T pump, DangerDen Maze 4 Acemetal for GPU and for CPU I'm proberly going for DangerDen TDX since Swiftech MCW6000 is unavailable and Storm is to expensive..

My current system (wich is mounted externally) cools my overclocked Pentium 4 to 49 degrees at full load. It's:
- Black Ice Xtreme II + 2x Papst 120mm @ 55CFM each
- Watercool Heatkiller rev2.0 (CPU)
- Innovatek InnovaGraph 2.0 (GPU)
- Eheim 1046 pump (300l/h)
- Big cylinder Innovatek reservoir

I know it's hard to predict the excact performance of a new system, but some experinces of people having a PA160 mounted in the case would be nice. Then it would easier to compare and determine whether it would cool good enough or not.

The storm is on firesale for $50. THe apogee, good or bad test results will perform on par with the WW or better and have better flow rates and both cost same as TDX

BTW. I remember a few people complaining about the accel plates being moved in the TDX

Hardware is very expensive in Denmark where I live compared to other countrys. The price of Swiftech Storm are 98$ in Denmark (danish price recalculated into US Dollar).
If I buy hardware out of danish borders I have to do in EU countrys (Sweden, UK, Germany...) to avoid expensive extra taxes.


But any feedback on my considerations for the new watercoolingsystem when it comes to choice of waterblocks + pump?

- CPU block: Swiftech MCW6000K-775
- GPU block: DangerDen Maze 4 AceMetal
- Pump: Laing DDC-1T
- Tubing: 10mm (3/8")

And how about the reservoir? would a reservoir like the one seen below (Watercool AB-G) limit the flow so it affects the performance noticably?

http://212.227.231.249/watercool/cms...pics/30310.jpghttp://212.227.231.249/watercool/cms...0310%2Dalt.jpg

DENMARK OW!

How much are the nexxos xp or similiar aquacomputer blocks? THey seem to perform pretty good

Instead of getting that res, Buy a ddc plexi tops that get rid of the 90 bend and improves flow rates.

Alphacool NexXxos are amongst the cheapest components in Denmark, NexXxos XP is priced 43,99$ with a plexi top and 38,6$ for the light weight version, but they have waterblocks priced as low as 31$ (NexXxos HP Pro), but they have 1/8" threads. A Aqua Computer Cuplex Pro is priced 60$.

But I going to try one of the american brands. The eurupean brands like Alphacool, Aquacomputer, Asetek, Innovatek and so on uses a small hose (6-8mm) and produces waterblocks that doesn't depend on flowrates and pressure... I had a couple of those systems and will now try the american way - big hoses (10-12mm) and powerful pumps.

I don't need money, but with the price of Swiftech Storm it's common sense to look for other waterblocks. Recommendations are welcome, DangerDen TDX and Swiftech MCW6000 (Socket 775) are what I'm considering.

This probably doesn't apply to the 160, but with thicker heater cores I've discovered that there's a better noise/cooling tradeoff with a pair of stacked 120mm fans (very undervolted) than with a single 120mm fan, even a bit undervolted. Easy enough to try if you find you need to run, say, a Nexxos at 12V...
Also pay attention to removing airflow restrictions inside your case (better to not have a restriction than to have to overcome it with fan power).

hmm - these guys have a pretty good rep, will ship to you via USPS (which is very cheap) and sell the storm @ $59.95
http://www.sidewindercomputers.com./waterblocks.html
even if you do get hit for Danish VAT (25% iirc?) that'll still probably be cheaper than getting it in Denmark

I just thought of something - wich is better? Thermochill PA160.1 or D-tek Procore (Chevy '86 heatercore)
Both using a Panaflo L1A or undervoltet M1A 120mm fan.

- Swiftech writes in their manual about the MCW6000 that is has a problem with the socketlever of socket 775 and is therefor placed assymmetric. I'm now got the impression that MCW6000 is not optimal for Socket 775 since the inlet is far placed from the core.
- I read about the Swiftech Storm the small jet channels could get stopped by anti-corrision coolant and that makes me think one extra time if it's the right waterblock for me.

If MCW6000 and Storm is excluded then what waterblock is to prefer for Socket 775? maybe the Apogee?

I just switched from the D-tek pro 120 HC to the PA 160. It is better my temps are 4-5 above ambient at idle and 10-12 at load. It is witout out a doubt much more quiet since the PA160 does with one Nexuss silent fan what the D-tek pro did with two fans that had to be turned up to get the same temps as I get with the PA 160. This is all on a DFI NF4 with a 4800X2 @2500 and a 7800 GTX OC with the Storm block and a Maze 4 on the 7800 with an MCP350 pump and 7/16 masterclear tubing.

Re: What performance to expect of a PA160.1?
 

My modelling would suggest the Chevy'86, by a short-head.

Edit: Inserted link to Chevy'86 to which I am referring

Re: What performance to expect of a PA160.1?
 

very confusing :D

Re: What performance to expect of a PA160.1?
 

I'm talking about the original Procore wich is 185x150x50mm (like the Chevy '86 heatercore) where the "Procore 120" model is 162x140x50mm and a D-tek specific model.

Okay, that sounds good since the Chevy86/D-tek Procore is easier to fit in my miditower. The surface (the space of fins) is actually theorectical bigger on the Procore since its 150x150x50mm where as I understood it the Thermochill PA160 core thickness is only 25mm = ~165x173x25mm (can't remember excact dimenssions).

EDIT: Could you explain what a value of "30" in "Specific Dissipation (w/c)" means? - that for every 30 watt dissipated the temperature rises 1 celcius above ambient?

Re: What performance to expect of a PA160.1?
 

Yes. It refers specifically to inlet temperatures of Air and Coolant.
It is the reciprocal of Bill's C/W
The model is only for the AirSide of radiator.
Actual values would be expected to be less: depending on coolant flow-rate.
It is only a Model and unvalidated.

Re: What performance to expect of a PA160.1?
 

Looks like you're assuming identical airflow through both. I'd think the PA160 would have a bit less restriction, which should make it even closer.

Re: What performance to expect of a PA160.1?
 

No, I am not assuming identical air flow.
Estimation of airflow is an essential part of the modelling

Re: What performance to expect of a PA160.1?
 

Sorry, Les. That vertical line fooled me. BTW, how many CFM did you calculate for each rad?

Re: What performance to expect of a PA160.1?
 

Only have in m^3/min.

Edit: Part of Legend obscured - fixed

Re: What performance to expect of a PA160.1?
 

Thanks, Les. That's more than good enough; I can multiply by 35 if needed. I'm impressed by how close the airflow is between the two. The HC's wider fin spacing really helps, I suppose.

Re: What performance to expect of a PA160.1?
 

I ask again - what waterblock would be optimal for my CPU using a DangerDen Maze4 Acemetal GPU block, a Laing DDC pump and dual pass "parellel" watercooler?

Does Storm require to high flowrates to be in a multi-block circuit and how high is the risk of the jet-channels getting clogged by coolant-fluid that prevents corrision?

The Apogee is'nt available in Europe yet and I don't know if I should wait for it.