What PSU is enough to start a car?

[burebista]

Let's see (http://www.youtube.com/watch?v=dUIKygE52kE). :crazy:

[Bobsama]

So that's all they're good for? You need 29A on a +12V rail for what? 5 seconds? Oooh wow major props to you, Cooler Master. :rolleyes: Good thing the car (much like your power supplies) don't need to be within ATX spec. :lol:

[Super Nade]

Interesting way of testing. I know nothing about cars, so is the peak draw only 10A?

I've moved it back to the PSU section. ;)

[Spectre]

So that is where Cooler Master's PR department went to..........

Interesting way of testing. I know nothing about cars, so is the peak draw only 10A?

I've moved it back to the PSU section. ;)

Depends on the car, the starter, the engine, the internal friction of the motor, weight of the crank, fuel, etc. To throw over certain motors you need higher torque starters which require even more current. Just one example, the 427 in my Impala has a specific high torque starter and a 1000A battery (no it doesn't actually throw all 1000A to the starter).

[Bobsama]

So why not toss a REAL power supply in a car? I'd like to see that EVGA 120A one start a HEMI or something!

[Spectre]

So why not toss a REAL power supply in a car? I'd like to see that EVGA 120A one start a HEMI or something!

The EVGA unit isn't 120A...............

[jonnyGURU]

Boy... they're grasping.

[Bobsama]

The EVGA unit isn't 120A...............

But you know they'll advertise it as "120A on +12V rails!!! OMG OMG OMG 1.2KW PSU FTW XXX Edition"

Anyways--I agree, they're really grasping. THey have nothing my old 300W PSU (bundled with a $20 chassis) couldn't do once!

[Solitaire]

Wouldn't that be 1.44KW?

"120A on +12V rails!!! OMG OMG OMG 1.5KW PSU FTW XXX Edition" :p

Yeah, peak for five seconds. Constant load would be more like 120W on +12V rails :lol:

And I like how CM have relocated their PR department to YouTube. Just seems the proper place for them :D

[Bobsama]

Wouldn't that be 1.44KW?

"120A on +12V rails!!! OMG OMG OMG 1.5KW PSU FTW XXX Edition" :p

Yeah, peak for five seconds. Constant load would be more like 120W on +12V rails :lol:

And I like how CM have relocated their PR department to YouTube. Just seems the proper place for them :D

No, because they claim to have 4 30A rails. In reality, it's something like 4 20-22A rails (when running all) plus full 3.3V and 5V, all adding up to a maximum of 1.2kW

[Spectre]

No, because they claim to have 4 30A rails. In reality, it's something like 4 20-22A rails (when running all) plus full 3.3V and 5V, all adding up to a maximum of 1.2kW

Come again?

[Bobsama]

Come again?

It's up to 30A on a single rail, if they are in fact quad-rail. The total combined amperage on the +12V rails is somewhere (far) south of 100A, of course. There was a thread about it at XS, about 3-5 pages back in Xtreme Trolling (News).

[Spectre]

It's up to 30A on a single rail, if they are in fact quad-rail. The total combined amperage on the +12V rails is somewhere (far) south of 100A, of course. There was a thread about it at XS, about 3-5 pages back in Xtreme Trolling (News).

I have a fair ide of how things work ;)

I said come again becasue this:

No, because they claim to have 4 30A rails. In reality, it's something like 4 20-22A rails (when running all) plus full 3.3V and 5V, all adding up to a maximum of 1.2kW

Makes no sense.

[Bobsama]

What I meant was each rail is supposed to have a max of 30A. I'm sure you have a grip on how your PSU is not 100% +12V (and for good reason), so when all's said and done, 120A is just marketing hype. I don't remember the real specs for the prototype (or if it was just here-say), but I think it was a total of about 88A when under load, which ends up being just north of 1kW. As I said, marketing hype from exaggerated specs plus gullible fools who hold anything EVGA brands as being made of gold.

[Spectre]

What I meant was each rail is supposed to have a max of 30A. I'm sure you have a grip on how your PSU is not 100% +12V (and for good reason), so when all's said and done, 120A is just marketing hype. I don't remember the real specs for the prototype (or if it was just here-say), but I think it was a total of about 88A when under load, which ends up being just north of 1kW. As I said, marketing hype from exaggerated specs plus gullible fools who hold anything EVGA brands as being made of gold.

What?

[Bobsama]

Marketing. When they release whatever they release (and let's hope it's a rumor), you'll see an actual unit and their marketing team work.

[Spectre]

Marketing. When they release whatever they release (and let's hope it's a rumor), you'll see an actual unit and their marketing team work.

No I wanted you to explain the bolded items. I don't think you quite understand what you are saying.

[Bobsama]

1) 4 30A rail thing...
Of course, most of these PSUs are overspec'd on the label. They can't deliver the full rated specs. Of course, DC VxA=W. That being said, 120x12=!1200, but 1440. There's also +5V and VSB rails, as well as +3.3V and -12V. Each are rated for specific wattages. Typically, there is an overall +12V cap, which is what I mean by saying that the prototype is said to have four individual +12V rails, each rated at 30A, though combined draw is still limited to something like 80-90A. Again, this is a prototype so nothing is in stone.

2) It's a PC power supply--so 3.3V and 5V rails are also required for proper function. I don't know what the ratings on those are (or -12V or 5VSB), but from current designs I would assume it's about 30A on 3.3V and 30A on 5V. Further, judging by current designs, I'd say the absolute max 12V draw is 95A, which leaves a measly 60W for everything else.

3) In practice and application, it would be much less than that, unless somebody was ONLY powering GPUs and fans with a unit like this. I'd assume based on those numbers, you'd get a typical max of ~85A on the four 12V rails.

Anyways--if you think I don't understand what I'm talking about, try shooting holes in what I say. And I mean that. I'm not an engineering student--I'm a business management student. By all means, I'm open to learning.

[Spectre]

1) 4 30A rail thing...
Of course, most of these PSUs are overspec'd on the label. They can't deliver the full rated specs. Of course, DC VxA=W. That being said, 120x12=!1200, but 1440. There's also +5V and VSB rails, as well as +3.3V and -12V. Each are rated for specific wattages. Typically, there is an overall +12V cap, which is what I mean by saying that the prototype is said to have four individual +12V rails, each rated at 30A, though combined draw is still limited to something like 80-90A. Again, this is a prototype so nothing is in stone.

4x30A is not over specing the unit.

2) It's a PC power supply--so 3.3V and 5V rails are also required for proper function. I don't know what the ratings on those are (or -12V or 5VSB), but from current designs I would assume it's about 30A on 3.3V and 30A on 5V. Further, judging by current designs, I'd say the absolute max 12V draw is 95A, which leaves a measly 60W for everything else.

False. There are and it is possible for units to output 100% of rated capacity on the 12v rails. Further, you can not "judge by current designs" that the absolute max draw is 95A as there are units on the market that easily do more than that and you aren't looking at the design being used or the components being used in this unit or any other unit to come up with that number.

3) In practice and application, it would be much less than that, unless somebody was ONLY powering GPUs and fans with a unit like this. I'd assume based on those numbers, you'd get a typical max of ~85A on the four 12V rails.

A typical max of 85A doing what or how? What are you talking about? 85A of draw? 85A of capacity?

Anyways--if you think I don't understand what I'm talking about, try shooting holes in what I say. And I mean that. I'm not an engineering student--I'm a business management student. By all means, I'm open to learning.

Perforated.

[Bobsama]

4x30A is not over specing the unit.
What I'm referring to is more deceptive advertising. IF this thing is going to be branded as 12V@120A, it's deceptive as there's simply no way. My opinion is that 4x12V@30A is overspecing it, because that 12V rail can't deliver that for any sustained period.


False. There are and it is possible for units to output 100% of rated capacity on the 12v rails. Further, you can not "judge by current designs" that the absolute max draw is 95A as there are units on the market that easily do more than that and you aren't looking at the design being used or the components being used in this unit or any other unit to come up with that number.
Yes, it's possible, but in actual use, how many systems us 100A on 12V rails? I don't know of any, so outside of specialized and secondary-market uses, saying 120A is pure bullshit. That being said, when this launches, I'd like see their combined 12V draw.


A typical max of 85A doing what or how? What are you talking about? 85A of draw? 85A of capacity?]
12V@85A total, and what I mean is running in an actual system (as sole PSU), that'd be a tad bit over 1kW. Of course, computers also draw from the other rails. Remember that this product has a primary purpose and primary target audience; to power the high-end rigs of computer enthusiasts. It's not designed to jump-start a car or do much else at all. Maybe that's a secondary purpose and market, but considering how little awareness there is of this particular use (to jump-start cars with 12V batteries), I would highly doubt that. Again, coming from a marketing standpoint.



Perforated.
Continue. I don't understand inquiries unless they're made clear to me. Again, I'm a Business Management major, not any sort of engineer, technician, or whatever.




On a semi-related note, I was talking to my friend, Loonym, about jump-starting his jeep (or recharging the battery) with a PSU before this particular topic was made known to me.

[Makalu]

What I'm referring to is more deceptive advertising. IF this thing is going to be branded as 12V@120A, it's deceptive as there's simply no way. My opinion is that 4x12V@30A is overspecing it, because that 12V rail can't deliver that for any sustained period.





It's not overspecing anything. 30A is the individual rail limit and somebody in marketing there at EVGA has made the common mistake of thinking it can output 30A on all the rails at once...which is a spec that will be given as the combined amount. The 30A limit for each rail is an equally valid but completely different spec.

[Super Nade]

Anybody know what is the ballpark number for transient draw in this case? What % of the average draw would it be, i.e + 30%, +20% etc?

[jonnyGURU]

I get what Bobsama is saying.. they would angle the marketing in such a way to make someone believe that 30A X 4 rails = 120A OMGWTFBBQ!

But you got the "XXX" thing wrong. That's XFX. You did get the "FTW" part right. ;)

[Spectre]

What I'm referring to is more deceptive advertising. IF this thing is going to be branded as 12V@120A, it's deceptive as there's simply no way. My opinion is that 4x12V@30A is overspecing it, because that 12V rail can't deliver that for any sustained period.

Again false about the 30A rail. A rail most certainly could deliver 30A @12v for sustained period. The division of 4 12v rails won't cut the individual OCP limit of a rail just because the units total OCP won't allow 120A @ 12v.



Yes, it's possible, but in actual use, how many systems us 100A on 12V rails? I don't know of any, so outside of specialized and secondary-market uses, saying 120A is pure bullshit.

Whoopie! Tha doesn't make your statement:

2) It's a PC power supply--so 3.3V and 5V rails are also required for proper function. I don't know what the ratings on those are (or -12V or 5VSB), but from current designs I would assume it's about 30A on 3.3V and 30A on 5V. Further, judging by current designs, I'd say the absolute max 12V draw is 95A, which leaves a measly 60W for everything else.

True by any stretch of the imagination. The unit could very well be capable of outputting 100% of its actual capacity on the 12v rails. There are units on the market that do it today. As I said earlier you aren't looking at the design being used or the components being used in this unit or any other unit to come up with your numbers on what a unit can do.


That being said, when this launches, I'd like see their combined 12V draw.

A power supply doesn't have a combined draw. Components requesting power form a power supply have a draw. A power supply has capacity. There is a HUGE difference.


12V@85A total, and what I mean is running in an actual system (as sole PSU), that'd be a tad bit over 1kW. Of course, computers also draw from the other rails. Remember that this product has a primary purpose and primary target audience; to power the high-end rigs of computer enthusiasts. It's not designed to jump-start a car or do much else at all. Maybe that's a secondary purpose and market, but considering how little awareness there is of this particular use (to jump-start cars with 12V batteries), I would highly doubt that. Again, coming from a marketing standpoint.

Then your numbers are completely made up again. A system will not draw 85A as some sort of set limit or value and the unit will not provide just 85A as some sort of set value all the time. A SMPS delivers the power that is requested from it dynamically. It is constantly changing. You are again confusing draw with capacity. A SMPS does not deliver its rated power all the time.

[Bobsama]

We're talking on different levels. Here's my basic points...
1) The primary market and purpose for this product is people wanting to run a computer with a single powerful power supply.
2) A lot of what I'm saying depends on what they'll market it as. The fact that it's been introduced as a +12V@120A power supply is deceiving to the vast majority of the market, as it can't sustain +12V@120A.
3) The manufacturer or brander, EVGA, has a lot of choices on final specifications and setup. If this is a true quad-12V-rail design, the final outcome will be drastically different than if it was a single-12V-rail design. Even so, these manufacturers get away with saying dual/triple/quadruple-12V-rail on single-12V-rail units. If it occurs that this is anything BUT a quad-12V-rail design, saying +12V1, +12V2, +12V3, and +12V4 are each rated to 30A is a blatant lie that the vast, vast majority of the market will not be wise enough to pick up.

So let's leave the rest of the discussion about the EVGA unit until it launches and somebody can review it properly.

[Spectre]

We're talking on different levels.

Not really.

Here's my basic points...
1) The primary market and purpose for this product is people wanting to run a computer with a single powerful power supply.

Uh ok?

2) A lot of what I'm saying depends on what they'll market it as. The fact that it's been introduced as a +12V@120A power supply is deceiving to the vast majority of the market, as it can't sustain +12V@120A.

Find me a source for this from EVGA not FUD that states the unit is doing something impossible.

3) The manufacturer or brander, EVGA, has a lot of choices on final specifications and setup. If this is a true quad-12V-rail design,

FUD says its a 2x12v.

the final outcome will be drastically different than if it was a single-12V-rail design.

How exactly would it be different than a single 12v unit subdivided into 4x12v with individual OCP limits on the 12v rials?

Even so, these manufacturers get away with saying dual/triple/quadruple-12V-rail on single-12V-rail units.

What do you mean get away with? What exactly are you accusing them of? Seasonic is the only real large name company that routinely has not set the individual OCP settings.

If it occurs that this is anything BUT a quad-12V-rail design, saying +12V1, +12V2, +12V3, and +12V4 are each rated to 30A is a blatant lie that the vast, vast majority of the market will not be wise enough to pick up.

/facepalm

No no it isn't. The units total OCP limit can be set below the additive value of the individual rails. That is what the whole combined section is for on multi12v units labels. Just like the 5v/3.3v combined limit.