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

Panaflo CPU Cooler Fan Upgrade?

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

psb962

Registered
Joined
Jan 23, 2005
Hi, I need some advice from you fan experts on what benefit I can expect from upgrading the Panaflo 120x38 fan on my SI-120 cooler on a 1090T CPU (AM3 socket). With the current fan, rated at 69 CFM and 30 dBA I'm getting up to 56 degrees with a 3.6 GHz overclock on stock volts when running the maximum heat torture test on p95. I think I need to lose at least 5 degrees for comfort or maybe additional headroom.

My current fan is a FBA12G12L1A with molex connector, and is no longer made, but there is a current equivalent -L1BX model with the mobo connector. I'm looking at getting either the FBA12G12M1BX that does 86.5 CFM at 35.5 dBA or even the FBA12G12H1BX which does 103.8 CF M at 41.5 dB. I guess the latter is pretty extreme, but I will be using the CPU FAN socket and letting the BIOS control fan speed according to CPU temp.

Has anyone got any idea how much benefit in terms of degrees C I will get by upgrading the fan from what I have to either of the two listed?

Thanks in advance...
 
You'd probably see at least a few degrees difference going with either of the two; more with the stronger fan of course. Watch the amperage though. The medium speed (M1BX / .28A) fan would probably be ok on a MB header, but the high speed (H1BX / .46A) would be pushing hard on the limits of most MB headers. If you go with that one, a fan controller may be a good investment. It's usually a good idea to err on the side of caution when it comes to MB headers. Most are probably ok up to .5A (6W) but pushing that isn't something I'd necessarily recommend.

That said, years ago I successfully ran one of the ultra-high-speed ones (U1BX / 114.7 CFM / .65A) off an abit motherboard for about a year before I knew better. I count my lucky stars it didn't kill the header.
 
Really depends on the model and the manufacturer. As I said, erring on the side of caution.

psb962, if you want to be sure (which is a good idea anyway), check with whomever manufactures your motherboard and find out exactly what your header's limit is. Never hurts to know exactly.
 
I wonder how the SI-120 stacks up against the Hyper 212+ ($15-25 if you're in the states), seeing as how the SI-120 is fairly old. I don't think a fan change will make a huge difference with the SI-120's small fins. It also has very low resistance, so 38mm fans shouldn't make much of a difference.

You might try putting a spacer between your fan and the heatsink with the sides taped off with aluminum tape or other tape. The idea is to remove the dead spot and to force all of the air through the fins. Stock the SI-120 does loose some airflow out the sides. If you tape it off without the spacer you may create more back pressure reducing the cooling ability.
 
I used to run a Delta GFB on an SI-120 back in the day of socket 939. I saw a sizable decrease in temps, but then that was pretty extreme too. :D I had it on a fan controller so I could throttle it down to manageable noise levels.

As for temp decreases with more cfm, I am guessing that you would see a max of around 3-4 C difference going from an L1A to an H1BX or Sanyo Denki H101 or H1011. It does scale a bit with more cfm, but it's not a big difference.
 
Really depends on the model and the manufacturer. As I said, erring on the side of caution.

psb962, if you want to be sure (which is a good idea anyway), check with whomever manufactures your motherboard and find out exactly what your header's limit is. Never hurts to know exactly.

It's a Gigabyte 880GA-UD3H.
 
Having spent some more time looking at the temps I'm pretty sure the only way I can get this rig hot is by running Prime95 on all 6 cores. It seems nothing else can make it work anywhere near hard. I even did some video rendering last night and the temps were fine. I think I'll give it some more time to burn in and stay at a 3.6 GHz clock for now. From what you've all said it seems a better cooler is the way to go so I'll probably try a Hyper 212+ at a later date
 
Having spent some more time looking at the temps I'm pretty sure the only way I can get this rig hot is by running Prime95 on all 6 cores. It seems nothing else can make it work anywhere near hard. I even did some video rendering last night and the temps were fine. I think I'll give it some more time to burn in and stay at a 3.6 GHz clock for now. From what you've all said it seems a better cooler is the way to go so I'll probably try a Hyper 212+ at a later date

I can't really suggest a better cooler just yet. I'm having a hard time finding comparisons between an SI-120 and newer coolers. Think google search is just getting worse and worse.

So here's a comparison of SI-120 + SI-128 First bench

Another user review

Ok so 2-3C probably with SI-128.

Then 1C from SI-128 to TRUE 120

So maybe 3-4C to TRUE 120 from SI-120. I'm not sure of the difference between TRUE 120 and hyper 212+ as it seems back and forth between true 120 and hyper 212+.

Ultimately the SI-120 is a pretty good cooler and seems pretty close to what air can do for lower fan speeds. Since you do have a 38mm I would think that it would improve the temp difference on say an SI-128 or something else with longer and more narrow fins. Just make sure you do some research or ask on the forums before you're ready to get a new heatsink as there might be some new heatsinks or benchmarks at that time.
 
Update: I discovered that on Auto my mobo is setting the CPU voltage at between 1.35 and 1.4v, so I went to Manual and set it to 1.30v. Temps on full load P95 are now down at least 4 degrees, topping out at 52 degrees after 20 minutes of running. All seems stable, but time will tell.
 
Back