• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Power Supply I won from Nvidia

L

Lord_Exodia

Supreme [H]ardness
Joined
Sep 29, 2005
Messages
7,008
I won a Rel 80 contest and won a SLi ready power supply which happens to be a Silverstone Strider 600Watt. I have read reveiws that say good things but what do you guys think about it? Anyone else have 1. Supposedly they are very future proof with Quad 12v rails. Your comments welcome.

Here is linkage to 1 review

http://www.techimo.com/articles/i275.html
 
Lord_Exodia said:
Supposedly they are very future proof with Quad 12v rails.
Click to expand...

I see in your future Dual CPUs :p

EPS12V split plane supplies have under the spec
2 dedicated rails, one for each CPU
one for the mobo
and one for drives ect
 
1 thing to consider for the strider 600W is that 16A is shared between the CPU and the PCIe cards. I have a feeling this isn't enough on an overclocked dual core + an sli setup.
 
http://www.silverstonetek.com/products-60f.htm
+12V1 13A
+12V2 18A
+12V3 16A
+12V4 8A

its pretty obvious that the rails arent routed to components by spec
its also fairly safe to assume they wired +12V1 up for the CPU
Im guessing +12V2 is for the mobo & PCI-E via the mobo
+12V3 likely for SLI auxillary power
and +12V4 drives & fans

but Im just guessing, what you wont see is the CPU sharing with anything, that is definately against spec and was the whole point of split rails in the first place to isolate fluctuation of voltage from the CPU

EPS12V 6.1.1 12V Power Rail Configuration

There are two types of 12V rail configurations for systems: 'Common plane' and "Split plane' processor power delivery. The 'common plane' system has both processors powered from a single 12V rail (+12V1) from the power supply. The 'split plane' system has both processors powered by seperate 12V rails (+12V1 and +12V2) one dedicated to each processor. The system in both cases, has an additional 12V rail to power the rest of the baseboard +12V loads and dc/dc converters. +12V1, +12V2 and +12V3 should not be connected together on the baseboard to ensure that 240VA protection circuits in the power supply operate properly
Table 6: 12V Rail Summary
........................................................................................................................................................................................
Common Plane System........................................................Split Plane System
+12V1........Processors.........................................................+12V1........Processor 1
+12V2........Baseboard components other than processors.......+12V2........Processor 2
+12V3........Drives and peripherals..........................................+12V3........Baseboards components other than processors
...........................................................................................+12V4........Drives and peripherals
Click to expand...
 
The ST60F is the same as an ST65ZF but is modular.

My review of the ST65ZF is here:

http://www.slcentral.com/silverstone-zeus-st65zf/

I have the ST60F to review (courtesy of Silverstone) but quad rails are such a pain to review I've been procrastinating. Espeically since I know it's the same thing, only modular.

12V3 goes to the motherboard's ATX connector (what would normally be 12V1) and the SATA connectors. What Silverstone labels as 12V1 and 12V2 are both on the EPS connector. If you use the 4-pin connector, the load of the CPU (what is normally 12V2) is split between 12V1 and 12V2. In other words, if your CPU puts an 8A load on the 12V rail, you'd put a 4A load on each of the 12V1 and 12V2 rails on the ST60F. All of the drives, fans, lights, etc are on 12V4.
 
so it is routed per spec :p

back to the crystal ball :p

it would be upto the mobo manufacturer to split +12V1 and +12V2 at the main connector and route them wherever, in the case of a dually board, each to its own CPU
 
Ice Czar said:
it would be upto the mobo manufacturer to split +12V1 and +12V2 at the main connector and route them wherever, in the case of a dually board, each to its own CPU
Click to expand...

Not really. EPS spec for a quad rail has each CPU on it's own rail. If you use it with a single CPU board with a four pin and use the four pin modular cable, the connector actually taps into BOTH rails. The motherboard wouldn't split that power. Ohm's law would just allow the CPU to draw from each rail equally, aside from any resistance differences between the two rails (caused by a loose connector, etc.)
 
:rolleyes: *sigh*

I guess listening to me is like trying to figure out code.
 
Lord_Exodia said:
This is starting to sound good. Sorry about the crappy review I put. This is a review more deserving of they eyes of the elite [H] readers.

http://www.pcmoddingmy.com/e107_plugins/content/content.php?content.322

Do you guys think it's a good win. I did get it free!!! :D
Click to expand...


Whats the issue? Its stable, has good reviews and its modular.
Its bigger then some, and the modular idea does have its own downside, but it was free.


I'd take this free before paying for a Seasonic S12 600W.
 
jonnyGURU said:
:rolleyes: *sigh*

I guess listening to me is like trying to figure out code.
Click to expand...

Listening to you isn't that bad Johnny, I understand why you are procrastinating since it IS the same damn thing just modular. Anything different on the PCB this time that would improve or hurt my new PSU? if they are the same but modular why go through all the pain in the ass testing right?

That is a good review too.
 
jonnyGURU said:
The ST60F is the same as an ST65ZF but is modular.

My review of the ST65ZF is here:

http://www.slcentral.com/silverstone-zeus-st65zf/

I have the ST60F to review (courtesy of Silverstone) but quad rails are such a pain to review I've been procrastinating. Espeically since I know it's the same thing, only modular.

12V3 goes to the motherboard's ATX connector (what would normally be 12V1) and the SATA connectors. What Silverstone labels as 12V1 and 12V2 are both on the EPS connector. If you use the 4-pin connector, the load of the CPU (what is normally 12V2) is split between 12V1 and 12V2. In other words, if your CPU puts an 8A load on the 12V rail, you'd put a 4A load on each of the 12V1 and 12V2 rails on the ST60F. All of the drives, fans, lights, etc are on 12V4.
Click to expand...

I possibly understood this rail assignment incorrectly.

So the motherboard (except the CPU) and PCIe cards is powered by the +12V3 and the CPU is powered by +12V1 and +12V2. Is this right? thanks
 
Correct.

Unless you use the 4-pin connector. Then the load of the one CPU socket is split evenly between 12V1 and 12V2.
 
Using Everest instead of a multimeter, and then blaming software and saying the BIOS which read from the same chip were my only probs w/the first review :eek:
 
jonnyGURU said:
.

12V3 goes to the motherboard's ATX connector (what would normally be 12V1) and the SATA connectors. What Silverstone labels as 12V1 and 12V2 are both on the EPS connector. If you use the 4-pin connector, the load of the CPU (what is normally 12V2) is split between 12V1 and 12V2. In other words, if your CPU puts an 8A load on the 12V rail, you'd put a 4A load on each of the 12V1 and 12V2 rails on the ST60F. All of the drives, fans, lights, etc are on 12V4.
Click to expand...

I just bought one of these saturday but I have not installed it. I was about to ask which rails the 4-pin ATX connector was on when I found the answer here. (yes jonny some of us understand the language you speak ;-)

I bought it because I can overclock my 7800 GT to 490/1.18 and I can overclock my 3800 X2 to 2750 but I can't overclock both at same time and be stable. And I waas 100% stable with a 6800 vanilla. I'm getting something like 485 watts when I plug in the overclocked figures to one of PSU calculators. which is probably too much for the enermax EG-565P-VE

What I find interesting is that if you have SATA drives it's on 12V2. which is shared with CPU/MOBO and not 12V4

Thinking about not using the SATA plug for all 4 of my drives but instead putting some or all of them on the 4 pin Ide plugs. Is plugs the correct terminology for where the cords plug into on the PSU?

Any way any thoughts on where to plug my sata drives are welcome. Especially since someday they may multiply.
 
jonnyGURU said:
Correct.

Unless you use the 4-pin connector. Then the load of the one CPU socket is split evenly between 12V1 and 12V2.
Click to expand...

somewhat confused with your reply.

I have an AMD64 3000+ and an ASUS board which requires a 20-pin and a 4pin connector.

Can you please tell me what rail would power the CPU and the motherboard?
 
jonnyGURU said:
12V3 goes to the motherboard's ATX connector (what would normally be 12V1) and the SATA connectors. What Silverstone labels as 12V1 and 12V2 are both on the EPS connector. If you use the 4-pin connector, the load of the CPU (what is normally 12V2) is split between 12V1 and 12V2. In other words, if your CPU puts an 8A load on the 12V rail, you'd put a 4A load on each of the 12V1 and 12V2 rails on the ST60F. All of the drives, fans, lights, etc are on 12V4.
Click to expand...

He tells you right there...
If you use the 4-pin connector, the load of the CPU (what is normally 12V2) is split between 12V1 and 12V2.
Click to expand...

Half the pins come from 12V2 and half come from 12V1...
 
Aren't they trying to phase out 12v from the SATA drives? I think I read somewhere that in the future (maybe now I don't know) that SATA drives will have 3.3v to the logic and 5v to the mechanics. 12v is there for legacy support. If that's right then (in the future) a 12v rail shared between motherboard and SATA wouldn't be as bad as a 12v rail being shared between motherboard and PATA drive or DVD/CD. Am I remembering correctly?
 
I think you're right about that. I think HDD's are migrating to all 5V. Not sure about the 3.3V to logic, though. I thought they were going to be all 5V.
 
Well, the Sata power connector has a 3.3v lead on it. and it makes sense to move logic to 3.3v. following lower power consumptopn trends. But it doesn't make a difference as no manufacturers have begun to move that direction. At least not that I know of.
 
ryan_975 said:
Well, the Sata power connector has a 3.3v lead on it. and it makes sense to move logic to 3.3v. following lower power consumptopn trends. But it doesn't make a difference as no manufacturers have begun to move that direction. At least not that I know of.
Click to expand...
not any sata connector ive ever seen, theyre all two grounds, a 12v and a 5v - same as molex. only thing that the sata power connector has over the molex is hotswap.
 
lithium726 said:
not any sata connector ive ever seen, theyre all two grounds, a 12v and a 5v - same as molex. only thing that the sata power connector has over the molex is hotswap.
Click to expand...

No. He's correct. PROPER SATA connectors have 3.3V. Pins 1, 2 and 3. The only kind that do not are "adapters" or modular PSU's with "universal" SATA cables on them. Unfortunately, I can't find where anybody discusses what the 3.3V is for. The SATA consortium states merely that it is for "future use."

But a move to 3.3V logic actually doesn't follow lower power consumption trends. If the drive needs to pull more wattage from a higher voltage rail, it will merely pull less amperage which will actually create less heat.

There's a reason why video cards regulate off of the 12V, as do CPU's, when they used to regulate off of the 5V and before that the 3.3V.

The same is true about power supplies: Why are 230V active PFC power supplies more common in Europe than full range active PFC? Because the amperage that a PSU would have to draw is too great at 115V for the unit to work as efficiently as desired.
 
jonnyGURU said:
No. He's correct. PROPER SATA connectors have 3.3V. Pins 1, 2 and 3. The only kind that do not are "adapters" or modular PSU's with "universal" SATA cables on them. Unfortunately, I can't find where anybody discusses what the 3.3V is for. The SATA consortium states merely that it is for "future use."

But a move to 3.3V logic actually doesn't follow lower power consumption trends. If the drive needs to pull more wattage from a higher voltage rail, it will merely pull less amperage which will actually create less heat.

There's a reason why video cards regulate off of the 12V, as do CPU's, when they used to regulate off of the 5V and before that the 3.3V.

The same is true about power supplies: Why are 230V active PFC power supplies more common in Europe than full range active PFC? Because the amperage that a PSU would have to draw is too great at 115V for the unit to work as efficiently as desired.
Click to expand...


Actually. CPUs and Video cards regulate off the 12v rail because of the lower level of noise from the larger voltage conversion to 1.x volt or 0.x volt needed to operate the chips. CPU's originally were 5v from the 8086 to the 80486 and early Pentiums. later Pentiums were 3.3v and required a voltage regulator on the motherboard. My old K6-2 was 2.4v which was regulated down from the 3.3v ATX line.

Another example of downward voltage trends is ISA cards movnig from 12v to 5v, PCI devices moving from 5v to 3.3v, RAM chips moving from 5v to 3.3v to 2.5 to whatever DDR2 and future DDR III are.

Also CPUs move to lower voltages to reduce the amount of heat produced. Heat production is a product of how fast an Electron (Current) is pulled (Voltage) through a wire (Resistance).

Power supplies are 230 volts in europe because the power grid is 230 volts. They generally don't have a 110v available (that i'm aware of). In the US most homes and businesses are wired for 110v. (Yes I know 220 exists, but it's generally used for appliances) So we have 110v supplies.
 
ryan_975 said:
Power supplies are 230 volts in europe because the power grid is 230 volts. They generally don't have a 110v available (that i'm aware of). In the US most homes and businesses are wired for 110v. (Yes I know 220 exists, but it's generally used for appliances) So we have 110v supplies.
Click to expand...

Umm.. Yes. The grids are 230V. You missed my point....

There's two different ways you can have an active PFC power supply: 230V only or Full Range.

The active PFC circuitry generates more heat with the greater amount of amperage it has to manage regardless if the output wattage is the same as it would be whether it had a 115V input or 230V input.

This is why some PSU manufacturers develop "EU Only" models that are not full range. They're 230V only. Because if they were full range and had to manage 4A @ 115V, they tend to run quite a bit hotter than 2A @ 230V and if they run hotter than the performance of the PSU is negatively effected. So they'll have a 230V active PFC model in the EU where active PFC is required and a non-PFC model with a 115/230V input switch elsewhere in the world.

And you're correct that there is less noise on the higher voltage rails and that, yes, heat production is a product of how fast an Electron (Current) is pulled (Voltage) through a wire (Resistance). But the equation for resistance is R=V/I. Voltage over Amperage equals Resistance. And resistive loss is P=I^2*R. Wattage equals Amperage squared times resistance. What this means in the hokey pokey of things is that resistance is reduced if amperage is reduced, but if we were to assume that resistance remains a contstant regardless of voltage then a reduction of voltage can allow for an increase of amperage.

Of course, this isn't good practice with electric motors as high voltage tends to push the magnetic portion of the motor into saturation, thus causing the motor to actually draw excess amperage in an effort to magnetize the iron beyond the point where magnetizing is practical. But we're not talking about motors. :D

Next time you go into a Radio Shack, look at the fuses. Fuses sizes vary by amperage, not voltage. A 230V 8A fuse is a tiny little thing. 230V @ 8A is 1840W! A 115V 15A fuse is huge by comparison. That's only 1725W. Now take a look at a 30A blade fuse in your car. That's a thick metal strip in there! Yet it's only 360W.
 
I do see what you mean in the power supplies. Sorry I misunderstood that.

jonnyGURU said:
And you're correct that there is less noise on the higher voltage rails...
Click to expand...

Not so much that there's less noise on the rail, but the regulation to a lower voltage eliminates the noise more effectively when coming from higher voltages

jonnyGURU said:
but if we were to assume that resistance remains a contstant regardless of voltage then a reduction of voltage can allow for an increase of amperage.
Click to expand...

Two words: Inrush Current.
A Logic device is a complicated mess of switches. When switches close and allow current to flow, the sudden application of voltage to a conductor cause a larger amount of current to flow for a brief moment. That inrush of current "hitting" the resistance cause the conductors to heat up and therefore increase the resistance of the circuit. Although the inrush current falls to a stable working current, but the working current still has to work through the higher resistance, producing more heat than normal. More heat means more resistance. .The cycle repeats until the thermal capabilities of the device are exceeded and it breaks down and no longer permits current to flow. Less complex logic devices can pass the excess heat off into the surrounding air, stabilizing the destructive cycle. However more complex devices (CPUs) can trap the produced heat within its layers letting the cycle run rampant. Active cooling (forced air, water cooking) is required to carry the heat away from the out layers of the device faster. The outter layers now have time to absorb the heat from the inner layers and pass it onto the outside air. This way is doesn't self destruct itself.

If voltage is too high, then the inrush current comes in faster heating things up more increasing resistance more, etc. So that doesn't work.

If a voltage for a device with a given inital resistance is too low, then the working current is greater (creating the same cycle above) So that doesn't work either.


The solution? Smaller components. A manufacturer has to make the devices smaller so that the conductive paths are shorter. Shorter conductors = less initial resistance.= less initial heatup slowing the cycle. A slower heat production cycle gives the cooling system more time to carry the heat away. More time means smaller, quieter fans. However, in the race to get their devices to do more and do it faster, manufacturers cram more heat producing components into the extra space they now have. Thus negating or reversing any gains they created from the smaller devices.

Imagine what the original Pentium processor's heat output would be if it were produced on a 65nm/.86v process instead of a 800nm/5v process. I don't think it'd need a heatsink.

On that note, imagine what the heat output of a Prescott P-IV would be at 800nm/5v

.
This probably doesn't make as much sense as I meant for it to. It's late (to me) and I'm not thinking straight anymore.
 
Dammit! Now I have to go read more books on electronics!!!!

:D ;) :p
 
you mean I actually made sense?

Equations will show you what will happen in a situation where everything is perfect. As we all know, nothing is perfect (except a ice cold beer on a hot day after working your arse off.)
 
ryan_975 said:
I do see what you mean in the power supplies. Sorry I misunderstood that.



Not so much that there's less noise on the rail, but the regulation to a lower voltage eliminates the noise more effectively when coming from higher voltages



Two words: Inrush Current.
A Logic device is a complicated mess of switches. When switches close and allow current to flow, the sudden application of voltage to a conductor cause a larger amount of current to flow for a brief moment. That inrush of current "hitting" the resistance cause the conductors to heat up and therefore increase the resistance of the circuit. Although the inrush current falls to a stable working current, but the working current still has to work through the higher resistance, producing more heat than normal. More heat means more resistance. .The cycle repeats until the thermal capabilities of the device are exceeded and it breaks down and no longer permits current to flow. Less complex logic devices can pass the excess heat off into the surrounding air, stabilizing the destructive cycle. However more complex devices (CPUs) can trap the produced heat within its layers letting the cycle run rampant. Active cooling (forced air, water cooking) is required to carry the heat away from the out layers of the device faster. The outter layers now have time to absorb the heat from the inner layers and pass it onto the outside air. This way is doesn't self destruct itself.

If voltage is too high, then the inrush current comes in faster heating things up more increasing resistance more, etc. So that doesn't work.

If a voltage for a device with a given inital resistance is too low, then the working current is greater (creating the same cycle above) So that doesn't work either.


The solution? Smaller components. A manufacturer has to make the devices smaller so that the conductive paths are shorter. Shorter conductors = less initial resistance.= less initial heatup slowing the cycle. A slower heat production cycle gives the cooling system more time to carry the heat away. More time means smaller, quieter fans. However, in the race to get their devices to do more and do it faster, manufacturers cram more heat producing components into the extra space they now have. Thus negating or reversing any gains they created from the smaller devices.

Imagine what the original Pentium processor's heat output would be if it were produced on a 65nm/.86v process instead of a 800nm/5v process. I don't think it'd need a heatsink.

On that note, imagine what the heat output of a Prescott P-IV would be at 800nm/5v

.
This probably doesn't make as much sense as I meant for it to. It's late (to me) and I'm not thinking straight anymore.
Click to expand...

I did not know that. /carson
 
ryan_975 said:
you mean I actually made sense?

Equations will show you what will happen in a situation where everything is perfect. As we all know, nothing is perfect (except a ice cold beer on a hot day after working your arse off.)
Click to expand...

Yes, it made sense. I learn as I go. I'm no EE. ;)

I do see your point, but I still don't think it applies throughout the entire PC platform. You can still assume some instances are "perfect" when it comes to power distribution with computers because the gage of wire often used is overkill and causes negligible resistance. The rule that higher voltage, lower amperage is less resistance can, and will often, still apply.
 
ryan_975 said:
(except a ice cold beer on a hot day after working your arse off.)
Click to expand...

get rid of the hot day and working your arse off part and youd be closer :p

That made quite a bit of sense and has parallels in LVDS at the board and drive level as well
(which also highlights the EMI part you didnt focus on)
 
Well I was referring more to the chip level conductors, which are usually microscopic and resistance is a factor. But wires can be a resistance point. Not in computers mind your. but take a 18awg wire and short out your wall socket. Resistance will very quickly show its ugly head.

I didn't even think about EMI. Which is a product of high switching speeds mostly. But higher amperages and voltages will propigate (sp?) stronger EMI when components cycle.

On the power supply side of things, higher resistance in the boards will work the power generating circuits harder. So manufacturer have to keep the power consumed by their devices down. Otherwise people willl have to start wiring their house with dedicated PC circuits. PSU's already pull around 8 amps from the wall. Which is I think 800-1000 watts depending on line conditions.. A 15amp branch circuit can only handle around 1725
Watts before tripping. Add on that same circuit everything else on has in their room (TV's Stereo's, Lava Lamps, 24" Monitors, etc) and you can see the needs to find a way to keep PSU consumption down. The way to do that is lower voltage and current levels in the devices that make up the computer so the PSU "sees" less resistance and works less hard.


I guess I really hijacked this thread. That was unintentional. I was originally trying to answer one person's post on the first page about whether or not to use his molex connectors to plug his SATA drive. Don't know if I ever really did that.
 
never worry about hijacks that answer related questions and "go deep"
they get quoted and requoted as examples and explanations ;)

however a "typical" load on a house circuit by a "typical desktop" wouldnt be 8A
http://www.silentpcreview.com/article265-page1.html
(even after conversion inefficiencies)

but your right about when you start adding everything else on
people rarely need less power :p

one of the things I always warn about is any kind of intermittent heavy amp draw on the same circuit (AC compressor for example) thus how important mapping your house circuits are
 
Circuit layout is important, yes. But it's very hard to control unless you've built your own house. I live in a house that was built 30 years ago (5 years before I was born). My computer is in the living room and every time my refrigerator in the kitchen kicks on, my desktop speakers pop (my kitchen is on a different circuit though). When my son turns on the TV my monitor screen wiggles like I just degaussed it.

When talking about power consumption, though, "typical usage" isn't as much of importance as "What is the most possible draw that each device will have on the circuit while still considered to be operating 'normally' ?." I've seen my teenage cousin's room where he's got two desktops going one with a 21" CRT and the other with a 17" CRT. A 27" CRT television with a PS2 game going, four or five bulbs burning, guitar amp running. and a mini fridge in the corner (No, he's not the only one in there when this is all going on. His room is the epicenter of his social life). This was all in one room which probably shared a branch with hsi sister's room. She also has a small TV and a computer of her own.. People generally don't think about how much power they're using, much the limits they're surpassing, so manufacturer's have to.


PS. And his parents complain that they have a $1000/month utilities expense. Makes me dread when my kids (TWO boys) hit 14-15.
 
You must log in or register to reply here.
Back
Top