Thermochill PA120.3 vs car heatercore

miniyazz

miniyazz

miniyazz

Soldato
Joined
25 Dec 2008
Posts
5,976
Location
Sheffield/Norwich
I know car heatercores used to be recommended above all else in the radiator/overclocking scene, but have things moved on? Does anyone know if the Thermochill PA120.3 or similar, when equipped with three decent-airflow 120mm fans, would outperform a generic heatercore (don't know much about the differences between them?) approximately 240mm x 120mm, with two decent-airflow 120mm fans?

Just considering which way to go if/when I go proper WC (instead of my H50).

Cheers
 
I think the issue is one of fin density, to get the better performance from the heatercore needs very powerful fans, whereas the thermochill runs well with relatively low pressure fans. Depends if you want low noise or performance I believe.

Could be wrong on this note though as I've never actually seen a heatercore.
 
Let's go with wanting both low noise and high performance :p

I'll probably plump with the thermochill then, if I can get one cheap enough for my liking. How do low pressure fans differ from low CFM fans? Are the two directly proportional as long as the fan size remains the same?

This one line on the description confused me slightly, as I know of no 120mm fans outputting 130CFM.
- Optimised for fans upto 130cfm
Does it mean 130cfm between the (three) fans that would be used to cool it, do you know?

Cheers
 
Ah, on more comfortable ground now. I don't know what you'll make of this analogy, but if airflow in cfm is considered equivalent to horsepower of an engine, the static air pressure is equivalent to the torque. A motorbike will go very, very fast, but not if it's dragging a caravan up a hill.

The stated airflow is how much air the fan can move against no back pressure. If you put a filter in the way it'll drop off a bit, if you put a wall in the way it'll drop to zero. A radiator is somewhere between these. Static air pressure is how hard it's forcing the air at the stated velocity. For a radiator, a fan with high static pressure will push more air through than one with a much lower static pressure but stated higher airflow.

This is why the 38mm fans are so popular for radiators. Static pressure also increases as the square of rpm, so a 1600rpm fan is likely to be four times better than an otherwise identical 800rpm fan. Makes it quite hard to optimise for low noise and high performance, you don't want a fan spinning very fast and moving no air, but neither do you want one spinning slowly and forcing too little air through the radiator.

I suspect the "optimised for fans up to x" to be marketing rubbish, but I'd be happy to be proved wrong.
 
Ah, would that also mean that if fan type A could produce a pressure of 5 bar, installing two type A fans on top of each other/in a push-pull thing with a radiator, would double the pressure, perhaps negating the need for a 38mm fan?


Edit: Also, OcUK seem to have no pressure data for fans! Ah well.
 
It does indeed. Fairly good results can be had from putting both fans on one side of the radiator I believe, though that feels clumsy to me. Shrouds are a really good idea, as fans leave an area of static air behind the hub, if the fan is spaced away from the radiator you can move this dead spot out of the radiator and improve performance. Fan + shroud is approximately equal to two fans in push pull according to martins liquid lab.

Very few fan manufacturers quote static pressure. I believe noctua do, and it's possible to find data on scythe s-flex (at least I found it somewhere when I last looked). In general I'd expect 38mm fans to quote it, as it'll be good, and 25mm fans to discuss other things, as it wont be so good.

If you really don't care about noise, you want counter rotating fans. Its two sets of blades in one chassis, one spins clockwise and the other anticlockwise. This is very, very effective and I believe louder than the average vacuum cleaner (may have forgotten that you said you want quiet when I wrote that)
 
Last edited:
Heh, I care about noise enough not to do that! Although I'd presumably not be able to do it with two identical fans anyway, no matter how I play around with voltages/positions.

I've got a shroud all planned - it'll basically be a (metal) mecanno frame bolting into the rad and the fans, suspending the fans 1-2" above the rad. Then I'll use masking tape or similar around the frame, making it airtight. I'll figure out whether to do push-pull or pull-pull at some stage.
 
Sadly not, the blades on the fans are twisted to optimse airflow, which means they run badly if run backwards. Pretty much limited to buying fans made for the purpose I think.

Mecanno is a brilliant idea. Maybe pvc tape instead of masking (don't know how well masking tape deals with being warm, or damp). Push pull should work better than pull pull, can't imagine it not doing so anyway. Do you have some sort of radbox in mind, or just a very large case?
 
Well even if the blades weren't twisted, swapping the voltage would mean the air would be blowing the other way, unless I turned it round, at which point the blades would both be going clockwise again!

I'll see what I've got in the cupboard in the way of PVC tape, that would probably be better, yes. Push pull will require slightly more meccano and space but will be doable.

I'm not really planning on a radbox at the moment, this is the one bit I haven't really considered. Obviously I may run into issues with things getting caught in the fans, so I at least need fan guards of some description, but tbh I'm probably just going to have it sitting on the windowsill behind my PC by itself - at 43cm long it's pretty big and finding something suitable would be something of a challenge.
Do you think it's worth making a proper frame to house it in? It's something I can do if it's going to be worthwhile, but enough of a hassle that I wouldn't do it otherwise!

I suppose if I made a frame I could then attach the pump (thinking an MCP655) in the frame, which would be one less thing to worry about.
 
Heh I'm crazily tempted by "120mm Delta Server Case Fan Focussed Flow FFB1212EHE 4000rpm Ball Bearing" :D "This is Not a quiet Fan - This Fan is specifically designed to MOVE AIR- FAST!" hehe
It's nearly as good as the Apache S-Flow for radiators apparently. But since it's more expensive and many, many times noisier and indeed more expensive, I think not. Not even for comedy value :D
 
The fan you want is the 38mm san ace. It's unbelievably, ludicrously good.

Can't see a box doing much for temperatures or for noise really, but it wins out on aesthetics. If you're on the ground floor and the computer is by a window you could trial my idea of radiator-submerged-in-waterbutt as an alternative to fans. Can't wait until I can try that, should be a cost of about £25 for the butt and a tenner in 1/2" hosepipe to submerge in the water, with only frost and locking the window as problems.

Is this your first trip into watercooling? Tricky to tell from this thread, but I'm guessing not
 
As a matter of fact it is :)

I've been experimenting a bit the last couple of weeks with a friend's custom loop and doing some research - he's definitely on a budget. He's got an old 45l jerry can for a reservoir/radiator, a central heating pump for a tenner, and some generic garden hose (which kinks, enjoyably). Not sure what waterblock. What's funny though is the level of gunk in the jerry can. It'd been thoroughly washed out repeatedly, but we opened it up at the weekend to properly look at temperatures and :eek:
Huge amounts of precipitated 'red lead' or whatever it is they use to keep the cans from rusting. Slightly more worrying was that some of it'd gone through the sock he was using as a filter and was coming through the system, he may need a new waterblock soon! :(

The waterbutt idea sounds like a good plan, but I'm on the 3rd floor so I'd need a reallly long pipe! Have you considered drilling a couple of holes through the wall? Might be a bit drafty, but no security implications, and frost (and snow) could be safely ignored by keeping the pump on 24/7, although condensation might then become an issue.

The Apache is 26.4mmH20 static pressure (and 57.53CFM).
Found a spec sheet for the san ace 120 38mm: 36.83mmH20 so yer that's quite an improvement. 64dB(A) though, vs 16.05dB(A)! Not to mention there seems to be minimal availability of these fans, unless you know where to look? I think the Apache's are better at home, the San Ace could be good in a server environment though.
 
Last edited:
Big grins at that loop set up. The sock in particular is brilliance. Hose that kinks is no fun at all. I'm now using 1.5mm wall thickness tygon with steel springs around it, which does solve the problem until such time as you run out of springs.

The quotes specs for the apache are a lie, the consensus reached on here was that they're probably out by a factor of ten. There's simply no way they can be true. It's a convenient typo though so it seems to be persisting.

There are more impressive fans than the san ace out there, but few running at 12V or with the some undervolting performance. I believe Greebo is running two of them at about 3V on a radiator. I'd be less enthusiastic about them if I didn't have one blowing far more air through two radiators than a 1600rpm fan is managing with one. They're very rare though.

Re waterbutt I'm on the second floor myself, so not a success yet. Shall try to live somewhere more convenient next year. Subtle holes in the wall are probably the way to go.
 
The Apache's static pressure you've stated here is a mistake mate, but a lot of sites that sell them still seem to advertise them with those specs..

The Akasa site has the correct specs listed.

::edit::

Beaten to it, should have refreshed the page first before posting!!!
 
Fair enough. Thanks for the linky as well. This makes me want the San Ace type more, but I'm not confident on finding any :(

I thought I might not be able to handle the noise level but, turns out, I keep underestimating how noisy my current things are :p
Pulled out my sound meter and my case is emitting ~50dB(A) at 0.5' so a San Ace at half capacity would probably be in keeping with my current sound levels!
 
All this talk about water butts etc, have you factored in the static head height the pump will have to deal with? I have a hard time imagining even an 18W ddc pumping up three storeys somehow, though I may be underestimating it, depends what its pump curve is like. What about the cold water tank in the loft anyway, imagine what it would be like in the winter, again assuming the pump has enough head.
 
Bubo said:
All this talk about water butts etc, have you factored in the static head height the pump will have to deal with? I have a hard time imagining even an 18W ddc pumping up three storeys somehow, though I may be underestimating it, depends what its pump curve is like. What about the cold water tank in the loft anyway, imagine what it would be like in the winter, again assuming the pump has enough head.
Click to expand...

We're not considering water butts from the third floor :p
But anyway, it wouldn't have to pump up three storeys - it'd only have to overcome the resistance of that length of pipe, since what goes up must come down. In other words, the water moving downwards counterbalances the water moving upwards, so once set up, it only takes a pump sufficiently powerful to move the water through that length of pipe + waterblock on the level.
The main problem with using the cold water tank would be that it's meant to be cold, not warmed up by a computer. :p
 
miniyazz said:
The main problem with using the cold water tank would be that it's meant to be cold, not warmed up by a computer. :p
Click to expand...

Actually the traditional cold water tank is only really there to provide pressure for your hot water. If anything heating it up via a computer rad will actually slightly reduce your hot water bills, at the expense of possibly having slightly warmer upstairs bathroom cold water, if that cold tap comes off the tank rather than the mains.
 
Bubo said:
Actually the traditional cold water tank is only really there to provide pressure for your hot water. If anything heating it up via a computer rad will actually slightly reduce your hot water bills, at the expense of possibly having slightly warmer upstairs bathroom cold water, if that cold tap comes off the tank rather than the mains.
Click to expand...

Ah fair enough, in that case it's not a bad thought at all. Still, it doesn't help me as I'm in a flat and if we have a cold water tank, we don't have access to it!
 
You must log in or register to reply here.
Back
Top Bottom