And finally.... My system has arrived...

[bandieramonte]

I like the look of this one:

http://www.aerocooler.com/shop.cart?act ... FANSC38122

If two of them are too loud at 12v, then you should be able to undervolt them to 7v and still get decent airflow.

Dan

Sure I would aim on those fans, but unfortunately those are not for sale here. I just bought two of the EverCool fans I just posted before, the ones that are 80CFM. Those were the ones of highest CFM here. They should arrive some time this week. I'm very anxious to see the performance increase between 37CFM and 80CFM. I hope it can at least lower idle temps from 39C to 35C, we will see...

One more thing:
I'm not sure about the connection maze I did with the 4 case fans (2 12cm and 2 8cm fans)from the Thermaltake Armor. See, what I did was to connect the 4 fans themselves using their 4-pin connectors, and then, the last 4-pin connector was connected to a single 4-pin connector from the PSU. Would this ensure proper power feeding to the 4 fans? I ask this because the two 12cm fans don't seem to blow almost any air if at all. I mean, I put my hand very close to the fan's blowing direction (say 3cm close) and I almost don't feel any air at all. But when I do the same with the 8cm fans, I do feel some considerable amount of air being moved. I don't know if the 12cm fans are improperly powered. Do I have to use another PSU 4-pin connector? (Also take in account that the Thermaltake Armor comes with a 25cm fan that I haven't even still connected, and don't know whether if use a single PSU 4-pin connector for this one... what should I do?).

It would be a good thing if I could by some means measure the RPM being performed by every fan, and control them too.

[Methious]

I've got the Thermaltake Armor (and two of their other cases), the Armor's 120mm and 90mm should be fine on one Molex, they don't put out a large volume of air and are more designed for silence.

The 250mm side fan I'd run on it's own molex just to keep the lead from stretching across the case but you could run it off the same molex the other 4 are connected to.

Typically when I tried some expensive Noctua 90+ CFM fans in place of the stock fans front and back it only made one or two degrees difference in the cases internal temps.

[DMB2000uk]

You have a 250mm fan on the side of the case and you haven't turned it on?

That should help with overall case temps

Dan

[bandieramonte]

Well... All the benches that I've posted in this thread were done having the system outside the case, on a solid surface because I had to re install the GeminiII many times and it was just too tedious to remove the motherboard from the case every time.
Did I did too much wrong?

But now I'm almost done installing the motherboard in the case. I'm waiting the two 80 CFM fans to improve GeminiII's performance some little bit. The problem I have now is that when I attach the motherboard to the case with the 9 screws I get memtest errors again; when I remove the screws, no errors appear. So I concluded that the screws are causing short circuits or grounding the system (not sure which one). So I think that if fasten screws using some plastic washers the issue should be solved. Is this correct?

[martini161]

no, screws will not short circut a system. if you notice, motherboards nowadays have this silverish stuff around where the screw goes in, to prevent it from interfereing with the PCB circutry in any way

[DMB2000uk]

Over-tightening a screw might do that, and some screws can overshoot that silver stuff.

Did you try testing the motherboard without screws by resting it on the case standoffs, with the case on it's side (so the motherboard doesn't fall out)?

If that is the case, then plastic washers should help. If you don't tighten the screws too much, you could even mock some up out of paper (provided the tightening won't rip the paper).

You didn't do anything wrong by having the mobo out of the case for testing (I'd have probably done the same if I had to re-seat a cooler that many times), but you might find that you get better temperatures with the motherboard installed in a case, as there will be airflow across the board from the case fans.

Dan

[bandieramonte]

The new fans are truly impressive....

I just installed them right now and my temps at idle are 32C having ambient temp at 27C. Now I feel secure to overclock this Q6600 to the way it's meant to work. Yes, good, I still haven't even plugged the 25cm fan.

I will research now on the ideal FSB, multiplier, and RAM memory FSB to have the system stable and the processor on 3.2 Ghz, for now. If you got any ideas, please post on.

[martini161]

sadly there is no way to get one ideal set of settings for every procesor or motherboard. you just have to experiment on your own, as hardware is like snowflakes, every one is different

[bandieramonte]

I'm getting crazy here; My idle temps are now 44C with the same fans. I have not done any kind of changes on system settings nor on the hardware setup. The only different thing right now is the ambient temps, which is 31C. But before changing the fans, when the ambient temp was 31C, the Q6600's temp at idle was 39C....

To summarize:

With the old fans:
-The idle temp was 39C when the ambient temp was 31C
-The idle temp was 35C when ambient temp was 25C

With the new fans (40 CFM more):
-The idle temp is 44C when ambient temp is 31C
-The idle temp was 32C when ambient temp was 27C

I can't see why my temp is now 44C...... Is there any other factor here?

[martini161]

is the big side fan plugged in?

[bandieramonte]

The 25cm fan was not plugged in. But now it is, and the processor is idling at 37C having ambient temp at 27C. It's strange because before, even when having the 25cm fan unplugged, the processor was idling at 32C under the same ambient temp of 27C.

[DMB2000uk]

Are the fans plugged into ports that have some kind of dynamic throttling turned on? i.e. they will run full speed when the CPU (or most likely the motherboard) is above a certain temperature. See if there is anything like that in the BIOS and disable it.

Dan

[bandieramonte]

All of the fans except for the dual fans of the GeminiII are connected directly to the PSU. The fans on the GeminiII are connected to 3-pin connectors of the motherboard.

If the dynamic throttling is on, should I shut it down in the BIOS?

[martini161]

it wont be called dynamic throtling, look for something along the lines of "Fan control" or something. and yes you should turn it off

[bandieramonte]

Hello everyone;

Sorry for the too late response, I had tons of things to do and I haven't almost had time to use my new system.

About the fan control, I could not find any "Dynamic Throttling" or "Fan Control" option in the BIOS. The closest thing I could find was, under the H/W Monitor menu, the following:
- CPU smart fan target , which is set to disabled
- Sys fan1 control , which is set to 100%

The temp problem remains unresolved...

Also, I just did my first overclock. Since it's my first time, I just unlinked the RAM and CPU, and only increased FSB from 1066 to 1200, attaining the 2.7 Ghz overclock. The problem with this is that I still don't know how to set the CPU voltage to a fixed level, because my system is all the time dynamically controlling it. In the Cell Menu of the BIOS, I did:
- D.O.T. Control is set to disabled, although I don't exactly know what is this
- The Intel Eist option is set to Enabled , which I think is good because the CPU throttles down when in idle (I don't know if I'm wrong with this)
- System Clock mode is set to Manual
- FSB is set to 1200
- This is something I don't understand. The CPU Voltage variable is at 0.00V, and when I change it, I can only set it up to a max of 0.3875V. Is this the CPU voltage for overclocking that I must fix and set to 1.15V if under 3 Ghz so as to attain the lowest possible temps while at the same time ensuring system stability? How can I set the CPU voltage to a fixed level?

Thanks for all help.

[martini161]

hmm, they are probably talking about how much voltage you are adding to the cpu, so if you put at its highest of .375, you actuall voltage would be the stock voltage plus .375

[bandieramonte]

hmm, they are probably talking about how much voltage you are adding to the cpu, so if you put at its highest of .375, you actuall voltage would be the stock voltage plus .375

Yea, but how may I set the CPU voltage to a fixed amount, and not to a dynamic amount set by the system?

[DMB2000uk]

Things are better for the temperatures if you let it be adjusted by the system, as it will go to a lower value when it isn't doing anything.

However for the purpose of overclocking you can keep the core voltage the same by disabling the EIST and C1E features.

Dan

[bandieramonte]

When I overclocked to 2.7 Ghz, I had to underclock because when I ran the memtest application to test system's stability, in didn't last even 1 second when it showed errors, and worse, memtest showed a Halted message and the system rebooted. It did this all the time I ran memtest. After underclocking back to 2.4 Ghz, memtest could run fine without errors.... any insights onto this?

After underclocking, I disabled the EIST feature, but the system is still controlling the CPU voltage and speed. But what's rare is that the Core Temp utility does show me that the CPU speed is fixed to 2.4 Ghz, and it also shows that VID is 1.2875v. But the CPU-Z application shows that the CPU's core speed is always being dynamically changed by the system, as well as the Core Voltage, which is ranging around 1.14 - 1.2 V. I did not know how to disable the C1E feature. The Bios does not show this. In which menu is it supposed to be?

[martini161]

cpuz doesnt acuratly read voltages