(Resistor) Mod a Laptop fan

[Antioch]

Hi guys.

It's a long story, but I have a machine that is identical to my friend's machine - same hardware, same temps. The only problem is, with the newer BIOS that I have the lowest setting the fan will run at is 3000RPM, while his BIOS gives a nice silent 1900RPM. Mind you this is the lowest setting, the only setting below that is 0RPM (off).

Until I can get the company to write some kind of fix in the BIOS, I'd like to see if I can resistor mod the fan so that it will run at 2000RPM instead of 3000RPM when at the same voltage. (I have a replacement part for the machine so screwing up is ok).

I was wondering if anyone can point me in the right direction or help me on how these little fans work. It only has 3 wires like a normal fan: orange/brown, red, and blue. Which is the RPM reporting wire? I'm guessing brown-RPM, red-hot(positive), and blue-cold(negative).

I don't know, but I'd like to figure out how this sucker works so I can temporarily lower the lowest RPM setting.

Thanks a lot, guys!

Here are some tech specs off of the sticker:
DC 5V 0.25A
MCF-211PAM05
Toshiba Home Technology Corporation

Based off that I think a 14ohm resistor ought to do the trick - lower it down to 3.5V. But I'm not sure if the fan is constantly drawing 0.25A of current. I would assume that at lower RPMs it is set at a lower voltage (and thus using less current).

 

[Super Nade]

Before you perform a mod, try using speedfan as a software mode control.

 

[Antioch]

Doesn't work.
The BIOS is FUBAR. The lowest setting it will switch the fan to is 3000RPM, that is at setting level 1. Level 2 is the same. Level 3,4, = 3500RPM, and 5,6,7 = 4200RPM. The control in the BIOS for ACPI and fan control only sets the fan to those speeds. However, as I said before, my roommate with an identical machine but an earlier BIOS, has setting 1 = 1900RPM.

My BIOS won't let me get it. TP Fan Control won't, nor will SpeedFan (SpeedFan actually doesn't work on my machine (Thinkpad T60, Core2Duo).

Unless I can get IBM to be smart and fix their crappy code, I'll have to stick a resistor in there and take care of it myself.

 

[SolidxSnake]

I'd believe the middle wire to be hot, left wire to be cold, and right to be tach...

I wouldn't be sure how to check it on a laptop fan, cause they don't have standard 3-pin fan headers, but smaller ones. What I'd do is open up the laptop, and then apply voltage to the fan (check the label for model no. or voltage rating), + to the middle pin, - to either side pin. If it spins, you've got the + and - leads. The remaining lead is tach.

What I'd do then, is cut off the + and - wires, and plug the tach wire back into the mobo. Then, I'd find a voltage source that would supply the fan, and then get a potentiometer, and then wire it so the pot could change the fan's speed externally and manually.

 

[Antioch]

I'd believe the middle wire to be hot, left wire to be cold, and right to be tach...

I wouldn't be sure how to check it on a laptop fan, cause they don't have standard 3-pin fan headers, but smaller ones. What I'd do is open up the laptop, and then apply voltage to the fan (check the label for model no. or voltage rating), + to the middle pin, - to either side pin. If it spins, you've got the + and - leads. The remaining lead is tach.

What I'd do then, is cut off the + and - wires, and plug the tach wire back into the mobo. Then, I'd find a voltage source that would supply the fan, and then get a potentiometer, and then wire it so the pot could change the fan's speed externally and manually.

Yea, that sounds like a good idea. I had IBM ship me a replacement heatsync/fan from my roommates machine type, they're the same except for the fan which is slightly different - but the different fan didn't change anything. So, I have a spare to experiment on.

I was thinking that I would want to wire up a pot like that, and I even found a small hole on the case to wire it through. The question is, do they make small ones? It'd only need to have less than 20ohms resistance so I would expect they could get small... but I don't know. More importantly I'd like one with a wired control (hopefully such a device exists) - one whose knob isn't affixed to the actual pot so that I could wire it out of the small hole in the case. Or if it is small enough I could just wire the entire pot externally and wire it's leads to the fan through the case hole. But again, I don't know what parts to look for or where to find them.

Could you point me in the right direction?

Thanks for the helpful suggestions.

 

[SolidxSnake]

http://www.allelectronics.com/cgi-bin/item/SVP-20/817/20_OHM_VERTICAL_TRIMMER_.html

There's a 20-ohm trimmer... You'd have to mount it in the hole and use a screwdriver or something of the sorts to change speeds.

http://www.goldmine-elec-products.com/prodinfo.asp?number=G15835

There's quite a large one.

I haven't found anything else as of yet.

 

[Antioch]

Thanks for the help.

The first one doesn't have any dimensions posted. I'll have to open the machine up again and see how much space there actually is... worse comes to worst, I'll still be able to figure out the best ohm resistor to put in after experimentation - and no matter what a resistor will fit.

 

[SolidxSnake]

THe second one, it's roughly 5mm*10mm*10mm.

 

[Antioch]

Where did you find the dimensions?

I found the following:
http://www.action-electronics.com/pots.htm#Trimmer

The first 2 look like they'd be good for setting in the case at wierd angles. However if yours is really 10mmx10mmx5mm then I'd have to go with it. These are .5" x .5" x .25"

Theres a smaller one down the page, but it has a higher max resistance. That shouldn't be a problem as I don't have to set it that high (and they are multi-turn).

 

[SolidxSnake]

Mine are the same size as yours, I just estimated. They're all fairly small though.

 

[Borisw37]

I am not sure if this will work:
The whole idea of having a tachometer in the fan is to have a closed loop control system. Basically the mobo receives the fans current speed and adjusts the PWM or voltage to keep it running at the set speed. So, if the friction in the fan increases over time, for example, the mobo will apply higher PWM duty cycle to keep it running at the same speed. Wouldn't the same thing happen if you put a resistor in there?

 

[Antioch]

SolidxSnake - Have you ever hard modded a fan before? Is there any electrical risk to the machine? So far as I can see it's perfectly safe - all the resistor does is drop some voltage between the motherboard and the fan, resulting in the fan running slower.

Boris, in theory that may happen, but I don't think it will for the following reasons:
The connector on the motherboard only goes to so high a voltage, so in the end it's capped.
The BIOS isn't controlling the fan based on it's RPM, rather by it's voltage. So with a resistor in series the motherboard terminal won't see any difference - the voltage difference will still be the same - it's the fan that sees a change in voltage.
I can say with little doubt that the BIOS isn't controlling the fan based on RPM, but instead using pre-configured voltage steps because on my roommates identical machine (with the exception that he has a CoreDuo, and I have a Core2Duo, thus the difference in BIOS rev) the fan runs at 1900RPM in the lowest setting while mine runs at 3000RPM (they are the same fan). Our temps are the same so I'm pretty sure it has to be the BIOS' setting that's too high. I've used TP Fan Control which interfaces with the ACPI to control the fan, and at setting 1 our machines run at the previously mentioned RPMs.

I don't think the fan is controlled by RPM because it's the simple 3-wired fan, not the complex 4-wired CPU fan that most processors use these days. Those 'old' 3-wired fans only report the RPM, which is why you can sever the RPM reporting wire and still have a fully functional fan.

Either way I really don't think putting a resistor will cause the BIOS to increase the voltage more, indefinitely - there is a max it can go to. True: with the resistor in place the max RPM will be lower (by about 1100RPM after I configure the 'idle' RPM to drop from 3000 to 1900) but if I really need the fan at max I can temporarily lower the potentiometer's resistance.

But your thought just might come into play - you never know...

 

[Antioch]

Speaking of those 4-pin CPU fan connectors (+volt, gnd, tach, & a green lead) - what does that 4th lead do? I assume it controls the RPM by some method other than voltage because I once used a standard 3-pin fan on a CPU-Fan socket, and the Q-fan didn't work on it, while it would work when plugged into a standard 3-pin motherboard fan socket.

Thats why I think my fan mod should work - my laptop's fan plus is the old 3-pin variation.