Turion Undervolt

[thi]

I did a search but did not find any posts on undervolting turions. Sorry if repost. I have just recently undervolted my v2000. It works great! I'm curious to know if anyone have any undervolted the ml-28. My system is running stable at an unusually low voltage.

V2000 Specs:
AMD Turion ML-28
512 ram
12-cell battery

oe volt:
4x = 1.000v
8x = 1.450v

undervolt:
4x = 0.875v
8x = 1.000v

I believe I can go even lower, I haven't crashed yet (Prime95 full test w/ cpu @ 100% and also a test playing WarCraft III). I will keep these settings and see how it performs under normal use for a few days.

Let me know your results if you've undervolted.

 

[frostedflakes]

It was set to 1.45V @ 8x? Seems kind of wierd, as default voltage for the ML (35w) Turion64 is 1.35V.

Anyways, I used to have an undervolted Turion64 MT-34 on the desktop. Was able to take it to 1.1V completely stable.

 

[Slicedbread]

Ive wanted to, but never had the time and I was afraid I might damage the system. Ive heard that the lower voltage requires a higher current, which could to a higher cpu stress - but I dont know if this is true since the charge is moving faster but its less powerful.

My temps (55c) and fan duration (barely ever) seem to be the same at 800MHz and 1600MHz, so I didnt think undervolting would change much.

I might try it now since you did it, because until now I have only seen dothans being undervolted.

I have ZE2000 series which is pretty much the same as yours.
Sempron 2800+
768 Ram
6-Cell bat
Maximal: 1.350v
Minimal: 1.075v
Using RMClock

 

[frostedflakes]

That is false, lower voltage does NOT require more current, but rather less.

Go ahead and give it a shot, don't worry, you won't mess anything up. Just take it slow, lower voltage in small increments testing for stability each time.

 

[thi]

youd be surprise at the diff, if you need help just post here. I'm new at this also.

 

[TuxDave]

Originally posted by: frostedflakes
That is false, lower voltage does NOT require more current, but rather less.

Go ahead and give it a shot, don't worry, you won't mess anything up. Just take it slow, lower voltage in small increments testing for stability each time.

Lower voltage will lead to lower drive currents from transistors. HOWEVER, lower drive currents means drivers are weaker and there will be longer RC slopes on the input of the load. This will lead to higher contention currents. In other words the input voltage has a longer period of time in the 'not quite on but not quite off' , so using CMOS inverters for example, the PMOS and NMOS pair would both be on creating a brief short between the supply rails.

But meh, if the net result is lower power consumption, then the tradeoff is probably still positive.

 

[Slicedbread]

My dilemna is that I rarely ever have it ony my lap or disconnected from an outlet, and even with its fan vents being covered the fans rarely go on. When monitoring the temps, the cpu can be at 53C and have the fans spin up or at 57C and the fans wont start.

However it would be nice for lower temps and more battery. Im afraid to try anything because I bought this for 415 at walmart on black friday, if something were to screw up, I wouldnt be able to replace it soon.

How long should I run Prime95? And are you using RMClock or NHC, because NHC doesnt let me modify the power options.

EDIT: RMClock does not let me undervolt the under 1.075 v (which is default minimal) so I wouldnt gain anything out of undervolting since it is rarely at 100% usage.

 

[Hacp]

how much does blife improve after undervolting?

 

[Aarondeep]

What tools do you use to undervolt the laptop? I don't have any bios options for voltage setings.
I already have a MT turion which is rated a 25w. I wonder if this would help any.

 

[thi]

I've noticed a 30 min increase in battery life so far.

Regarding software, I use RM CPU clock (to adjust the multiplier & voltage) & Prime95 (to test the stability of the changes done in RM CPU clock).