He's not measuring molecules...and rust has nothing to do with the instrument's precision.Quote:
Originally Posted by Iconyu
Printable View
He's not measuring molecules...and rust has nothing to do with the instrument's precision.Quote:
Originally Posted by Iconyu
the divets or whatever they are called that hold the tip to the tape, are clearly able to move by about 1mm
its damn well close enough to 2" i think everyone can calm down
First: To be even remotely accurate the tongue of the tape measure would have to be OVER the edge of the radiator, not just pressed into the gap of the curved edge and the fan.Quote:
Originally Posted by Baron_Davis
Second: The cut at the start of the tape is curved.
Third: There is obvious play between the rivets holding the tongue to the tape.
All this before we get to 30mm, sure the accuracy might improve a whole lot once you get to the 50cm mark but at the 50mm point it's not trustworthy.
I just adjusted the data and reinput the variables into the quantum over drive. The H70 is in fact:Quote:
Originally Posted by Iconyu
48.019384765999947401017mm thick give or take a micron.
The world can breathe a sigh of relief...
Keep up the good work son.Quote:
Originally Posted by Baron_Davis
Dude that was a review quote. The rad thickness is 48mm.Quote:
Originally Posted by dengyong
Has anyone got some pics of the H70 mounted in a case? :)
Im thinking about one of these but am worried about the thickness of the rad. 50mm, or 48mm whatever, seems really thick!! i like having space around the CPU area so that other components can get some air flow
http://forum.aquapc.com/showthread.php?t=93013Quote:
Originally Posted by a5h
http://img121.imageshack.us/img121/9135/28665297.jpg
I see now. :up:Quote:
Originally Posted by Biker
I've been testing this H70 . First of all i set the fans as intake . This heats up my entire case and this adds to the noise of my videocards (they start to scream with their tiny fans) .
So I have mounted the fans now to be outtakes , this is also most fair to compare it to my previous cooler (thor's hammer) which also had to cool the cpu with air from inside my case .
fans used on the rad are :
push papst 4412 GP (100cfm 2900rpm max , PWM controlled)
pull scyte slipstream 1600rpm (88cfm max , always running 100%)
room temp 24°C
thuban X6 cpu 4216mhz , 1.48 idle 1.52 load
idle temp 37° , with firefox open this is 40° . all in all 1° better than my hammer .
load temp during boinc (100% cores , 75% load) 57° . during prime goes to 70° (pwm fan at 2700rpm) then i had to stop (max cpu temp for thuban is 62°C)
seems like this cooler can't handle the cpu's load .
good points :
lower internal casetemp/motherboard/NB/SB/videocards and better airflow
somewhat less noise due to the fact that i could drop 1 fan with this setup
very clean case with lots of room
Given the fact that it is a watercooling setup and costed me 96€ , I am fairly disappointed with the unit as of now .
Thanks for the information CrimInalA..but to compare..how high was the load temp from your thor's hammer!
And what where the temps when you use it with fans as intake?
JP.
Thanks! the rad looks massive compared to the H50Quote:
Originally Posted by dr.noob
the hammer couldn't take prime either at these speeds and voltages .Quote:
Thanks for the information CrimInalA..but to compare..how high was the load temp from your thor's hammer!
And what where the temps when you use it with fans as intake?
JP.
But it's a bit apples to peaches here , because on the hammer i had a papst 4412GLL (44cfm 1200rpm) compared to the slipstream (88cfm 1600rpm) which i have now installed on the rad .
the temps when i had the fans as intake on the rad weren't much different (37° idle , prime : not able to handle) . But at that point i had the noctua NF-P12 as push and papst 4412GP as pull . however , my GPU idle temps were +8° , NB +9° , SB +5° , MB +3° .
According to the reviews of this cooler, it does no better than my Asus Triton 88.
that sounds like a bad mount. ive tested higher voltages with lower clocks and never broke 50C. i dont know if any high end cooler lets a amd chip ever reach past 60C due to how little power they consumeQuote:
Originally Posted by CrimInalA
Were those core temps? I didnt break 50c either with a 1090t and H50 at 4ghz 1.48v. 4.2ghz 1.6v load ~54c, this is cpu temp. not core temps. As core temp is bs.Quote:
Originally Posted by CrimInalA
While I've not had much experience overclocking AMD's recently, I do think Criminal had mounting issues. He couldn't tighten the screws without boot failure/problems. If that's still the case, then I'd imagine all his results are compromised.Quote:
Originally Posted by Manicdan
all temps are cpu temps , as core temps are irrelevant .
I redid my AS5 and mounted and tightened the screws .
As you can read above , the mounting issues were in fact bios issues caused by removing the cpu and reseating it .
http://img638.imageshack.us/img638/1734/omgmacprice.png
Hahaha, that's pretty funny! :rofl:
Sorry for OT. :p:
Thats why you should never test coolers on AMD. The sensors are so bad it screws up everything. Try using Everest to read the temps and small fft to load the cpu. Below you can see what a NH-D14 can do with a 1500 rpm Slipstream:
http://forum.lab501.ro/attachment.ph...4&d=1282052589
I always/only use everest to show temps on my G19 lcd . Other programs show the same temp , but everest is well implemented for the G19 so i'm going with that .
Temps for cpu are the same during blend or small fft (maybe +-1°C) .
Tell the guy in the screenshot to run a decent HTlink speed and NB speed :rolleyes:
You'll see temps go up by a fair amount .
it shouldnt be hard to know the accuracy of the temps. if you load 200W into an AMD or Intel using the same heatsink, they should have very similar temperatures. im not saying within 1C, due to shape of the chip and which parts heat up, things can vary by probably 3-4C. but a 10C differences comes off as an outside issue.Quote:
Originally Posted by halfwaythere
Unfortunately testing a Thors hammer and a Boss 2 couldn't handle the heat and the motherboard failed. I guess 4 phases for a serious mobo is a bit on the low side. As for Ht link and NB speed their influence was very low since he was trying to achieve higher cpu frequencies. Anyways testing heatsinks on AMD is a great exercise in futility since you don't really know which program to trust when reporting temperatures. Pushing over 1.5v into the cpu is a serious no-go if you don't own the very best in air cooling or some serious water.
The same TDP argument is not valid simply because theres no way of telling which program reports the accurate temperature on different platforms.
1.5v is a joke for AMD, turbo uses 1.45v AT STOCK
and you clearly didnt understand the theory about how you can translate wattage into heat and return an estimated temperature thats close enough to let you know which values to trust
and btw i dont see many issues with reporting AMD temps. OCCT, AOD, Everest, all show the same exact temps to me
For your reading pleasure:
http://forum.lab501.ro/showpost.php?...0&postcount=65
Romanian use translators.