Latitude E6400/E6500 Owners Fan Noise?

Thanks for the info, I didn't know about those differences. The graphics memory is only different on the Nvidia right? If I chose the Intel 4500 integrated graphics it'll be the same on both machines I'm assuming.

 

Yes, I think so.

BTW, a guy on the E6400 owner's lounge says that with Intel graphics his fan doesn't come on, period. Even under load. I'm not sure if all E6400s with Intel graphics are like that, but it's good to know

 

E6400, BIOS A02, P8400; 1.000v at 8.0x (other specs are in my sig). I'd say my ambient temperature is about room temperature. Idling, the fan never runs. Period. Browsing the web and playing music... it almost never runs. Watching SD video (by SD I mean like 480p)... it almost never runs. Load it up more than that, and it'll start to run, but still not that often.

Overall, the fan runs less than than the one on my Nvidia-based machine. Idle temps are around 43C chipset, 40C core 0, 43C core 1. Most often the fan spins up when the chipset hits 50C, and spins down when the chipset goes back under 40C. Sometimes it just starts out of the blue. Honesly, if they could have made this thing run from 60C to 50C, it would be perfect.

i8kfangui does not work on the macine Fan control just spins the fan up and down, up and down. Fn + z does do something to the thermal sensors such that the fan usually shuts off if the chipset is below 50C. Also, the Nvidia heatsink is different from the Intel heatsink.

 

I have a E6400 with Intel GPU which I normally run in the docking station with the lid closed and with one external monitor attached. I am using Windows XP. BIOS A12.

Light usage (Web surfing etc), room temperature 23C:
The fan stays off approximately 30-60 minutes after turning on the computer. During this time the chipset and other internal components slowly heat up. When the temperature reaches 45C, the fan starts running at the slow 3000 RPM speed. The temperature of chipset etc decreases slowly. In the docking station it rarely goes below 42C which is the threshold to turn off the fan. Instead the chipset temperature seems to stabilize at 43-44C and the fan is running at 3000 RPM all the time. This is ok since the noise at 3000 RPM is not that bad.

I have noticed that the chipset temperature seems to depend a lot on the room temperature. E.g. with a room temperature of 20C the fan will probably not run very often at all. On the other hand, a room temperature of 27-28C or higher will probably cause the fan to run at the high speed mode of 4800 RPM. That is very noisy. From my experience the fan starts running at 4800 RPM when the chipset temperature reaches 50C.

So when you ask people if E6400/E6500 is noisy you should also ask what room temperature they have!

Edit: I just checked the temperature on my E6400 and I would like to make one correction: It seems that with BIOS A12, the fan starts running at 3000 RPM when the chipset temperature reaches 46C and not 45C as with earlier BIOS revisions.


Örjan

 

Yes, there is a correlation with ambient temperature. I'm running my E6400 in a room at 26C and, with light usage (forum browsing) the various hardware temperatures gradually rise to the fan trigger temperature, then cool down when the fan comes on until it goes off. The cycle repeats over many minutes.

It's not annoying me since I can't hear the fan but we can hope that Dell sees fit to slightly increase those temperature thresholds in the next BIOS update.

John

 

John, are you saying that your E6400 cycles between low speed fan mode (3000 RPM) and fan off in a 26C room? And that it doesn't go to the high speed fan mode at all?

Are you still undervolting? Could that explain the low temperatures you get?

Örjan

 

I can get the fan to go into fast mode. I just have to fully load the CPU for about 10 minutes until the heat spreads far enough to get the chipset temp to 50C. The attached includes a Mobilemeter plot where the vertical grid lines are every 2 minutes. First I tried the wPrime 32m benchmark but 5 attempts at that didn't have the required effect so I started up the 1024M test.

Note that Mobilemeter reads the ACPI temperature. The RMClock plot shows the temperature of one core. That is the hot one. Core 0 runs about 7 to 8C cooler when under load (and 4 to 5C cooler under normal use). Yes, the CPU is undervolted to 1.05V. The default maximum voltage is 1.1375V. I've never thought of a good reason to stop the undervolting.

I also noted that I8KfanGUI's chipset temp shows a maximum of 70C. That's a bit of garbage and is a bug with either the program or the BIOS. The fan switched to fast mode when the reported chipset temperature reached 50C.

I'm sure there are other ways of getting my E6400's fan to go fast, but it is a very rare event under my normal usage pattern. Something which loads the chipset / GPU more might reach the temperature threshold faster.

John

 

John, as you say the chipset temperature seems a bit strange. I think the default sensor configuration in I8kfanGUI is wrong for E6400 when not having a dedicated GPU.

I have changed the internal sensor configuration according to the attached screen shot. Since I don't have a dedicated GPU I have set the GPU sensor to index 0 which I believe corresponds to CPU temperature. It is normally a few degrees higher than the Core 1/2 temperatures.

I am still not sure I have the correct configuration set for chipset and memory,it could be the other way around. But these two values are almost always the same anyway.

Örjan

 

It's misleading to set sensor 0 as the GPU. As you note, with the Intel graphics the GPU is part of the chipset so the sensor is probably the same.

Give the Intel GPU more work (eg higher resolution or multiple displays or some 3D graphics) and the trigger temperatures may get reached sooner. However, I can't run 3D graphics and watch the temperatures at the same time.

John

 

By that reckoning, the smaller chassis M2400 would run hotter still due to a hotter GPU & still just one cooling circuit...