Thermal Load Calculator by Innovatek

Want to talk about one of the articles in SPCR? Here's the forum for you.
Post Reply
MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Thermal Load Calculator by Innovatek

Post by MikeC » Tue May 09, 2006 2:25 pm


vitaminc
Posts: 306
Joined: Mon Mar 27, 2006 8:11 am
Location: Silicon Valley, California

Post by vitaminc » Tue May 09, 2006 2:31 pm

Always wondered why SPCR didnt make one :p

That calculator is still very lacking at the expansion cards, FDD, HDD, CD, motherboard, etc part, especially CD/DVD, FDD, 4/6/8 phase PWM mobo, case fan, and 15k RMP HDDs. :(

MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Post by MikeC » Tue May 09, 2006 3:16 pm

vitaminc wrote:Always wondered why SPCR didnt make one :p

That calculator is still very lacking at the expansion cards, FDD, HDD, CD, motherboard, etc part, especially CD/DVD, FDD, 4/6/8 phase PWM mobo, case fan, and 15k RMP HDDs. :(
Motherboard is covered... but it seems a little iffy. The rest of your list is not really significant -- because you cannot really access this at the same time that the GPU & CPU are at full tilt. So if you gave yourself a 25% margin, you'd be fine. IE, if the total comes to 300W and you get a 350~400W PSU...

Slaugh
Posts: 774
Joined: Thu Dec 02, 2004 2:27 am
Location: Quebec, Canada

Post by Slaugh » Tue May 09, 2006 3:20 pm

Here's another one with more options...

http://www.extreme.outervision.com/psucalculator.jsp

paulesko
Posts: 35
Joined: Sat May 06, 2006 1:53 am
Location: Spain (Europe)

Post by paulesko » Tue May 09, 2006 3:40 pm

Slaugh wrote:Here's another one with more options...

http://www.extreme.outervision.com/psucalculator.jsp
this one is crazy. I´ve donde my configuration and it tells me that I need 380 watts while I´m using a 300w psu with no problem.

I´ve tried a hard configuration with two opterons 260 (overclocked to 2900 with water :D ) and two 7900 gtx sli 2 hd´s and tells me more than 1000watts!!!!! that´s not true anyway

frostedflakes
Posts: 1608
Joined: Tue Jan 04, 2005 4:02 pm
Location: United States

Post by frostedflakes » Tue May 09, 2006 8:29 pm

Yeah, the eXtreme calculator is up to date with the latest parts, but it tends to overestimate wattage quite a bit.

For example, it tells me I need 250w DC, yet while gaming (basically the highest real-life load I could achieve) power draw was only 140w AC (~110w DC), so... take what it tells you with a grain (truckload?) of salt. :lol:

teknerd
Posts: 378
Joined: Sat Nov 13, 2004 5:33 pm

Post by teknerd » Wed May 10, 2006 1:34 am

I think the confusion with the innovatek calculator arises from the fact that it is not a PSU calculator but rather a calculator for determining the amount of cooling that your water cooling system needs to be able to provide.
This is also why there are options for northbridge/northbridge and mosfets - whether you are cooling just the northbridge or also adding a block to cool the Voltage circutry.

Ackelind
Posts: 467
Joined: Sun Mar 20, 2005 1:18 pm
Location: Umea, Sweden.

Post by Ackelind » Wed May 10, 2006 5:46 am

I'd have to say that it roughly overestimates my system draw about 100%. Yes that is true, 100%.

There is no way in h*** my system will draw 310+W no matter how hard I push it. I'd say my system is roughly around 150W based on earlier SPCR tests.

I also had to set my CPU to Opteron 144 instead of A64 3000+. Using the later gave me specs of [email protected] which is based on the Clawhammer core. There is no way to even lower the voltage to the standard 1.4V, or below it.

Interesting enough, a system based on a Opteron 852 running at 5.2GHz (yeah right) @1.75V with two X1900XTX in SLI and 10*10krpm drives gave me roughly double the amount compared to my current system (670W vs 310W).

Sendorm
Posts: 56
Joined: Wed Jun 08, 2005 10:21 pm

Post by Sendorm » Thu May 11, 2006 2:46 am

teknerd, seems to have covered the idea. Let me reexplain.
Firstly this is not a psu calculator.

It only adds the heats of several pc component you'll be water cooling and give you a rough wattage.
Using that value you choose a radiator suited to your needs. Radiators are also watt rated, so this is just apples vs apples.

The watt values are clearly not the electricity usage of the components. The values are nearly of by %100 in some cases. For instance my agp 6600gt is rated at 120watts in the calculator, which uses 55watt at peak usage.

Ok so the calculator tells me that if I want to cool my graphics adapter adequately (in the expertenmodus; you can select cool-matic or graph-o-matic for gpu cooling. You'll see that ram combo-box disables, because one solution has ram cooler incorporated, the other has not) I'll need to get a radiator rated at 120watts.

supastar42
Posts: 6
Joined: Fri Sep 16, 2005 7:40 am
Location: UK
Contact:

Post by supastar42 » Fri May 12, 2006 3:19 am

Thank you teknerd and sendorm. This is a THERMAL calculator, not a power calculator. The figures used in the calculation are derived from manufacturer's data and will probably never be reached in practice. The best figure, therefore, to take is the middle, Typical wattage.

It allows a user to fit a radiator (or radiators) for performance or quietness, depending on your need.

Ackelind
Posts: 467
Joined: Sun Mar 20, 2005 1:18 pm
Location: Umea, Sweden.

Post by Ackelind » Sun May 14, 2006 3:16 pm

Power calculation and thermal calculation is the same in a computer. Everything is converted into heat.

jaganath
Posts: 5085
Joined: Tue Sep 20, 2005 6:55 am
Location: UK

Post by jaganath » Sun May 14, 2006 3:57 pm

Power calculation and thermal calculation is the same in a computer. Everything is converted into heat.
I've always wondered about this. Surely some energy is required to switch the transistors on and off? So maybe only 99.9% of the electrical energy is converted into heat.

supastar42
Posts: 6
Joined: Fri Sep 16, 2005 7:40 am
Location: UK
Contact:

Post by supastar42 » Mon May 15, 2006 1:12 am

You are completely missing the point here.

The Innovatek Thermal calculator is designed to give figures for the specific components that the user selects and wants to watercool. This result is then used primarily to spec a radiator.

A Power calculator would cover the total energy requirements of the PC in order, say, to size a power supply.

loz
Posts: 49
Joined: Wed Dec 28, 2005 1:29 am
Location: Grenoble-France

Post by loz » Mon May 15, 2006 4:15 am

jaganath wrote:
Power calculation and thermal calculation is the same in a computer. Everything is converted into heat.
I've always wondered about this. Surely some energy is required to switch the transistors on and off? So maybe only 99.9% of the electrical energy is converted into heat.
Yes, "some energy is required to switch the transistors", but what do you think think this energy become ? 100% heat.
The only exceptions I can think of are:
- mechanical wear in mobile parts (hdds and fans)
- emitted light (monitor and leds) which will... heat the walls (and other objects, including you) it hits !
Neverthless, it's probably so low nobody could measure it.

Ackelind
Posts: 467
Joined: Sun Mar 20, 2005 1:18 pm
Location: Umea, Sweden.

Post by Ackelind » Mon May 15, 2006 6:43 am

It is true that some energy is converted into kinetic energy when the HDD and fans are accelerated to their set speed. But when the computer is shut off, this energy is converted into heat, which maked the fans/HDDs grind to a halt.

Everything IS converted into heat (except for light perhaps). Even if this calculator is used to calculate the thermal load on radiators, it still overrates some parts. Is this to sell more radiators?

jaganath
Posts: 5085
Joined: Tue Sep 20, 2005 6:55 am
Location: UK

Post by jaganath » Mon May 15, 2006 7:31 am

Yes, "some energy is required to switch the transistors", but what do you think think this energy becomes ? 100% heat.
No, it won't be 100% heat, because in transforming energy from one form to another (electrical to thermal) there are always conversion losses. There is no conversion process that is 100% efficient (second law of thermodynamics).

Devonavar
SPCR Reviewer
Posts: 1850
Joined: Sun Sep 21, 2003 11:23 am
Location: Vancouver, BC, Canada

Post by Devonavar » Mon May 15, 2006 7:54 am

Yes, it will become 100% heat (assuming energy that remains as light/sound is negligible, which it is). Conversion losses are inevitably losses in the form of heat. The total energy in the system does not change. Looking at the system as an isolated box, energy in has to equal energy out or it would would not obey another law of thermodynamics: Conservation of Energy in a System.

On another note,
supastar42 wrote:Thank you teknerd and sendorm. This is a THERMAL calculator, not a power calculator. The figures used in the calculation are derived from manufacturer's data and will probably never be reached in practice. The best figure, therefore, to take is the middle, Typical wattage.
Regardless of what it is for or how it gets there, this "thermal calculator" provides a much more realistic estimate of total power than any other PSU calculator that I've seen. Total combined maximum power (i.e. maximum possible power draw for every single component in the system) is simply not a good way of gauging power needs — it invariably over-estimates because no system will every see every component fully stressed simultaneously.

kimandsally
Posts: 11
Joined: Mon Dec 27, 2004 9:48 am

Post by kimandsally » Tue May 16, 2006 12:40 pm

I disagree it cannot all be converted into heat otherwise electric fires would not be required or made, I agree quite a lot is turned into heat but a computer is a very inefficient heater.

Ackelind
Posts: 467
Joined: Sun Mar 20, 2005 1:18 pm
Location: Umea, Sweden.

Post by Ackelind » Tue May 16, 2006 1:35 pm

kimandsally wrote:I disagree it cannot all be converted into heat otherwise electric fires would not be required or made, I agree quite a lot is turned into heat but a computer is a very inefficient heater.
What other types of energy will it be converted into then?

Since it isn't accelerating, moving towards gravity or providing any pneumatic preasure, all will become heat. There is just no other explanation.

jaganath
Posts: 5085
Joined: Tue Sep 20, 2005 6:55 am
Location: UK

Post by jaganath » Tue May 16, 2006 3:11 pm

What other types of energy will it be converted into then?

Since it isn't accelerating, moving towards gravity or providing any pneumatic preasure, all will become heat. There is just no other explanation.
A computer can do useful work, right? So some energy must be used to do this work:

work (physics)
Types of work
Forms of work that are not evidently mechanical in fact represent special cases of this principle. For instance, in the case of "electrical work", an electric field does work on charged particles as they move through a medium.
So the electrical field in the transistors in a CMOS device do work on the charge carriers.

Ackelind
Posts: 467
Joined: Sun Mar 20, 2005 1:18 pm
Location: Umea, Sweden.

Post by Ackelind » Tue May 16, 2006 9:10 pm

Energy is required only to accelerate particles or objects, not to maintain a constant speed without friction.

However, when these particles stop moving, as they surely will unless going on forever without friction, the kinetic energy will convert into heat through frictional losses.

Vicotnik
*Lifetime Patron*
Posts: 1831
Joined: Thu Feb 13, 2003 6:53 am
Location: Sweden

Post by Vicotnik » Wed May 17, 2006 6:22 am

jaganath wrote:No, it won't be 100% heat, because in transforming energy from one form to another (electrical to thermal) there are always conversion losses. There is no conversion process that is 100% efficient (second law of thermodynamics).
No conversion process that is 100% efficient, that is correct. There are always losses. Losses in the form of heat usually.
So, again: 100% heat.
jaganath wrote:A computer can do useful work, right? So some energy must be used to do this work:

work (physics)

So the electrical field in the transistors in a CMOS device do work on the charge carriers.
What matters is that in the end all energy will be converted into heat. That there is some other form of energy conversion before that is really beside the point.
The only energy that doesn't end up in the form of heat is the energy that escapes in the form of sound and light. That amount of energy is negligible and is most likely converted into heat eventually to, just not inside the computer.

loz
Posts: 49
Joined: Wed Dec 28, 2005 1:29 am
Location: Grenoble-France

Post by loz » Wed May 17, 2006 7:10 am

So, the noisier my computer is, the less heat it generates... sounds crazy :lol:

More seriously : if you doubt 99.99% energy is converted into heat, tell us what it becomes...

jaganath
Posts: 5085
Joined: Tue Sep 20, 2005 6:55 am
Location: UK

Post by jaganath » Wed May 17, 2006 8:10 am

So, the noisier my computer is, the less heat it generates... sounds crazy
There's nothing crazy about it; power MOSFETs' on-state resistance is positively correlated with their temperature, so as their temperature rises so does heat dissipation. In extreme circumstance thermal runaway can occur.
More seriously : if you doubt 99.99% energy is converted into heat, tell us what it becomes...
Power supplies and long traces on motherboard PCB's emit electro-magnetic radiation; this manifests itself as noise in nearby radios, TV's and wireless phones (sound and light). I never said 99.9% of the energy is not converted to heat, but it is not 100%. The pressure wave that reaches your ears is converted into electrical signals in the brain which you perceive as sound. Same with light.

loz
Posts: 49
Joined: Wed Dec 28, 2005 1:29 am
Location: Grenoble-France

Post by loz » Thu May 18, 2006 3:00 am

jaganath wrote:The pressure wave that reaches your ears is converted into electrical signals in the brain which you perceive as sound. Same with light.
And "electrical signals in the brain" finally turn to ... heat !
So the noisier my computer is, the more heat is generated in my brain, and I've got headaches, it does make sense...

(It's not meant to be serious)

supastar42
Posts: 6
Joined: Fri Sep 16, 2005 7:40 am
Location: UK
Contact:

Post by supastar42 » Fri May 19, 2006 12:55 am

The power figures used in the Thermal Calculator are all based on the manufacturer's data. And as Devonavar rightly says, no system is ever stressed to the full so the load figures will never be reached in practice.

This is why the centre reading, Typical (typisch) is used for design purposes. So, Ackelind, the question of over sizing radiators does not arise.

Post Reply