Why BFG custom coolers "suck" ?

[ozziegn]

Make sense i agree. But wtf is up with the NV5 Silencer using the same HS for core and memory?

apparently you can be an effective cooler with this problem.

well yes but you have to take into account the much bigger fan on that cooler vs. the two teeny-wheeny fans thats on the BFG (copper) cooler which like I said before, they hardly move any air at all.

it seems like BFG wanted those fans to "oooh and ahhh" any potential (and future) sales but it seems like they couldve cared less if they actually moved any air at all.

just as long as they looked nice and purdy.

 

[Badger_sly]

.........
(boy, you just gotta love the tons of people on computer tech hardware forums who get offended sooooo easily when someone says something negative about a piece of computer hard that they may own...)

....... or other who get offended sooooo easily when someone contradict what they may say.

 

[UltimateMan]

well yes but you have to take into account the much bigger fan on that cooler vs. the two teeny-wheeny fans thats on the BFG (copper) cooler which like I said before, they hardly move any air at all.

it seems like BFG wanted those fans to "oooh and ahhh" any potential (and future) sales but it seems like they couldve cared less if they actually moved any air at all.

just as long as they looked nice and purdy.

but apparently any of that stuff i said about thermal transfer doesn't matter?

 

[Dragonetti]

My unmodded PNY does 425/1150 no sweat 24/7 . All I had to do is adjust some sliders..
I have to agree that BFG should have let well enough alone.

 

[saintrobyn]

heh, you'll certainly get no arguement(s) from me there.

its a PITA for someone like myself to do that whole BIOS / floppy disk thingy considering my Shuttle SFF doesnt have a FDD so I have to take the cover on/off everytime I need to use a FDD.

I have read that the X800 Pro cards built by ATI tend to have the cores that truely did not make it as an X800 XT. I just can't remember where I read it.

As for your Shuttle dilema, I completely understand. Which one are you using, I have the SK41G and it doen't have a Floppy in it either. It's my HTPC, it doesn't need a floppy, I used the bay to put in a second Hard Drive for recording TV.

 

[saintrobyn]

My unmodded PNY does 425/1150 no sweat 24/7 . All I had to do is adjust some sliders..
I have to agree that BFG should have let well enough alone.

Hey no fair Dragonetti, I already used that sliders line. Get some new material.

All kidding aside, are you using rivatuner or coolbits to overclock your card? I can only get 1050 memory using coolbits and I'm wondering if I would get better results with rivatuner.

 

[perfektrtw]

My PNY is running pretty damn hot. I think I need a new case, my whole cooling situation sucks. But at the moment I have the side of the case off and my GT is at 60-61c Idle. When the case was closed it was idling 65-66c. This is at stock speeds. How hot can you take these things without damaging them? I don't know why mine is running so hot. I would love to see how high I can overclock it, but I am concerned about the tempatures... even at stock speeds. Any comments?

 

[Darth_Fluffy]

gee, whats the matter Skippy?

you wouldnt happy to own one of those bad-@ss, mama-jama BFG cards and maybe have gotten your feelings hurt by me not giving it a two thumbs up, ehh?

(boy, you just gotta love the tons of people on computer tech hardware forums who get offended sooooo easily when someone says something negative about a piece of computer hard that they may own...)

I don't even have a a 6800 of any flavor to begin with, let alone a BFG one. But with the words you use, you really sound like a hick

 

[Taurus]

That actually isn't tue and i will explain but first observe: the NV5 Silencer using the same HS for core and memory?

apparently you can be an effective cooler with this problem.

and basic science it may not be. If you really go into the details (which i can only skim really) you can then begin to talk about:

how the copper heatsink will act to distribute the heat evenly throughout the entirety of the metal's mass. But no matter what, the copper will not be AS hot as the GPU. In fact, the copper itself, will not be as hot as the Ram either. Therefore, it would still under those circumstances be drawing heat AWAY from the ram.

If not, then the ram WOULD burn out, rather quickly. Because of the thermal properties of the copper as opposed to the ram, it will always be cooler than the memory (with some fans in the area.)

so it wouldn't transfer heat to the ram after all.

what WOULD happen, since the heatsink touches both the GPU and RAM is that since the copper touching the ram is also being heated by the GPU, the copper is naturally warmer than copper that is, say, not touching the GPU. Therefore, it does not draw heat away from the memory AS QUICKLY as it would otherwise.

and before you debunk the "copper heatsink is cooler than the ram" theory, think about this. If it weren't, the ram would fry, because you would not only be NOT coolling it, you would be heating up an already hot piece of memory. the copper is cooler than the ram.

which method is MORE EFFECTIVE at cooling the ram and GPU? well, all other factors being equal, the two separate pieces of copper cooling the ram and GPU individually.

but no heat is being transfered TO the ram by the heatsink.

YES it is TRUE

Listen, No matter what you do your STILL gonna have a tiny whiny little bit of heat transfer to the ram!
EVEn though the ram will be cooled, it will still not be cooled the best that it could be. When you have a solid piece of metal touching the gpu and the memory, the Hooter of the two(GPU) will infact transfer some heat throughout the SOLID piece of metal and it eventually will reach the memory. Even though the memory will still be cooled, but it will not be cooled effectivley.

Always go with this : Cooling Notice Seperate Ram Sink

rather then this: not as good cooling

 

[Badger_sly]

YES it is TRUE

Not necessarily.

Listen, No matter what you do your STILL gonna have a tiny whiny little bit of heat transfer to the ram.......

The fact of the matter is that heat may or may not transfer from the gpu to the memory. Whether it does or not will depend on the distance from the gpu to the memory modules, and, if the fans push enough air to disapate the heat before spreading that far.

Here's a goofy example:
We take a long 6 foot piece of electrical fence wire and wrap one end around your hand, and the other end around the electric burner of a stove top, turn a heat adjuster knob to medium, then watch the burner get slightly red hot. Heat will travel from the burner into the wire, then travel down the wire. Now, if that wire was only 1 foot long, you'd feel the heat in your hand pretty quick, but since it's 6 feet long, it would take perhaps a few minutes for the heat to travel that distance before your hand starts getting warm. The normal air temperature will absorb some of the heat from the wire, how much depending on the amount of wire exposed. Then, with the wire getting hotter and almost burning your hand, if we take a 1 foot diameter tabletop fan and blow it on the wire, you'd feel the heat reduce. If we added a second 1 foot fan, the heat would reduce even more. If we replaced the two 1 foot fans with a 3 foot barn fan, blowing on the wire, you'd still feel the heat reduced.
Adjusting the dial on the stove burner to high, to simulate OCing, more heat would travel the length of the wire to your hand. Turning the 2 fans or 1 big fan to a higher setting (video card adjusting to the temp detected) should reduce the heat.

No, that example isn't perfect, but it is similar to the heat from the gpu travelling, while the fans are removing the heat. Until someone can attach thermal probes down the length of the heatsink to measure the heat transfer, we won't know for sure if heat is reaching the memory, or, at what OC it starts to.

My theory is that the engineers at BFG were smart enough to design the heatsink/dual fan combo so that the gpu does not affect the memory....... to a certain point. Once a certain OC is reached (give or take 5 degrees for room temp) then the memory is affected, thus, hitting a cards max OC.

 

[Syphon Filter]

There is NO heat transfer from the HSF TOO the RAM. This is not what is happening. Check this out:

(figures are made up and heat source temps are constant just for clarity)

Situation 1: the heat source (aka the RAM) is at say 70*C and the heatsink on it is at 20*C.

Situation 2: the heat source (aka the RAM) is at say 70*C and the heatsink on it is at 60*C.

Situation 3: the heat source (aka the RAM) is at say 70*C and the heatsink on it is at 100*C.

In both situations 1 and 2 heat will transfer from the RAM to the heatsink overall, if you look at the deep physics of it then there will be energy moving from RAM to heasink and vice versa, BUT much more will be moving from RAM to heatsink thus keeping the RAM cool. The difference between situations one and two is that heat will be moving from RAM to heatsink much more quickly/effectively/efficiently in situation 1 due to the greater temperature differential.

In situation 3 (which is totally made up) heat/energy will be overall be moving from the heatsink to the RAM causing it to heat up, because that is the direction of the temperature gradient.

Having the RAM and GPU on seperate heatsinks lends to a situation for the RAM which is like situation 1. Having the GPU and RAM all on one heatsink leads to something more akin to situation two, because the heatsink is warmer due to contact with the GPU as well as the RAM.

The new BFG design is not warming the RAM up. It is just not cooling it as effectively, there is a difference. Its the temperature differentials that are important. If you wanna correct me then feel free, but make sure its not BS and get some opinions off ur physics teachers/prof's.

 

[Taurus]

Not necessarily.


The fact of the matter is that heat may or may not transfer from the gpu to the memory. Whether it does or not will depend on the distance from the gpu to the memory modules, and, if the fans push enough air to disapate the heat before spreading that far.

Here's a goofy example:
We take a long 6 foot piece of electrical fence wire and wrap one end around your hand, and the other end around the electric burner of a stove top, turn a heat adjuster knob to medium, then watch the burner get slightly red hot. Heat will travel from the burner into the wire, then travel down the wire. Now, if that wire was only 1 foot long, you'd feel the heat in your hand pretty quick, but since it's 6 feet long, it would take perhaps a few minutes for the heat to travel that distance before your hand starts getting warm. The normal air temperature will absorb some of the heat from the wire, how much depending on the amount of wire exposed. Then, with the wire getting hotter and almost burning your hand, if we take a 1 foot diameter tabletop fan and blow it on the wire, you'd feel the heat reduce. If we added a second 1 foot fan, the heat would reduce even more. If we replaced the two 1 foot fans with a 3 foot barn fan, blowing on the wire, you'd still feel the heat reduced.
Adjusting the dial on the stove burner to high, to simulate OCing, more heat would travel the length of the wire to your hand. Turning the 2 fans or 1 big fan to a higher setting (video card adjusting to the temp detected) should reduce the heat.

No, that example isn't perfect, but it is similar to the heat from the gpu travelling, while the fans are removing the heat. Until someone can attach thermal probes down the length of the heatsink to measure the heat transfer, we won't know for sure if heat is reaching the memory, or, at what OC it starts to.

My theory is that the engineers at BFG were smart enough to design the heatsink/dual fan combo so that the gpu does not affect the memory....... to a certain point. Once a certain OC is reached (give or take 5 degrees for room temp) then the memory is affected, thus, hitting a cards max OC.

Your Right, partly

Now take that wire and cut it in TWO, but still have that one end touching the burner and the other part wraped around your hand. THE HEAT WILL NOW not reach your HAND RIGHT? IT WILL SIMPLY not reach it. Thats what Im trying to say here! Now if you have two seperat cooling soultions you will genereate less hear.

I have tried this experiment in high school electronics class. I took two seperate pices of metal put a tourch underneath each peice and then i measured the temp. I then took BOTH pieces and WELDED then together(METAL SHOP CLASS) then again I put the TOURCHES underneat he metal in the exact same spots ,but this time with the metal welded toghther(remember). And the temp diffrence I got was 21C (not sure what that is for American Folks).
Therfore No matter what cooling solution you have its still gonna run hotter then two seperate pieces with the exact cooling solution.

The best Cooling solution is two have seperate heatsinks on the ram and seperate heatsinks on the gpu WITH a fan blowing air onto both! This is what I have! (http://www.thetechlounge.com/review.php?directory=zalman_vga_cooler&page=3)

I have also made custome ramsinks for it

 

[Taurus]

is just not cooling it as effectively, there is a difference. .

There it is the money shot. BINGO. Thats what im saying

THERFORE it is BETTER to have seperate cooling solutions for the gpu and ram

Ive been saying that forever LOL

BFG cooling is good, but its not better then redisigning it and having two seperate soultions for ram and gpu. That would be the best case senerio.

 

[Badger_sly]

........
.........
BFG cooling is good, but its not better then redisigning it and having two seperate soultions for ram and gpu. ...............

Maybe not better than seperate pieces for the gpu and memory, but also maybe not worse, if, as I stated before, the heat from the gpu isn't reaching to the memory. Everyone is assuming that because BFG's cooler is one piece, that that gpu is making the whole thing hot all the time.