• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Push-pull and more, Xigmatek S1283

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

pejsaboy

Member
Joined
May 1, 2005
Location
nor-cal
In another thread, Mustanley inquired whether or not this 'sink would benefit from a push-pull setup with 120x38mm fans. I told him I would do some testing when I picked up a second Ultra Kaze fan. This thread is for the results of that testing. I'm still putting the new test subject together, but hopefully I'll have some numbers within a few days.
 
Last edited by a moderator:
Round one of testing consisted of a single Ultra Kaze pushing through the 'sink, with the cpu at its stock speed of 2.66Ghz. Idle temps across the cores:
36 33 34 34
Load temps:
44 38 41 39

Round two of testing consisted of two Ultra Kazes, the original pushing, and the secondary pulling through the 'sink. Again, stock speed of 2.66ghz. It looks like the second fan helps out by about 1-2 degrees.
Idle temps:
36 33 34 34
Load temps:
42 37 40 39

Round three of testing consisted of two Ultra Kazes, both pushing air into the middle of the 'sink. As before, stock speed of 2.66ghz. I stopped the test prematurely because, as you can see by the load temps, the push-push configuration is counter-productive even to a single fan pushing through.
Idle temps:
36 33 34 34
Load temps:
45 40 43 41

The last rounds of testing were done at 3.6ghz with 1.328 vcore- which is [I think] a ton more than what I need to be stable. I'll be backing that down and re-testing for stability as I get more time. The idle temps for both push and push-pull configurations were the same as stock speeds, so I can pretty well say that my sensors are stuck at those points.
3.6ghz push, Idle temps:
36 33 34 34
Load temps:
56 52 54 52

3.6ghz Push-pull, Idle temps:
36 33 34 34
Load temps:
55 51 52 50

So, it looks like a push-pull configuration is still good for about 1-2 degrees with my particular choice of fans.
 
Last edited:
I plan to do the same, it's just not at the top of my priority list yet. Looking forward to your results.
 
Testing will be done using Prime95 small FFT, and recording the highest temps as reported by RealTemp 3.00 over the course of a 10 minute session. The Ultra Kaze fan will be run at full power, and whatever case fans I happen to have running right now will run whatever speed they're set to on my fan controller. I won't be touching them any time soon :)

Round one of testing has been recorded in post #3, and I hope to have round 2 sometime tomorrow. When all testing is finished, I'll try to work up a nicer looking chart and/or graph to consolidate the information.

EDIT: I forgot to mention that vcore at stock speed is 1.164v idle via cpu-z, and 1.152 under load. Vdroop of .012v, the least of any board I've owned! I'm running the push-pull test now, and should also get to the push-push test this morning. I've decided against the pull-pull as I don't feel like messing with the little rubber fan mounts on the primary fan. Also to note, I've worked up a bolt through mount for this 'sink as opposed to the push pins that it comes with. It's much sturdier, and improved the temps on my e2160 by 2-3 degrees.
 
Last edited:
The testing at default core clock and vcore are complete. It appears that a push-pull setup is only slightly better than a single fan, and probably not worth the cost of an extra fan and shipping. The push-push setup is completely worthless, and I won't be testing it at higher clock speeds. Once I get all my normal programs and games installed and get everything to my liking, I'll work on an overclock to see if the push-pull setup comes into its own with a heavier load. It will be at least a couple days, possibly next weekend before I'll have any more results.
 
Sure thing. I'm testing 3.4Ghz right now, and am sitting pretty good as far as temps and voltage goes. Only a couple degrees warmer than stock, actually. I'm hoping to get 3.6 prime stable for a couple hours before the next round of testing, but if it takes too much tweaking I might just test at whatever I can get stable enough easily enough.
 
Man, the post thing isn't the only thing weird happening with the forums today. I went and tried to delete the original post and learned something: if you delete the original post the whole thread gets deleted. That wasn't too weird and I learned something new, but what happened next was just strange. I copied your second post into the original one, then went to delete the second post and the whole thread deleted again.:screwy: After undeleting the thread for the second time, I was then able to delete the second post from the thread to clean things up.:beer:

On topic now; I am looking forward to seeing your results with the experiment, pejsaboy. :thup:
 
Yes nice results, I have a S1283 and have straped a YateLoon medium to it for now. Since my case is BTX(upside down and backwards) I can just set an extra fan in there on the other side of the Xiggy, and I have only noticed about 1deg C difference when doing that with 2 YL's.

I was wondering if you have any lesser fan(like the stock one) to compare temps with to show how good the Kaze fan is. I have been eying one for myself, just not sure if its worth it or if the 3000rpm version is TOO loud?

I wasnt worried about it when I had my E8400 in here as it would cool that easily even at 4.4Ghz, but this new Q9650 is a beast so I thought I would get a better fan.
 
Last edited:
I don't have the stock one available anymore, as it's now a case fan in the box I passed down to my wife. However, I can tell you that I dropped between 7-10 degrees just by changing to the Ultra Kaze. To be perfectly honest, I don't mind the noise of the 3000 rpm fan at full tilt, but I came from my original custom built computer having about six 80mm fans that were quite loud. The 110cfm slipstream I bought however, is a bit too much for me on full speed though. Luckily it's an intake case fan, so I run it slower. Realistically, you could run the 2000 rpm version of the Ultra Kaze and probably have the same performance level with this particular 'sink. I noticed that the 3k doesn't really do much else for the 'sink above what I estimate to be about 6-7 volts on my fan controller.

Also, thanks for getting the thread cleaned up mud, much appreciated :thup: I'll get you all some more results as soon as I can get 3.6ghz. I've decided that HAS to be the testing point. Just gotta figure out some settings :bang head
 
.

got this cooler a few days ago, and its great but need to replace current fan! it runs at 1200 rpm? and i think it is only around 50 cfm! ordered some generic 2100rpm/75cfm 120mm fans - hope they make a difference!
 
Yeah, you'll definitely see improvement by replacing the fan with a faster/higher cfm one.

Anyway, I've been pretty busy the last few days but I got 3.6ghz stable. Was actually working on 3.8, but I didn't want to push too much vcore just yet- brand new $200+ cpu and all ;) I'm hoping to have some more results for you all later this afternoon or evening though. I can't make any promises, but I"m trying!
 
got some higher cfm fans: ( when i say load i mean running set )

fan @ 1200rpm load temps 52,51,48,48 ( same speed as xiggy fan )
fan @ 1900rpm load temps 49,49,45,43 ( similar cheapo fan frm ebay )
 
idle temps:

1200rpm = 33,34,28,28
1900rpm = 33,33,27,27

wow a fan with higher cfm makes a difference of quite a few degrees, especially at load temps!
 
I'm surprised you didn't try a pulling fan. I might make a miniscule difference, but if it does help and little and you only have one fan you might want to go with pull. Pulling would allow more air to enter directly onto the fins instead of being blocked by the fan and it would also pump it directly out of the case where as a push might push it right into the case circulation.
 
The final results are in! check back to post number 2 for the latest numbers.

ChanceCoats- There are 2 reasons I didn't try a single pull setup. First, I didn't want to mess with the rubber fan mounts on the 'sink with the whole thing in my case. Also why I didn't do the pull-pull setup. Second, the rear case fan actually intakes for me as opposed to most people's [I assume] being exhaust. I didn't want to come up with a way to ghetto band the fan on the opposite side of the 'sink. I barely had the right sized "supports" to hold a second fan on. I needed the overall size of 2 fans and the 'sink to get it done :)

Anyhow, hope you all enjoy the results. As it goes, my second ultra kaze will probably replace the slipstream as rear intake, and the slipstream will be tastefully mounted to help cool my 4870. For me, 1 or 2 degrees isn't worth the extra fan.
 
Back