Real risk in useing 8pin adapter?

[OCGuy]

You could obviously say im not a power supply expert....

What are the dangers in using the adapter that takes 2 of my 6 pins and makes them a single 8? I just installed them on my new GTX280. EVGA sends a warning in the box that the adapter is to be used short-term only.

This thread is getting closed because it's become horribly off-topic.

jonnyGURU
Power Supplies Moderator

 

[Zap]

They're just trying to CYA.

 

[HOOfan 1]

CYA?

shouldn't it be CTA? Cover their ass?

 

[beray]

Originally posted by: Ocguy31
You could obviously say im not a power supply expert....

What are the dangers in using the adapter that takes 2 of my 6 pins and makes them a single 8? I just installed them on my new GTX280. EVGA sends a warning in the box that the adapter is to be used short-term only.

The basic inherent problem is power factor.

Let's say your GPU used 200W from 12V, using the 8-pin adapter boosted the 12V power connection to a total final impedance of 0.03O. Your PSU would have to put out the real power of 208W for the 200W apparent power used by your GPU.

In the short term 8W is nothing as long as your PSU could handle the job, especially if the PSU is really powerful.

In the long term, the 8W real/apparent power difference would cause accelerated oxidation at the power connectors, the PSU would put out even greater real power to compensate. --> Your power connectors would melt or catch on fire and they nearly always do.

There are other problems but they are minor in comparison.

 

[Aluvus]

Originally posted by: beray

The basic inherent problem is power factor.

Let's say your GPU used 200W from 12V, using the 8-pin adapter boosted the 12V power connection to a total final impedance of 0.03O. Your PSU would have to put out the real power of 208W for the 200W apparent power used by your GPU.

I have never seen anyone use power factor or apparent power this way, and in fact both terms have well-defined meanings... for AC power systems.

I had started to type up an explanation of how these terms are used, but Wikipedia seems to have it mostly right.

In the long term, the 8W real/apparent power difference would cause accelerated oxidation at the power connectors, the PSU would put out even greater real power to compensate. --> Your power connectors would melt or catch on fire and they nearly always do.

WTF? Emphasis added. Unless you are using "long term" in a sense similar to a famous Keynes quote ("Long run is a misleading guide to current affairs. In the long run we are all dead."), I suppose...


@OP: Adapters of that type introduce a few concerns:

The adapter adds some series resistance. This can have several effects, but we will focus on the extra power that is turned into waste heat, and the voltage that is then lost across the connector. The extra heat adds to the heat to be evacuated from the case. The voltage loss creates a voltage droop, such that the input voltage seen by the video card is not quite as high as the output voltage of the power supply. In practice, these are normally not major problems.

The adapter also creates a new potential point of failure. The additional set of connectors can come loose or otherwise fail. Again, this normally does not happen (why would anyone sell you an adapter that they expect to fail?), but it can. It can happen without the adapter, as well; the adapter just adds more connections that can potentially fail.

In general, most people that use these sorts of adapters don't have any major problems with them. Ideally, you would have an 8-pin connector on your power supply (well, ideally you would have a cable soldered between the power supply and the card...). But using this sort of adapter will not make your computer suddenly explode in 6 months. In the event something does go awry, EVGA doesn't want to be blamed.

 

[beray]

Originally posted by: Aluvus

Originally posted by: beray

The basic inherent problem is power factor.

Let's say your GPU used 200W from 12V, using the 8-pin adapter boosted the 12V power connection to a total final impedance of 0.03O. Your PSU would have to put out the real power of 208W for the 200W apparent power used by your GPU.

I have never seen anyone use power factor or apparent power this way, and in fact both terms have well-defined meanings... for AC power systems.

I had started to type up an explanation of how these terms are used, but Wikipedia seems to have it mostly right.

Here's a tip - ActivePFC altered the input impedance of the PSU and correct the power factor dynamically. Try to to think of static instead of dynamic.

In the long term, the 8W real/apparent power difference would cause accelerated oxidation at the power connectors, the PSU would put out even greater real power to compensate. --> Your power connectors would melt or catch on fire and they nearly always do.

WTF? Emphasis added. Unless you are using "long term" in a sense similar to a famous Keynes quote ("Long run is a misleading guide to current affairs. In the long run we are all dead."), I suppose...


@OP: Adapters of that type introduce a few concerns:

The adapter adds some series resistance. This can have several effects, but we will focus on the extra power that is turned into waste heat, and the voltage that is then lost across the connector. The extra heat adds to the heat to be evacuated from the case. The voltage loss creates a voltage droop, such that the input voltage seen by the video card is not quite as high as the output voltage of the power supply. In practice, these are normally not major problems.

The adapter also creates a new potential point of failure. The additional set of connectors can come loose or otherwise fail. Again, this normally does not happen (why would anyone sell you an adapter that they expect to fail?), but it can. It can happen without the adapter, as well; the adapter just adds more connections that can potentially fail.

No one sold ATX PSUs and expect the ATX power connector be melted, they are made for long term use and still melted regularly just the same --> Thousands of examples.

In general, most people that use these sorts of adapters don't have any major problems with them. Ideally, you would have an 8-pin connector on your power supply (well, ideally you would have a cable soldered between the power supply and the card...). But using this sort of adapter will not make your computer suddenly explode in 6 months. In the event something does go awry, EVGA doesn't want to be blamed.

Of course they would, EVGA expected the adapters to have problem in long term use, including melting. Idiots will ignore it as usual.

 

[VirtualLarry]

Originally posted by: beray

Originally posted by: Aluvus

Originally posted by: beray

The basic inherent problem is power factor.

Let's say your GPU used 200W from 12V, using the 8-pin adapter boosted the 12V power connection to a total final impedance of 0.03O. Your PSU would have to put out the real power of 208W for the 200W apparent power used by your GPU.

I have never seen anyone use power factor or apparent power this way, and in fact both terms have well-defined meanings... for AC power systems.

I had started to type up an explanation of how these terms are used, but Wikipedia seems to have it mostly right.

Here's a tip - ActivePFC altered the input impedance of the PSU and correct the power factor dynamically. Try to to think of static instead of dynamic.

I thought that Power Factor was always 1.0 in DC systems?

Active PFC doesn't come into play here, that's for the AC input side of the PSU, not the DC output side.

 

[beray]

Originally posted by: VirtualLarry

I thought that Power Factor was always 1.0 in DC systems?

Active PFC doesn't come into play here, that's for the AC input side of the PSU, not the DC output side.

It is, as long as the output impedance is so small that it can be ignored. As the output impedance goes up, it can't be ignored.

ActivePFC basic function is matching the PSU input impedance to the output power source in use for maximum power transfer. AKA the highest possible power factor.

 

[Aluvus]

Originally posted by: beray
Here's a tip - ActivePFC altered the input impedance of the PSU and correct the power factor dynamically. Try to to think of static instead of dynamic.

And? You are still using the terms incorrectly and have yet to explain what PFC actually has to do with anything being discussed.

No one sold ATX PSUs and expect the ATX power connector be melted, they are made for long term use and still melted regularly just the same --> Thousands of examples.

Thousands (maybe) out of how many million ATX power supplies on the market? 17 million PCs (including laptops) were sold in the second quarter of this year in the United States alone. By my estimate (a combination of UN/ITU data per-capita and CIA World Factbook population estimate, since the UN site seems to be down) there were around 229 million PCs in service in the United States in 2004. That would mean, and I'm being very generous here, a 0.0052% rate of incidence.

No one is saying this type of failure never happens (in fact I said the opposite of that), but if you are going to claim that the ATX connector melts "regularly" you might want to provide some actual evidence of that. There is a huge gulf between "happens once in a great while" and "happens regularly". Like 3 orders of magnitude.

Of course they would, EVGA expected the adapters to have problem in long term use, including melting. Idiots will ignore it as usual.

And how would you demonstrate that EVGA expects this, or even that the incidence of such failure is high enough for you to brand people "idiots"?

 

[beray]

Originally posted by: Aluvus
And? You are still using the terms incorrectly and have yet to explain what PFC actually has to do with anything being discussed.

"It is, as long as the output impedance is so small that it can be ignored. As the output impedance goes up, it can't be ignored.

ActivePFC basic function is matching the PSU input impedance to the output power source in use for maximum power transfer. AKA the highest possible power factor."

Thousands (maybe) out of how many million ATX power supplies on the market? 17 million PCs (including laptops) were sold in the second quarter of this year in the United States alone. By my estimate (a combination of UN/ITU data per-capita and CIA World Factbook population estimate, since the UN site seems to be down) there were around 229 million PCs in service in the United States in 2004. That would mean, and I'm being very generous here, a 0.0052% rate of incidence.

Don't worry, the burn up connector numbers will get higher and higher when 750Ws and higher PSUs are even more common.

No one is saying this type of failure never happens (in fact I said the opposite of that), but if you are going to claim that the ATX connector melts "regularly" you might want to provide some actual evidence of that. There is a huge gulf between "happens once in a great while" and "happens regularly". Like 3 orders of magnitude.

For old low power PSUs not counted on to be melted down at all "Thousands of examples" from googling alone is very regularly.

And how would you demonstrate that EVGA expects this, or even that the incidence of such failure is high enough for you to brand people "idiots"?

"EVGA sends a warning in the box that the adapter is to be used short-term only" <-- Hey!! EVGA knew whats going on with higher power demands.

Where is yours which EVGA din't expect this?

 

[MrOblivious]

Originally posted by: Ocguy31
You could obviously say im not a power supply expert....

What are the dangers in using the adapter that takes 2 of my 6 pins and makes them a single 8? I just installed them on my new GTX280. EVGA sends a warning in the box that the adapter is to be used short-term only.

Its a cover their ass issue. If your power supply does not have sufficient 6 pin or 8 pin connectors than it may not be sufficient for powering the card (ever notice the inflated minimum specifications for powering videocards? Same deal). Additionally, adapters can add some resistance with all of its resultant issues but this is generally negligible so long as you are not "daisy chaining" connectors.

 

[beray]

Daisy chaining.

Let's see, one adapter made it 0.03O - 8W, three adapters should be 0.06O - 17W, five adapters should be 0.12O - 27W for the same 200W GPU.

Yes, 5 adapters should make long term into something much shorter. Though that usually wouldn't be having the spectacular oxidation effects when one adapter degraded to the point where it had to handle 27W instead of 5 adapters.

 

[JEDIYoda]

Originally posted by: beray

Originally posted by: Ocguy31
You could obviously say im not a power supply expert....

What are the dangers in using the adapter that takes 2 of my 6 pins and makes them a single 8? I just installed them on my new GTX280. EVGA sends a warning in the box that the adapter is to be used short-term only.

The basic inherent problem is power factor.

Let's say your GPU used 200W from 12V, using the 8-pin adapter boosted the 12V power connection to a total final impedance of 0.03O. Your PSU would have to put out the real power of 208W for the 200W apparent power used by your GPU.

In the short term 8W is nothing as long as your PSU could handle the job, especially if the PSU is really powerful.

In the long term, the 8W real/apparent power difference would cause accelerated oxidation at the power connectors, ---------have you ever seen oxidation on the power connector? or do you believe all the FUD you read on the PC Power & Cooling Myth`s page??the PSU would put out even greater real power to compensate. --> Your power connectors would melt or catch on fire and they nearly always do.

There are other problems but they are minor in comparison.

 

[Modelworks]

My biggest concern with doing this would be the source of the 12v in the supply itself.
In a supply that uses two different rails for those 6 pin connectors, they might not like a device "mixing" 12V from two different rails, especially if they were truly independent inside the supply. Also depends on how the video card converts the 12V to 3.3.,2.1, or whatever at the connector.

 

[beray]

Originally posted by: JEDIYoda
have you ever seen oxidation on the power connector? or do you believe all the FUD you read on the PC Power & Cooling Myth`s page??

You don't read English very well, please go and pick up another $30 worth of English skill.

I don't brainlessly aped "mumbo jumbo" from "reliable sources", you are confusing me for yourself and other experts. For example....

Your own faithfully believable expert learned "crossloading" by paying for and reading SMPS text books, I learned "crossloading" by reading free advertisement like this one --> "crossloading of a non-group/non-independently regulated PSU". AKA a true single-rail PSU.

 

[VirtualLarry]

Originally posted by: beray
Your own faithfully believable expert learned "crossloading" by paying for and reading SMPS text books, I learned "crossloading" by reading free advertisement like this one --> "crossloading of a non-group/non-independently regulated PSU". AKA a true single-rail PSU.

So let me get this straight. You are touting your knowledge as superior, because you learned it from an advertisement, rather than a textbook???

You might want to crack open some textbooks, then you would learn that PF isn't even an issue with DC circuits, because it is ALWAYS 1.0.

 

[beray]

Originally posted by: VirtualLarry
So let me get this straight. You are touting your knowledge as superior, because you learned it from an advertisement, rather than a textbook???

Reading and comprehending the proper "crossloading" advertisement would not let you do something stupid like this below...

"Here we've 2 independently regulated power rails, we'll heavily loaded down one independently regulated power rail but not the other independently regulated power rail and vice-versa then check for crossloading data.

See there, everything worked perfectly in specs, no crossloading problem.

And there had never been any independently regulated crossloading problem for the last 3 years in all of our hundreds of tests."



Here's a tip --> If it crossloaded it's not independently regulated.

Originally posted by: VirtualLarry
You might want to crack open some textbooks, then you would learn that PF isn't even an issue with DC circuits, because it is ALWAYS 1.0.

You should learn to recognize that ideal DC circuits had the PF = 1.0, non-ideal DC circuits had the PF < 1.0

Learn power transfer properly DC or AC.

 

[VirtualLarry]

Originally posted by: beray
non-ideal DC circuits had the PF < 1.0

Yes. We commonly call those "AC".

 

[beray]

Long-term use ATX connectors had burned and melted down for thousands of times, but high power short-term use 8-pin adapters would never have long-term problem.

Because all "experts" and "reliable sources" with SMPS text books said so.

 

[JEDIYoda]

Originally posted by: beray
Long-term use ATX connectors had burned and melted down for thousands of times, but high power short-term use 8-pin adapters would never have long-term problem.

Because all "experts" ---you mean every expert in the world? Please name a few...thx and "reliable sources" with SMPS text books said so.--- please name the reliable sources...thx or do you have a clue?