[To all newer AMD Turion owners]Strange power handling on hp dv5

Yes i have seen those pauses in just about every Windows OS. It's caused by poor mobo design. I know that my cousin's computer would boot XP in 10 seconds on a fresh install, but after getting all drivers installed it would take a 2-minute pause, then the same 10 seconds of HDD activity.

Anyway, i don't really know for sure if my optical is parallel ATA, but i've always been under this impression. I'll check... And no, i get no pauses during bootup on either of my laptops. The HDD is churning all the time.

 

It is advisable to disable devices you don't need in Control Panel
This reduces hardware polling in Windows hence conserving power and speeding up boot time.

 

Yes it is, but let's not go over the top with it. I was "hardcore" in XP too, but ultimately there are few configurations that really benfit from disabling services and hardware devices. My dual-PIII still running XP is one of them, where disabling the onboard SCSI controller (i have a SATA card installed in it) reduced boot time from 1:30 to 0:20.

I still have full functionality with exactly 11 system processes running. But that's a special case, and definitely a very special computer. Gigabyte 6BXDS, dual 700MHz PIIIs, 1GB SDRAM, voltmodded Radeon 9800 VIVO 256MB, Sil 3112A PCI SATA card + Western Digital 320GB, Creative AWE64 ISA soundcard, 4-port PCI USB 2.0 card, A4Tech wireless kb/mouse, Frontier Tricorn case with custom light job, and a kickass sound system with a custom-built 12" sub (my neighbours are half deaf so i can make noise ). Currently displayed on a 24" Thomson flat CRT TV.

It's my baby, over 3 years in the making and still not complete. I have two 1400MHz Tualatin processors coming from the USA and i will be buying a 21" Trinitron CRT monitor tomorrow. 2048x1536 ftw. You have no idea how many things this computer can do, and no matter how hard i tried i could never crash or hang it. I have a C2D rig at home (this PIII is at my grandparents' place) with 8GB RAM and HD3870, and i must say it's been a whole lot more troublesome. The dualie "just works", and i wouldn't sell it even if someone gave me $1000 for it.

In Windows 7 on my laptops i never felt the need to start tweaking system services.

Besides, Control Panel can't disable anything, Device Manager can.

 

Yeah, the bootup process is much faster than Vista snail edition.
But still there's a small pause (small... 3-4 seconds) with no disk activity at all just after the kernel is loaded.

 

I don't really tweak services, just devices.
There are devices that I probably won't even use at all like the dial-up modem and the wired NIC.

 

Well i don't think the modem is going to really take any clock cycles... If you want to be really sure there's the Performance Monitor. As for the wired NIC, i use that all the time. My router is wired.

 

Well I just tried again the parallel ata trick: enabling parallel ata controller from device manager results in a longer boot time with longer "thinking" period.
I wish I could disable parallel ata controller directly from bios, I think it would be great.

Th3_uN1Qu3, is your dvd reader PATA so? You can discover it inside Device Manager using View -> Devices by connection and finding your parallel ata controller to see what's attached to.

 

Th3_uN1Qu3, I had the chance to work on a acer 5536g notebook with Athlon QL processor.
Using latest turionpowercontrol, i found that altvid is set to 0.950v by default. According to datasheets, altvid is used when processor is in low activity state. I don't know *exactly* what that means, but probably altvid is used when processor is in C1E state.

May you try to set altvid to lower voltage and see if it gives you some more battery life? You should set altvid to the same voltage of pstate1, using the -altvid switch.

This is the current pstate configuration of the test machine:

Code:

C:\Users\gabriella\Downloads\tpc-0.12\x86>TurionPowerControl.exe -l
Turion Power States Optimization and Control - by blackshard - v0.12
Detected CPU:
Family: 0xf             Model: 0x3              Stepping: 0x1
Extended Family: 0x11   Extended Model: 0x3
Package Type: 0x2       BrandId: 0x1400
Detected Physical Cores: 2
Processor has 3 Power Planes
Detected processor: Athlon X2 QL processor
Processor has 2 cores
Processor has 8 p-states
core 0 pstate 0 - En:1 VID:40 FID:13 DID:0 Freq:2100 VCore: 1.0500
core 0 pstate 1 - En:1 VID:60 FID:13 DID:1 Freq:1050 VCore: 0.8000
core 0 pstate 2 - En:0 VID:48 FID:13 DID:1 Freq:1050 VCore: 0.9500
core 0 pstate 3 - En:0 VID:0 FID:0 DID:0 Freq:800 VCore: 1.5500
core 0 pstate 4 - En:0 VID:0 FID:0 DID:0 Freq:800 VCore: 1.5500
core 0 pstate 5 - En:0 VID:0 FID:0 DID:0 Freq:800 VCore: 1.5500
core 0 pstate 6 - En:0 VID:28 FID:13 DID:1 Freq:1050 VCore: 1.2000
core 0 pstate 7 - En:0 VID:28 FID:13 DID:1 Freq:1050 VCore: 1.2000
core 1 pstate 0 - En:1 VID:40 FID:13 DID:0 Freq:2100 VCore: 1.0500
core 1 pstate 1 - En:1 VID:60 FID:13 DID:1 Freq:1050 VCore: 0.8000
core 1 pstate 2 - En:0 VID:48 FID:13 DID:1 Freq:1050 VCore: 0.9500
core 1 pstate 3 - En:0 VID:0 FID:0 DID:0 Freq:800 VCore: 1.5500
core 1 pstate 4 - En:0 VID:0 FID:0 DID:0 Freq:800 VCore: 1.5500
core 1 pstate 5 - En:0 VID:0 FID:0 DID:0 Freq:800 VCore: 1.5500
core 1 pstate 6 - En:0 VID:28 FID:13 DID:1 Freq:1050 VCore: 1.2000
core 1 pstate 7 - En:0 VID:28 FID:13 DID:1 Freq:1050 VCore: 1.2000
Processor Maximum PState: 1
Processor Startup PState: 2

Minimum allowed VID: 52 (0.900v) - Maximum allowed VID 28 (1.200v)
Processor AltVID: 60 (0.800v)
Processor Northbridge VID: 52 (0.900v)

SMAF7 is disabled; processor is using LMM Configuration Registers for Power Mana
gement
DID to apply when in C1E state: 0

Voltage Regulator Slamming time register: 2
Voltage Regulator AltVID Slamming time register: 2
Voltage Regulator Step Up Ramp Time: 0
Voltage Regulator Step Down Ramp Time: 0
Done!

 

Will try, but now i'm not at home and i only have the DV9000 with me, as it is not only larger but gets twice the battery life because it has the 12-cell.

I'll be home after the 10th of September... But anyway, by then i'll probably afford to buy the DV5 a larger battery as well.

 

Got it. In the meantime I did some undervolting tricks on that Athlon QL and I went stable down to 0.975v volt at 2.1 Ghz. That's impressive!

Then I pushed it to 0.950v, windows crashed, wrote its memory dump to disk and at next reboot the MBR was GONE! All partitions turned into smoke, damn!

 

Wow you own that Acer?
I always wanted the new Acer 4535G with 4570HD Graphics.

 

fdisk /mbr

 

No it's not mine, it's from a friend.
It's an aspire 5536G with Athlon QL-64, 4Gb of ddr2 667 mhz, 320gb hard drive, ati radeon hd4570 w/ 512mb of ddr2, and 15,6" screen.

I did a 3dmark2006 and got almost 3000 points. My hp dv5 with hd3450 scores 1800
It's cooler than my dv5, but the fan is always spinning kinda fast and loud.

 

Unfortunately I think it would not solve my problem. MBR was completely gone, and since it carried the main partition table too... there were no chances to recover it. It was just zero-filled

BTW, I've found another little thing: minimum VID on QL processors is set to 52, meaning that any undervolt below 0.900v is uneffective! Just take a look at the post with the pstate config table above.

 

Blackshard,

I noticed your threads on multiple forums and I appreciate your work.
I noticed my ZM-82 ( in DV6-1264) behaves exactly like yours and I see that 1.2V thing popping up once in a while.

 

You never know unless you try.

I noticed that minimum VID in the latest version of your program when i downloaded it to my DV5 a couple weeks ago, but i didn't pay any attention and set it to 0.7625 as usual. I'll double check when i get back home, and if there is no change in temps it means you are right.

 

Thank you exercise. I just updated the first page of the thread with some useful instructions to solve the issue and a link to the latest version of the program if you need it.

 

I'll wait, in the meantime I will check if there's a way to reduce the minimum vid. On my zm-80 it's set to 0.750 and unfortunately I can't go below.

 

thx for the update but it don't compile under linux (gcc 4.4.1)
mario@tiulk ~/TurionPowerControl/linux src $ c++ TurionPowerControl.cpp Processor.cpp cpuPrimitives.cpp -o TurionPowerControl
TurionPowerControl.cpp: In function 'void processorStatus(Processor*)':
TurionPowerControl.cpp:222: warning: NULL used in arithmetic
Processor.cpp: In member function 'virtual uint32_t Griffin::minVID()':
Processor.cpp:451: error: 'Rdmsr' was not declared in this scope
Processor.cpp: In member function 'virtual uint32_t Griffin::maxVID()':
Processor.cpp:462: error: 'Rdmsr' was not declared in this scope
Processor.cpp: In member function 'virtual uint32_t Griffin::startupPState()':
Processor.cpp:473: error: 'Rdmsr' was not declared in this scope


thx
in av

 

Whoops
Try again, I think I corrected the missing routine (had no time to test it now, I'll double check later BTW).

 

thx compiles now

now i need to find out how to use

is ther something like a optimal settings script that ramps down the frequency and let the burnMMX (or similar prog) running until the cpu crashes and restart with this setting for the next ramp like http://www.s3pp.de/misc/linux-phc-optimize.bash

 

There are some examples on the first page, also there's a readme.txt file attached and if you launch the program with no switches it shows a list of available commands with some description.
Finally, there's me here

 

Sorry, no automated scripts out there, you'll have to find the processor limits yourself.

BTW, on recent turions and athlons pstate 1 (around 1 Ghz) can be safely pushed down to 0.800v.
If you have turion zm or rm, you can push pstate 2 (around 550-600 Mhz) down to 0.750v with no problems.

pstate 0 (full speed) is much more delicate, you have to drop some voltage step by step.
Turion ZMs go below 1.050v with difficulty, and don't know how RMs behave.
Recent Athlon QLs (QL-64, for istance) can go below 1.000v easily.

 

here the output for my cpu
./TurionPowerControl -l
Turion Power States Optimization and Control - by blackshard - v0.12
Detected CPU:
Family: 0xf Model: 0x3 Stepping: 0x1
Extended Family: 0x11 Extended Model: 0x3
Package Type: 0x2 BrandId: 0x560
Detected Physical Cores: 2
Processor has 3 Power Planes
Detected processor: Turion Ultra ZM processor
Processor has 2 cores
Processor has 8 p-states
core 0 pstate 0 - En:1 VID:36 FID:16 DID:0 Freq:2400 VCore: 1.1000
core 0 pstate 1 - En:1 VID:48 FID:16 DID:1 Freq:1200 VCore: 0.9500
core 0 pstate 2 - En:1 VID:60 FID:16 DID:2 Freq:600 VCore: 0.8000
core 0 pstate 3 - En:0 VID:48 FID:16 DID:1 Freq:1200 VCore: 0.9500
core 0 pstate 4 - En:0 VID:0 FID:0 DID:0 Freq:800 VCore: 1.5500
core 0 pstate 5 - En:0 VID:0 FID:0 DID:0 Freq:800 VCore: 1.5500
core 0 pstate 6 - En:0 VID:28 FID:16 DID:1 Freq:1200 VCore: 1.2000
core 0 pstate 7 - En:0 VID:28 FID:16 DID:1 Freq:1200 VCore: 1.2000
core 1 pstate 0 - En:1 VID:36 FID:16 DID:0 Freq:2400 VCore: 1.1000
core 1 pstate 1 - En:1 VID:48 FID:16 DID:1 Freq:1200 VCore: 0.9500
core 1 pstate 2 - En:1 VID:60 FID:16 DID:2 Freq:600 VCore: 0.8000
core 1 pstate 3 - En:0 VID:48 FID:16 DID:1 Freq:1200 VCore: 0.9500
core 1 pstate 4 - En:0 VID:0 FID:0 DID:0 Freq:800 VCore: 1.5500
core 1 pstate 5 - En:0 VID:0 FID:0 DID:0 Freq:800 VCore: 1.5500
core 1 pstate 6 - En:0 VID:28 FID:16 DID:1 Freq:1200 VCore: 1.2000
core 1 pstate 7 - En:0 VID:28 FID:16 DID:1 Freq:1200 VCore: 1.2000
Processor Maximum PState: 2
Processor Startup PState: 3

Minimum allowed VID: 64 (0.750v) - Maximum allowed VID 28 (1.200v)
Processor AltVID: 64 (0.750v)
Processor Northbridge VID: 52 (0.900v)

SMAF7 is disabled; processor is using LMM Configuration Registers for Power Management
DID to apply when in C1E state: 0

Voltage Regulator Slamming time register: 5
Voltage Regulator AltVID Slamming time register: 2
Voltage Regulator Step Up Ramp Time: 0
Voltage Regulator Step Down Ramp Time: 0
Done!

model name : AMD Turion(tm) X2 Ultra Dual-Core Mobile ZM-86

so what would be the arguments ?

 

You have to use the -pallc command switch, followed by pstate#, VID, DID and FID.
VID is used to set a voltage, and DID and FID are used to set an operating frequency. If you want the formula, you can check the readme.txt file. For now, we will just address vcore voltages. If you want to downclock too (not overclock, these processor are upward limited), feel free to toy around with DID and FID and expect crashes

However these will surely cause you no crashes:

1) TurionPowerControl -pallc 2 64 2 16

2) TurionPowerControl -pallc 1 60 1 16

1) will set pstate2 voltage to 0.750v to both cores, 2) will set pstate1 voltage to 0.800v to both cores, giving you some more battery lifetime and less heat.

Then you can start to issue:

TurionPowerControl -pallc 0 38 0 16

This will lower pstate0 voltage to 1.0750v for both cores. Use some stressing software to see if your computer crashes
If your computer doesn't crash, stop the stress toll, change 38 with 39, issue the command then restart the stress tool and let it work for another minute. So on, raising the value each time. Every time you raise the value, voltage supplied to the processor is lowered.

You may also issue this command to stop transitions to pstate 6:

TurionPowerControl -pallc 6 48 1 16

And also this command to mitigate transitions to pstate 7 effect:

TurionPowerControl -pallc 7 36 0 16

Eventually, Substitute 36 with the value you found toying around with pstate 0 above.

 

Good work Blackshard83,

But before I start fiddling with all that, can you tell how much of a temperature drop you get from undervolting (with those settings)?
I'm interested in lowering the temps when set at max performance (during gaming sessions).
Also, is there a way to manually control the fan settings?

Thanks for your time!

 

hi blackshard,

thank you for this nifty little program it works rather well on my tx2 (rm-74). im using it with k10stat (to do the clock control), and have it set to address the pstate 6 and 7 issue and to set the nb voltage, as well as to restore the voltage settings on wake from sleep (actually on unlock). all seems good... i *think* im getting better battery life, from 1 minute per percentage point on the 4-cell to maybe 1:15 or around 2 hours of normal use (probably 2:30 of idle) with wifi. Haven't really run any scientific measurements, but as long as it works long enough for my needs i'm happy!

temps still seem a bit high, but i think they're lower than what they were before (from 62Cish idling to 54Cish idling). these turions really run hot! and i didn't get quite so lucky with mine because the lowest stable voltage i could get at 2.2Ghz was 1.075, so my full-load IntelBurnTest temp still goes up to 98C or so.

huge thanks for this program it really helped me get the setup i wanted for my tx2!

k

EDIT: is there a way to underclock the NB?

 

Thank you exercise
Mostly depends on how much you downvolt your processor. Some people was able to gain even 10°C, but on average expect something like 4-5°C
Note that during gaming there's also the video chip producing heat and probably the heat dissipation is shared between processor and gpu, so each chip warm the other!

About the fan, actually I found no way to control it on hp dv notebooks.

 

Hi kix_r, thank you for greetings

you may use the -nbvid command to alter northbridge voltage, but looks like to me that's not influencing very much heat or battery life.

BTW I talked to HP technician once, and he told me that those temperature are "perfectly normal" and that if something goes overheating, to notebook just shuts down. My zm-80 went happily over 100°C under wprime, but it's still working fine so I don't really care about max temp...

 

What bugs me is that these things like leaving red marks on your feet in the shape of their air exhausts, so lower load temperature is something to be desired. And like any overclocker will confirm, the lower the temperature, the better. It does not matter if the chip itself is rated for 100C, there are a lot of other factors in play. See the 8400 and 8600 mobile GPUs.

 

Yeah, that's sure. Higher temps means more troubles in the near or far future, that's the reason why I'm gonna investigate my heatspreader problem.
BTW toying with the acer with athlon QL I had to format and clean, I just saw that they have a bit better heat dissipation, and processor never went over 85 °C

 

85C is pretty high as well, 80 is where i would start to worry. Well, at least it isn't nearly as bad as the Macbooks.

 

Before I bought the tx2500 I would have been worried about anything higher than 85C but its actually quite 'normal' for any of the cpu's in the tx2500 series to go over that temp just by visiting a flash heavy website. My ZM86 throttles everyday(max 103C~98Cundervolted). I'm surprised it lasted this long(year plus) without any side effects. I havent looked at how hot the macbooks ran but I would say my tx2500 ranks high amongst the hot runners!

 

Well, the Macbooks use C2Ds exclusively, so it's not like we could compare them. I'd say you're lucky, at that temp there's plenty of things that can go wrong.

As for the Macbooks, quite a few are known to only kick in the fans at full blast when the CPU reaches 95C. The maximum thermal limit declared by Intel is 100C, beyond that the machine should shut down. Apple doesn't do that in all cases... I'll let you draw the conclusions. In my experience it's safe to say that most Macbooks run in the high 90s at load (i remember there was a chart with Macbook temps somewhere), and the Apple people say that is perfectly normal. After all, they are the ones that explicitly state that the new Macbooks should not be kept in your lap or on your bed - they aren't laptops anymore.

I could see this warning being reasonable on a crammed thingy like the Air, or like your tx2500, but it is the same on the 15" and even 17" Macbook. My dv9000 was able to run for over 20 minutes without the fan. That says enough.

So for now, i'll stick to large laptops. My dv9000 runs at 74C full load CPU/80 GPU, and that's with room temps of 30C and an overclocked GPU. It's also pretty quiet, due to the large size it dissipates heat effectively so the fan doesn't work hard. I don't leave it on the bed without keeping something under its left side to keep the vents clear, but if i'd somehow forget it on the bed, i'm confident it won't melt. I can't say the same for my DV5, but still 84C on a chip rated for 100 is something i can live with.

Last 12.1" lappy i had was a Toshiba Portege, and it never got over 60C load. It was a Pentium III 933 though, and its vents were on the sides.

Sure, some may disagree with me on the safe temp range, but i build and fix audio amplifiers for use in professional sound events, and any electronic component exceeding a temperature of 85C cannot be considered reliable.

 

I'm not disagreeing with you but my tx2500 is not the only one which runs this hot. Pretty much every tx2500 out there will have similar high temps and as of now I haven't heard of any massive failures.

In any case HP should have designed it so as not to throttle under max load, doesn't matter whether the laptop is 12" or 20".

 

My suspicion is that the design is alright, but the build is wrong.

 

Nope My RM-72 averages 65-70 degrees on flash websites.
That is with undervolting.
HP AMD Tablets have design issues.

 

Try youtube HD and the temp should increase a bit more. Ofcourse there will be slight variances between the RM and ZM series but I have seen even the RM series get very hot under load.

Nah, its bad design. People have replaced the thermal paste and seems to have helped a little but overall the whole series has issues with thermal management.

 

Hmm, now youtube has become the favorite stress test. That's not good news... Adobe should get their heads out of their rears and improve the performance of Flash.

 

Agree. Probably that's why dv5 weight is a bit less than other comparable notebooks. They just made economy on aluminium/copper heat spreader. Actually taking a look inside the vents, I see gray metallic material, so probably HP used aluminium for heatspreaders, instead of copper

 

As far as i remember, weinter has an acer

 

I have opened my DV5 up so i'll confirm that. The cooler's fins are aluminum, while the heatpipe that transfers from the GPU to the CPU and the bases that sit on the chips are copper. But the dv9000 is the same and heat is much lower... However it is a couple inches larger.

 

No the temps are rather fixed because I did some modding myself on my Aspire 4530.
I optimized the layout for GPU gaming.


It is a rather extreme mod and there is sufficient space for it, most of all it did solve the gaming throttling issue.
Imagine RM-72 and 9100M G runs at max frequency
Originally the temp sensor was on the CPU heat pipe

 

Good job Weinter,

Yes it's a rather extreme mod but every little bit helps.
I'll pass on that kind of mod until warranty period ends to open my DV6 case, but I'll take good note of that tho.

 

I was referring to the tx2500 not Acer or any other laptop when I mentioned the temperatures.

 

Still, that's a cool mod.

 

Eheh nice job indeed

 

Interesting. I've been wondering whether adding some mass to my dv5's heat sink could help. Would have to make sure it doesn't block whatever poor airflow is in there at the moment. It's pathetic; this thing's temperatures with the stock heatsink and ICD thermal paste are still as hot as most processors without any heatsink at all.

 

Methinks the only thing that would help would be more vents. Remove the mobo from the case, stick out the dremel and get creative.

 

Meh, dunno. I think there could be some better engineering and bigger heatspreaders. Also I found my memory modules are very low density, then there are tons of chips on them: 4 gb on two banks, don't remember if they have 8 chips or 16 chips per bank, so many chips to feed surely produce lot of heat.
There's also something wrong with power handling and power usage on these dv5's, because battery depletion is faster than comparable machines, even when they are just idling. And I'm sure the processors don't matter.

BTW, I just found that using fedora linux 11, there's the USB bluetooth module that is eating lot of battery and doesn't allow processor to get into deeper sleep. Issued hciconfig hci0 down and magically mi idle temps went from 55°C to 47°C