Ryzen.5 1600 Fake Cooler and Fan Curve

gm-madi

Respectable
Nov 26, 2017
277
2
1,815
recently i got my ryzen build but looks like i got a fake Cooler, super noisy with 7 blades while in youtube it is very quiet with 5 blades. but looks loke someone in MSI Forum got the same as me without announcing any problem at all.

link to msi Thread: https://www.msi.com/blog/AMD-Ryzen-5-1600-Eight-Core-Revealed-B350-TOMAHAWK

1: is this cooler 'Different' or 'Fake', i mean can this be a problem?
2: am looking for a good Fan curve
( CPU @ 3.825Ghz at 1.300v 73° max test with Aida64 but 85°C test with Prime95 small FFTS )
3: so which stats i should consider, Aida's or Prime's ?

 
Solution
I just looked at one I have in a box and it has 7 blades and I am 100% sure it is not a fake. I don't use the OEM Ryzen heatsinks for that reason, they tend to run loud (they do cool well though).

I think you have an authentic product. There are variations of the Wraith coolers.

COLGeek

Cybernaut
Moderator
I just looked at one I have in a box and it has 7 blades and I am 100% sure it is not a fake. I don't use the OEM Ryzen heatsinks for that reason, they tend to run loud (they do cool well though).

I think you have an authentic product. There are variations of the Wraith coolers.
 
Solution

InvalidError

Titan
Moderator
Out of the four variants of the Wraith coolers, only the Wraith Spire without LEDs has a five-bladed fan, all others (including the red LED Spire variant) have seven-bladed fans.

You can't really gauge noise from YouTube videos, too many variables from case acoustics, microphone location and properties to all the processing from initial recording to eventual playback. People say they find Intel's stock HSF loud, I don't really notice mine.
 

gm-madi

Respectable
Nov 26, 2017
277
2
1,815
i made some google researchs and seems 2 companies were making AMD coolers, Corsair ( 5 blades , silent ) and he second one i forgot the name ( 7 blades, loud ).
my cooler do some job but double RMP of other people, i have to make it fan run at 65%speed at 50% CPU load and 85%speed at 100% CPU stress test to maintain 73°C. but its damn loud.
i just made this thread to check if other People got same problem of me, should i replace the heatsinks ?
 

gm-madi

Respectable
Nov 26, 2017
277
2
1,815


that is really wierd, how can it be different from what we see in youtube unboxing and benchmarking !!!
 

gm-madi

Respectable
Nov 26, 2017
277
2
1,815


mine is ryzen 1600- wraith spire with 7 blades, same in the MSI link i put above, but it is supposed to be 5 blades.
 
7 blades are delta , 5 blades are cooler master.
Doesn't seem to be differentiation on batching or production date.

I ahve 2 ryzen 1700 very close to release date coolers here.
One has a 5 blade , one has a 7 blade.

The 5 blade at similar rpm is substantially quieter - the 7 blade has a humming even at low rpm that isn't there on the other fan.
 

Karadjgne

Titan
Ambassador
The pic posted comes from AMD website. However, it's entirely possible that the 5blade design was dropped in favor of the 7blade design, since they are all the same fan, different heatsinks/shrouds. Manufacturing 1 fan to fit all is far cheaper and easier than multiple designs.
 
7 blades are delta , 5 blades are cooler master.
Doesn't seem to be differentiation on batching or production date.

I ahve 2 ryzen 1700 very close to release date coolers here.
One has a 5 blade , one has a 7 blade.

The 5 blade at similar rpm is substantially quieter - the 7 blade has a humming even at low rpm that isn't there on the other fan.

Edit - would appear that some more recent batches use avc fans too.

So there are actually 3 variations

P7mm9YY.jpg

 

Karadjgne

Titan
Ambassador
Fan curves are set in the bios, set by mobo software packages or by 3rd party software like SpeedFan. Once you fine tune the fan curves it'll change the sound patterns. My cpu fans max is 1500rpm. Little noisy. The curves are set to cap at 900rpm, but only reach that under torture testing. Generally they run 450rpm idle, 500-600rpm under load. Considering that any temp under @70°C is all the same to the cpu, loosing the 3°C under gaming loads and going from 52°C to 55°C, is of no consequence. Setting fan curves is just a balance of what temps you can live with vrs what pc volume you can live with. If you can't get that balance, time to bite the bullet and get a bigger cooler.
 

Karadjgne

Titan
Ambassador
Oh no, I only use p95 26.6 small fft. At 4.9GHz on kraken x61 I was getting 72°C after ½hr run, but since knocking back to 4.6GHz,I've significantly lowered fan rpm, basically silent now, even under gaming. Don't even hear the gpu since it's a fractal define R5, Asus 970. As I said, I can live with the temps, they are well under 70°C, so I have no need to try and max out cooling potential.
 

gm-madi

Respectable
Nov 26, 2017
277
2
1,815
i manually set the fan curve since i did some Overclocking ( 3.8Ghz at 1.3v ) the Fan run at 1400/1500rmp (35%) idle and 2100rmp (85%) at load test.
aida64 50min test 73°c max but Prime95 small fft temps go up to 83-85°c. this is why am asking for screen of fan curves so i can balance.
 

Karadjgne

Titan
Ambassador
Ok. Aida64 uses some linpack, basically it uses partial ram as well as the cpu. What this means for temps is that you don't get a 100% cpu load, you'll spike 100% but drop to @80% in stages and back up. So temps are not consistent, especially on a aio which takes longer to get hot due to liquid absorption rates compared to metal heatsinks.
Prime95 v26.6 and prior, small fft is 100% cpu load, and the most accurate as it's consistent, so after about ½hr test, AIO's have stabilized and you'll get an accurate bench of exactly what you can expect as maximum temps under any normal conditions. Any version of p95 after 26.6 uses AVX instructions (and a few others) which are specialized for some software. AVX in the large amounts used by p95 are closer to 130% cpu loads in comparison, so as is, that's an unreliable baseline for gaming, as games don't really use AVX at all. Newer bios take AVX into consideration and have an offset for OC, recommended is 2, which when the cpu detects its using AVX will drop the OC 200MHz, and a small amount of vcore, to compensate for the added heat. It's one of things sometimes overlooked and ppl freak out over extreme temps at lower OC than what they were looking for.