Re: 100% Fanless midrange system - is this possible?
Posted: Tue Oct 23, 2012 10:16 am
Thank you for the comments!
I actually tried to buy an extra sett with six heatpipes, like the one I have for the PCU, from Streacom. At first they were positive and I also got a price, but after a while I did not hear anymore from them. Maybe they got cold feet and are afraid to get complaints if someone fries their components after pushing it too far. I guess they have compared the risk and cost of complaints up with the safe way advertising the chassis just for use that they know will work safely. Everyone may not use the thermal paste the best way. But I still think they could have taken the chance to recommend lower end graphic cards, like GTX 650. They could just have set the same recommended max tdp of 65-95Watts.
Anyway I was content using the HFX heatpipes, because they are more adapteble. It is easy to mount several blocks with four heatpipes on top of each other. It's just to use a longer screw and that follows the package of the extension set. I am planning to order a extra set of the VGA cooler and use on the VRM. Other components on the graphic card is cool, like the ram. The only other component that gets hot to touch is the northbridge on the MB. I allready have a Zalman ZM-NB47J, but yet the temp don't seem critical. The RAM on the MB also seem to do fine, eaven if it's low profile. The CPU holds very low temps and the heatsink seldom feels warm (Right now it almost feels cold so it can't be over 37 C).
I use the GPU-Z from TechPowerUp to monitor the temp on the GPU. It is the first time I use this, with the HFX classic case I always felt safe. I also wondered what the 50% TDP meant. I also guess it must mean the card only goes on half load. I have understood from reviews of graphic cards that they seldom or never uses 100% load. For example the GTX 670 seldom uses more than about 160W and the load for GTX 660Ti is normally in the mid 130s. So maybe 50% of TDP means more than 50% of what the card practically can reach, but I am not shure. I have focused more on the use on games and the important thing for me is the possibility to play the games I want for some hours. The testing I have done is with the cargames Grid and Need for speed Shift. The games where on the high setting as far as I could understand. There is a new Need for speed game coming out soon and then I can test out with that one. The testing I talked about earlier I played Grid for some hours and the temp seemed to round of about 65 Celsius and kept around this level. The cooling system seem to respond fast. After the heavy gaming and if I go on surfing the temp goes rather fast down to under 50 C but then it take some time to go under 40C. If I went for another long period of gaming soon after this maybe the temp would go further up. I will know after some time. For now it seem to work fine, except VRM.
I should also mention the heatsinks do get warm, espessialy after long gaming. It feels too hot to touch for a long time, but not like I get burned. I can understand why Streacom would not recommend this kind of load. The front plate also gets warm but here I can hold the hand as long as I want. After long gaming the VRM is very hot, like burning, so I think something must be done, like the VGA cooling set I mentioned. I am also not shure if the GPU-Z program monitors the heat of the VRM or just the graphic processor itself.
I agree with the consern of the extra heat from an internal PSU. I saw a review of the Seasonic fanless 460 from 2010. He said the room temp was 25 C and that the extra temp from the PSU was 10 C at 80% load and 13 C at 110% load. I hope the Seasonic 520 will be eaven more efficient since it's a Platinum, so hopefully the extra temp will be minimal and I don't think the load will be more than 80% from this kind of system. After some reviews we will know more. One could maybe try to isolate the PSU from the rest of the case. The holes on the top could be like the ones on the FC5, if it gives enough air that would not compromise the design too mutch. I drilled about 700 small holes extra in the FC5 case (took about 4 hours).
At the moment I feel the PC works fine, except for the VRM, and if I can play some games on full settings now and then this works for me. I can always monitor temp while gaming.
I actually tried to buy an extra sett with six heatpipes, like the one I have for the PCU, from Streacom. At first they were positive and I also got a price, but after a while I did not hear anymore from them. Maybe they got cold feet and are afraid to get complaints if someone fries their components after pushing it too far. I guess they have compared the risk and cost of complaints up with the safe way advertising the chassis just for use that they know will work safely. Everyone may not use the thermal paste the best way. But I still think they could have taken the chance to recommend lower end graphic cards, like GTX 650. They could just have set the same recommended max tdp of 65-95Watts.
Anyway I was content using the HFX heatpipes, because they are more adapteble. It is easy to mount several blocks with four heatpipes on top of each other. It's just to use a longer screw and that follows the package of the extension set. I am planning to order a extra set of the VGA cooler and use on the VRM. Other components on the graphic card is cool, like the ram. The only other component that gets hot to touch is the northbridge on the MB. I allready have a Zalman ZM-NB47J, but yet the temp don't seem critical. The RAM on the MB also seem to do fine, eaven if it's low profile. The CPU holds very low temps and the heatsink seldom feels warm (Right now it almost feels cold so it can't be over 37 C).
I use the GPU-Z from TechPowerUp to monitor the temp on the GPU. It is the first time I use this, with the HFX classic case I always felt safe. I also wondered what the 50% TDP meant. I also guess it must mean the card only goes on half load. I have understood from reviews of graphic cards that they seldom or never uses 100% load. For example the GTX 670 seldom uses more than about 160W and the load for GTX 660Ti is normally in the mid 130s. So maybe 50% of TDP means more than 50% of what the card practically can reach, but I am not shure. I have focused more on the use on games and the important thing for me is the possibility to play the games I want for some hours. The testing I have done is with the cargames Grid and Need for speed Shift. The games where on the high setting as far as I could understand. There is a new Need for speed game coming out soon and then I can test out with that one. The testing I talked about earlier I played Grid for some hours and the temp seemed to round of about 65 Celsius and kept around this level. The cooling system seem to respond fast. After the heavy gaming and if I go on surfing the temp goes rather fast down to under 50 C but then it take some time to go under 40C. If I went for another long period of gaming soon after this maybe the temp would go further up. I will know after some time. For now it seem to work fine, except VRM.
I should also mention the heatsinks do get warm, espessialy after long gaming. It feels too hot to touch for a long time, but not like I get burned. I can understand why Streacom would not recommend this kind of load. The front plate also gets warm but here I can hold the hand as long as I want. After long gaming the VRM is very hot, like burning, so I think something must be done, like the VGA cooling set I mentioned. I am also not shure if the GPU-Z program monitors the heat of the VRM or just the graphic processor itself.
I agree with the consern of the extra heat from an internal PSU. I saw a review of the Seasonic fanless 460 from 2010. He said the room temp was 25 C and that the extra temp from the PSU was 10 C at 80% load and 13 C at 110% load. I hope the Seasonic 520 will be eaven more efficient since it's a Platinum, so hopefully the extra temp will be minimal and I don't think the load will be more than 80% from this kind of system. After some reviews we will know more. One could maybe try to isolate the PSU from the rest of the case. The holes on the top could be like the ones on the FC5, if it gives enough air that would not compromise the design too mutch. I drilled about 700 small holes extra in the FC5 case (took about 4 hours).
At the moment I feel the PC works fine, except for the VRM, and if I can play some games on full settings now and then this works for me. I can always monitor temp while gaming.