Slipstream 500 rpm as case fan?

Tzupy · Post by **Tzupy** » Sun Jul 06, 2008 4:34 am

I'm considering building a minimalist single-fan workstation, in a NSK 3480 with Zen 300 PSU and the Slipstream 500 rpm as case fan.
No other fans, low power components and the best heatsinks I can buy. Like HR-01+ for the ~45W CPU, HR-05 for the chipset, SQD for the HDD.
I'm not sure yet how much heat can the SSS 500 rpm exhaust and how much can the passive (convection cooling only) Zen 300 take.
I need some advice to decide if I should stick with integrated gfx, in which case I go for AMD CPU, or I could add a gfx and then I go for Intel CPU.
Thank you in advance for your inputs.

Luminair · Post by **Luminair** » Sun Jul 06, 2008 5:14 am

You will find that multiple very slow fans do much much better than one slow or even one faster fan. For instance, SPCR more often recommends a PSU with a silent fan, rather than one with no fan. No-fan is a very NON-fool-proof part. It is easy to screw that up.

I shouldn't comment on the fan choice because I don't have a 500rpm slipstream. That said, while I think the 500rpm would work, I think a better solution (again more fool-proof) would be to get the 800rpm or 1200rpm slipstream and hook it up to a fanmate2, and then choose which speed works best for YOU.

Only you can prevent forest fires, and only you can decide which graphics solution you need.

Tzupy · Post by **Tzupy** » Sun Jul 06, 2008 5:57 am

I know that multiple low speed fans do better then one higher speed one, but this is a SINGLE-FAN project, the Holy Grail here at SPCR.
Well, if the Zalman TNN300 would still be available, it would be a 0 fans project, but AFAIK it's out of stock everywhere.
I already thought of using the 800 rpm SSS and a fan controller, but the 800 rpm isn't available in my country, unless I buy a Ninja Copper just to get it.
I still consider this option, but I dislike the Ninja Copper, way too heavy and not the best tower cooler, so I wouldn't use it.
The 1200 rpm SSS has a minimum rpm of about 700 at 5V, which is much noisier than at 500.
The 500 rpm SSS should generate about the same noise as a WD GP in SQD, which would be inaudible for me from 1 m.
I could use a Nexus and fan controller for starters, and then, if everything turns OK at 5V, switch to the 500 rpm SSS.
But the reason I ask this in the forums is to spare myself time and a major dissapointment if I fail in my project.

dhanson865 · Post by **dhanson865** » Sun Jul 06, 2008 7:14 am

The Ninja2 SCNJ-2000 comes with a 1000 RPM slipstream and is about 700g.

SY1225SL12SL (500 rpm version)
SY1225SL12L (800 rpm version)
SY1225SL12LM (1,000 rpm version)
SY1225SL12M (1,200 rpm version)

As far as I know you won't be able to get the SY1225SL12LM as a stand alone fan though we can hope they will offer it as a stand alone fan in the future.

Of course it hasn't been reviewed so we can only guess the starting voltage on the 1000 RPM model.

You could always buy a Ninja2 or Ninja Copper, swap the fan, and resell the Ninja.

dhanson865 · Post by **dhanson865** » Sun Jul 06, 2008 7:26 am

Code: Select all

Optimal RPM for targeting <18 DBA
Slipstream SH 1900  5V <18dBA  400RPM
Slipstream  H 1600  5V  21dBA  890RPM
Slipstream  M 1200  4V <18dBA  620RPM
Slipstream LM 1000  ?v         ???RPM
Slipstream  L  800  5V <18dBA  420RPM
Slipstream SL  500 12V <18dBA  440RPM

Here is a crazy thought. You could get a Slipstream SH 1900 RPM and run it at 5v to get your super slow CFM and still have headroom to ramp it up later assuming you have a way to feed voltage other than the hardwired 5v/7v tricks as 6v would be your next step on fan that can ramp as much as the SH does.

and just for kicks here is the same table at slightly higher voltage/RPMs

Code: Select all

Optimal RPM for targeting <20 DBA
Slipstream SH 1900  7V  18dBA  630RPM
Slipstream  H 1600  5V  21dBA  890RPM
Slipstream  M 1200  5V  18dBA  720RPM
Slipstream LM 1000  ?v         ???RPM
Slipstream  L  800  9V  18dBA  640RPM
Slipstream SL  500 12V <18dBA  440RPM

I leave the 1600 H model in the chart just to show how bad of a fan it is for SPCR use but oddly enough the SH can do lower RPMs than the H.

seemingly.random · Post by **seemingly.random** » Sun Jul 06, 2008 7:43 am

Tzupy wrote:I know that multiple low speed fans do better then one higher speed one, but this is a SINGLE-FAN project, the Holy Grail here at SPCR.
Well, if the Zalman TNN300 would still be available, it would be a 0 fans project, but AFAIK it's out of stock everywhere.
I already thought of using the 800 rpm SSS and a fan controller, but the 800 rpm isn't available in my country, unless I buy a Ninja Copper just to get it.
I still consider this option, but I dislike the Ninja Copper, way too heavy and not the best tower cooler, so I wouldn't use it.
The 1200 rpm SSS has a minimum rpm of about 700 at 5V, which is much noisier than at 500.
The 500 rpm SSS should generate about the same noise as a WD GP in SQD, which would be inaudible for me from 1 m.
I could use a Nexus and fan controller for starters, and then, if everything turns OK at 5V, switch to the 500 rpm SSS.
But the reason I ask this in the forums is to spare myself time and a major disappointment if I fail in my project.

I would guess that the goal is not necessarily a single fan pc (or any other number for that matter) but a quiet, if not silent, pc. If you can't hear the rear fan, you're not going to be able to hear another fan attached to the hsf inside the case unless you have one of these cases built "from screen doors". Heat is _not_ your friend. Remember, one rear case fan is the second pc fan including the psu.

You also might reconsider on the 500rpm - it provides very little air flow. That 20cfm rpm measurement in every spcr fan review isn't there by chance. Maybe a slipstream 1200rpm 5 volted for 720rpm might actually provide some air flow. Another possibility is the 1600rpm sflex 5 volted for 580rpm.

The case design is a huge factor in cooling so not everyone's experience with individual fans will apply. I think you've already got the best idea - use a zalman fan controller on a nexus so things can be adjusted to your individual tastes. And, of course, the ambient temp plays a huge role. I can get by with half the cooling fan power in the winter as I can in summer.

---

The great thing about a mb with an igp is that it can be used without a gpu and if video speed is found to be insufficient, a gpu can be added later.

Tzupy · Post by **Tzupy** » Sun Jul 06, 2008 9:11 am

I appreciate your help, fellow SPCR dwellers, but my question remains mostly unanswered. So I'll rephrase it:
Assuming ALL components (CPU, GPU, chipset, HDD) will we properly cooled, how much heat can one 500 rpm SSS throw out of the case?

@seemingly.random: the 500 rpm SSS would be the single fan, since the PSU would be passive (Zen 300).
For the single-fan project the configuration that seems sensible to me is an undervolted and underclocked 4850e cooled by ducted HR-01+, on an nVidia 8200 mobo.

I have an alternative project, which would involve 3 (three) 500 rpm SSS fans: one for an E8400 with HR-01+, one for an 8800 GT with AC S1, and one as the case fan.
Which raises another question: does an 8800 GT with AC S1 fit in the NSK3480? From the SPCR review it seems they should by a narrow margin.
The PSU would be Enermax Modu82+525, but I've read that it's fan has a bad signature and the top vents of the NSK3480 would make it worse.

jaganath · Post by **jaganath** » Sun Jul 06, 2008 1:03 pm

Assuming ALL components (CPU, GPU, chipset, HDD) will we properly cooled, how much heat can one 500 rpm SSS throw out of the case?

you can calculate it from the heat transfer equation:

http://www.comairrotron.com/high_altitude_cooling.shtml

CFM = Q /(Cp * r * DT)

rearrange the equation then you can calculate how much heat (Q) can be moved for a given amount of CFM (20) and a given delta T.

http://www.comairrotron.com/engineering_notes_02.asp

seemingly.random · Post by **seemingly.random** » Sun Jul 06, 2008 2:44 pm

Tzupy wrote:I appreciate your help, fellow SPCR dwellers, but my question remains mostly unanswered. So I'll rephrase it:
Assuming ALL components (CPU, GPU, chipset, HDD) will we properly cooled, how much heat can one 500 rpm SSS throw out of the case?

@seemingly.random: the 500 rpm SSS would be the single fan, since the PSU would be passive (Zen 300).
For the single-fan project the configuration that seems sensible to me is an undervolted and underclocked 4850e cooled by ducted HR-01+, on an nVidia 8200 mobo.

I have an alternative project, which would involve 3 (three) 500 rpm SSS fans: one for an E8400 with HR-01+, one for an 8800 GT with AC S1, and one as the case fan.
Which raises another question: does an 8800 GT with AC S1 fit in the NSK3480? From the SPCR review it seems they should by a narrow margin.
The PSU would be Enermax Modu82+525, but I've read that it's fan has a bad signature and the top vents of the NSK3480 would make it worse.

Ok, I missed the fanless psu - didn't know that these were still available anymore. I still don't get that one fan is a holy grail though - I would think that zero fans would be. I suggest that you just get a couple of fans that you think might work and try them.

I've had an 8800gtx in an antec 3400. I also have the 3480. The main difference is that it now accepts a standard psu and has more vents on top. There is a caveat with the size though - it needs to be a short standard psu - 140mm deep as opposed to 150 or larger (corsair 520hx is 150). If it's larger than 140, then there will be trouble fitting the dvd and cables. I also flipped side panels since air intakes were added to the new version.

I've read reviews that show that the 780g igp is much better than the 8200. This might come at the cost of more heat though.

What do you feel is acceptable for an upper limit hd temp?

dhanson865 · Post by **dhanson865** » Sun Jul 06, 2008 4:14 pm

seemingly.random wrote:I suggest that you just get a couple of fans that you think might work and try them.

And my suggestion was to get one fan that would have enough range that voltage would allow you to do those tests without having multiple fans.

I have no idea what 500 RPM of slipstream will cool but you can get any one of 4 different slipstreams that will go below 500 RPM and one that will get down to about 600RPM. To me that says if your retailers don't carry the 800 RPM just look for a 1900 or 1200 and you can run them at their lowest RPM and see if it is too hot, and ramp them up manually or use speedfan to do it automatically if you think you need to.

Tzupy, I think you are going to find it hard for anyone to give you an exact answer because there are just way to many variables. Your ambient temp (Air Conditioning and heating), your case, sample variables on how much power is drawn and heat thrown off by each and every part you put inside that case.

If you build it I'd love to know the results of your testing but I'm not about to make a prediction.

porkchop · Post by **porkchop** » Sun Jul 06, 2008 11:49 pm

barring the passive psu, i can tell you that cooling everything with a single fan is possible, because that's what i did with my nsk3400.

my specs:
nsk3400
su380 psu- came with the nsk3400, fan swapped and spinning at 770rpm
x2 3600+- 1ghz @0.8v, 1.9ghz @.975v
abit an-m2
wd3200aaks -suspended
coolermaster hyper tx cpu heat sink - w/o fan
rear fan is a glacialtech silentblade @550rpm
120mm fan duct

current idle temps, ambient ~15C:
sys 24, cpu 20, aux 28, hdd 27
cpu load temp 36

my summer idle temps with a passive 2600pro peaked at about 40C all round- this is from memory, and please note that summer temps topped 40C outside and i don't like using the ac. the 2600pro was around 55C.

choose the right parts and do some airflow guidance and you're good to go.
BUT- i don't know how a passive psu would fare, and also note that the sqd will increase hdd temps, i'd recommend a 5400rpm hdd.

edit:
the fan is a 1000rpm model, set to 60% in the bios- i agree with all the people who suggested a higher rpm fan, its good to have options

. also, considering how efficient the slipstreams seem to be, i think its fair to compare it to my 550rpm glacialtech.

Tzupy · Post by **Tzupy** » Mon Jul 07, 2008 4:43 am

@jaganath: Great resources, thank you, I saved the pages to my HDD for future reference.
From those pages it seems that my planned 65W system (undervolted and underclocked 4850e with 8200 integrated gfx) would be properly cooled by the single 500 rpm SSS, which is rated by the SPCR review at 14 cfm. Bucharest is only about 150m above sea level, so the high altitude stuff shouldn't apply. But I'm getting high room temperatures during the summer, now about 25C and may rise upto 30C with a bad and lasting heat wave.

@seemingly.random: The Zen 300 seems to be standard size, altough I couldn't find the sizes in the SPCR review.
It did very well with a 65W load, at 80% efficiency. I would turn it upside down for better fitting of the cables in the NSK3480.
I choose the 8200 over 780G for programming reasons, I'm writing software that uses OpenGL and prefer nVidia for this.
As for HDD, I plan to use the WD GP 750GB, which is cheap at ~100 euros and uses a bit less power then the 1GB one, in Scythe Quiet Drive.

@dhanson865: for starters I'll probably use a Nexus as case fan and have the 500 rpm SSS on the HR-01+.
The Nexus would be controlled by a Scythe Kaze Master 5.25", for the 4 temperature sensors that I'll use to monitor CPU, chipset, PSU and HDD temps.
If the undervolting and underclocking goes well with the Nexus at 5V, I'll remove the Nexus and use only the ducted 500 rpm SSS.

@porkchop: it's encouraging to see that your system, close to what I plan, is a success, thank you for the details.
I plan to undervolt the 4850e to 1V (from 1.25V AFAIK) and downclock to 2 GHz (from 2.5), which should cut power in half, to about 23W.

@All posters: Now I feel more confident, so as soon as I have spare time, I'll start the build.
My only unclear point is which 8200 motherboard has good underclocking and undervolting capabilities, please give me a recommendation.

dhanson865 · Post by **dhanson865** » Thu Jul 10, 2008 6:03 am

I don't know if the 8200 GPU is affected but considering the fallout on other 8xxx GPUs I'd hold off on that purchase until I was darn sure.

http://www.theinquirer.net/gb/inquirer/ ... 84-g86-bad

thejamppa · Post by **thejamppa** » Thu Jul 10, 2008 11:18 am

The affected line seems to be G84 and G86 mobile versions. Like 8400M and 8600M GT. The problems seems not to be presence in desktop versions. What comes to the inquirer, I'd ber very cautious what to believe from them. Especially when it comes nVidia reporting by Charlie Demerjian...

That guy has deep hate relationship with nVidia and his reporting with nVidia is usually excessive, scandalous and more speculation and guesses than actual facts...

dhanson865 · Post by **dhanson865** » Thu Jul 10, 2008 12:01 pm

Do you feel the same about Adrian Kingsley-Hughes http://blogs.zdnet.com/hardware/?p=2203

Now you can take the Inquirer with a large pinch of salt if you wish, but last week when news broke of this issue I followed the trail of destruction that that led me to conclude that the GeForce 8400M was a likely suspect. More digging, along with news of an update from HP made me add the GeForce 8600M to the list of suspects. If some are affected, itâ€™s not a big leap to assume that all (or a large proportion) are affected, making this a very big problem indeed for NVIDIA. Want to dig the hole even deeper? Well, consider that these GPUs have also been used in countless notebooks, including Appleâ€™s MacBook and MacBook Pro lines. If this really turns out to be this big, then $200 million isnâ€™t going to go very far.

Until NVIDIA come clean on this issue (and according to one expert Iâ€™ve spoken to, it could take months to fully get to the bottom of this problem) uncertainty will cloud the company.

I'm not saying this is confirmed. I am saying there is lots of grumbling on them there InterTubes and you might want to avoid being burned.

Now get off my lawn.