NotebookForums.com › Forums › Notebook Manufacturers › Sager & Clevo Notebook Forums › Sager & Clevo Notebook Reviews › Turion CPU upgrade - Sager 4750V (D470K)
New Posts  All Forums:Forum Nav:

Turion CPU upgrade - Sager 4750V (D470K) - Page 5

post #81 of 197
Quote:
Originally Posted by Stevew

After installing 2.04 I had some odd interrupt assignment messages when booting. I immediately reset the configuration data. That didn't help. I went back in and re-enabled the parallel port. I figured that would make it re-assign interrupts, and it did.
Stevew, was that a bios setting or through device manager? I got BSODs with Sager's 2.04 as well. It also got through a PassMark Performance test benchmark before the BSOD and scored 20 points lower than Eurocom's.

Eurocom gave an IRQL error while Sager's gave a page fault in an unpaged area.

I'm going to try a fresh install with Eurocom's bios over the weekend.
post #82 of 197
Thread Starter 
Quote:
Originally Posted by WFO
was that a bios setting or through device manager?
That setting is on the BIOS.
post #83 of 197
Quote:
Originally Posted by msantos
That setting is on the BIOS.
Thanks m. I appreciate your help. I'm glad some have successfully upgraded. It's looking like I won't be one unless I find the magic bios settings.

I've tried Eurocom's 2.04 and Sager's 2.04. They both cause BSODs alternating between IRQL not less than or equal too or Page fault in non paged area errors.

I've tried default settings, resetting configuration data, resetting the Serial Port and Parallel ports to auto, disableing and re-enabling them and nothing is working. It's not the OS or drivers. I get BSODs trying to reinstall Windows too. 4 attempts with different settings.

I'd love to use the Eurocom bios even without the chip upgrade simply because of the DMA 5.

If anyone could post the bios setting that are working it would be greatly appreciated.

Serial port A
Base I/o address
Interrupt

Parallel Port
Mode
Base I/O address
Interrupt
DMA channel...

and anthing else that may be pertinent.

TIA
post #84 of 197
Thread Starter 
Quote:
Originally Posted by WFO
If anyone could post the bios setting that are working it would be greatly appreciated.
These are my settings:

I/O Device Configuration
Serial port A: Enabled, 3F8, IRQ 4

Parallel Port: Enabled, EPP 1.9 & ECP,378,IRQ 7, DMA 1

Regards;
post #85 of 197
Well, my ML-44 came in. I flashed to 2.04a, and then installed the CPU. I've had windows up long enough to reinstall the processor driver and setup RMClock. However, on most boots, Windows (XP SP2) will freeze shortly before or after I login. The CPU heatsink gets *extremely* hot.

Both the old and new cpus are bare cores (no heat spreader), I used AS5 on the install.

Ideas?

Thanks!
-Ken

edit: put the old 3400+ Clawhammer back in, works flawlessly. What now?
Thanks!
post #86 of 197
Sounds like the heat sink isn't sitting on the processor correctly. Did you remove the cpu contact like MS did on his cpu cooler?

Edit - Ok that was someone else I guess. But it sounds like the heat sink isn't doing it's job. Like it's not transfering heat correctly. Make sure it's seated on the proc alright.
post #87 of 197
Quote:
Originally Posted by Poobah
Well, my ML-44 came in. I flashed to 2.04a, and then installed the CPU. I've had windows up long enough to reinstall the processor driver and setup RMClock. However, on most boots, Windows (XP SP2) will freeze shortly before or after I login. The CPU heatsink gets *extremely* hot.

Both the old and new cpus are bare cores (no heat spreader), I used AS5 on the install.

Ideas?

Thanks!
-Ken

edit: put the old 3400+ Clawhammer back in, works flawlessly. What now?
Thanks!
Do you have monitoring programs installed so you can monitor temps and perhaps more important processor voltages? What are they?
post #88 of 197
Thread Starter 
Quote:
Originally Posted by Poobah
Both the old and new cpus are bare cores (no heat spreader), I used AS5 on the install.
The main catch here is that the new CPU's have a much smaller core area - so it is critical that it makes good contact with the heat sink.

I would recommend that we do not power on the machine after placing the new CPU in. In fact, the idea is to:

1- Clean the heat sink contact surface well.
2- Apply a very thin amount of AS5 on the CPU core only
3- place the CPU in the socket and lock in in with the ZIF lever
4- Attach the heatsink and gently screw it in.

Then before powering it on - Remove the heatsink and see how much of the core actually made contact with the heat sink. If you have a perfect rectangular AS5 print on the heat sink then all's fine. This check can probably save you many headaches as well as money.

I suspect (like Digital and WFO) that your core in not transferring the heat to the heat pipe sink properly.

Also, if you verify that good thermal contact exists between the CPU and the Heat Sink then I would attempt to start Windows in safe more (F8 at windows startup). It may very well be that RMClock is feeding the processor the wrong voltage. Too much voltage can render a cpu unstable and be the cause for lockups.

Regards;


MS
post #89 of 197
m, just wanted to say thanks for posting the bios settings. Those are what the 1.04 bios loads on my machine. Did your ram timings change? If so, when I try again over the weekend, I may be able to correct things with A64 Tweaker. On a lark I ran Prime 95 for 4 hours with no errors on the 1.02 bios. The 1.04 won't stay booted long enough to complete a test.
post #90 of 197
Thread Starter 
Quote:
Originally Posted by WFO
Did your ram timings change?
The ram timings did not change with the BIOS only upgrade. However, when I inserted the new CPU then the RAM timinings and speed changed (as indictated in the review). I been using CPUz to check the effective ram timings.

Regards;
post #91 of 197
Ok, removed everything, cleaned it all up. Put AS5 on the cpu, set the heatsink on, pressed down, removed heatsink. Nice grey rectangle. Screwed it all down, fired it up. Same problem. heatsink above CPU is very warm, as are the heatpipes and the finned section. Still locking up. Don't get far enough along to run any temperature monitoring programs.

Grrrr!

Some Q's
should I remove the 3 strips of spacer material that are on the heat sink? (these are outside the core's contact area). I think they serve the same function as the circular spacer dots on desktop cpus.

Perhaps the BIOS is over-volting the CPU?

opened the bottom, propped up the laptop and pointed a 12" desk fan at it... still crashes booting windows. Linux will die occasionally, but generally survives a *lot* longer.

Edit:

Got it to run for a little while. Idling at 43 deg C (mobilemeter) when it puked and did a BSOD.
post #92 of 197
Thread Starter 
Quote:
Originally Posted by Poobah
heatsink above CPU is very warm, as are the heatpipes and the finned section. Still locking up.
That is actually a good sign as it shows the heat is tranferring to the heatsink adequately.
I am curious though... is the CPU fan actually working?

Quote:
Originally Posted by Poobah
should I remove the 3 strips of spacer material that are on the heat sink? (these are outside the core's contact area). I think they serve the same function as the circular spacer dots on desktop cpus.
The three strips are there to ensure that debris due to the air turbulence does not lodge itself around the core and the adjacent resistors on the substrate. So I would suggest keeping them there.

Quote:
Originally Posted by Poobah
Perhaps the BIOS is over-volting the CPU?
I don't think so. These BIOS updates are supposed to recognize the CPUs (whether Clawhammer ot Turion) and properly initialize them with the right voltage for the multiplier they start at.

Quote:
Originally Posted by Poobah
Got it to run for a little while. Idling at 43 deg C (mobilemeter) when it puked and did a BSOD.
It would be nice to observe what the effective multiplier and voltage the CPU is set at when it BSOD's. Can you have CPUZ (or RMClock monitor that?).

The Turion's (ML) voltage typically varies from .8 Volts (at 4x) to 1.35 Volts (at 12x in case of the ML-42 and ML-44).

Regards;

MS
post #93 of 197
It's difficult to keep it running long enough to install any software, but I will give it a whirl.

The CPU fan doesn't run all that much, tho i can hear it kick up and down speed wise. Definately a LOT less than with the clawhammer core.

For grins and giggles, I fired up my Linux partition, and let it run one of the OpenGL screensavers to try and stress the system a little, it kept running until the battery ran out - go figure

Not sure what's up with the windows side. Can't be drivers, safe mode pukes pretty quick too.

My current plan of action:
1- Do memtest86 on her. I was getting an occasional error with windows explorer prior to all of this
2- slap in another hard drive and do a vanilla windows install

Is it possible I got a bad chip?

Thanks!
-Ken
post #94 of 197
Thread Starter 
Quote:
Originally Posted by Poobah
It's difficult to keep it running long enough to install any software, but I will give it a whirl.
CPUZ does not need an installation. You just run it from wherever the executable is stored.

When you operate on battery alone your multiplier (regardless of CPU rev) is typically set to 4 (800Mhz) . It can be throttled explicitly to other multipliers-the same way RMClock does.

Hummm... It is very rare for anyone to get a bad chip. Typically the process quality control that both AMD and Intel use makes "bad chips" almost impossible. While statistically possible I really don't believe you got a bad chip (unless it got damaged since it left the manufacturing facilities ) .

I agree, a fresh Windows Install can probably help. I would also run HD tests(manufaturer specific - typically downloadable) before trying another drive.

Regards;

MS
post #95 of 197
memtest86 has been chugging along now for close to 2 hours (on AC power), no memory errors, no crashes.

gave windows a shot, BSOD right after logon.

Edit:
Well, trying a different Hard drive and new windows install didn't solve the problem. It's all very strange. We have a 2nd 4750 at work that took the upgrade no sweat. However, I have a new theory, My 4750 is old (like September 2004 old), I know there have been motherboard revisions, its possible that mine just cant do it. Anyway, I want to put the clawhammer back in, and flash down to the 1.04 BIOS and see if that works again.

Can someone send me the 4750 1.04 BIOS?
Thanks!
-Ken

(email: nospamken@badcow.net - remove the nospam)
post #96 of 197
Quote:
Originally Posted by Poobah
BSOD right after logon.
Can someone send me the 4750 1.04 BIOS?
Ken,
why don't you use Bios 1.05 nstead of 1.04?

My first experiences up to today are similar to yours. Had a few BSOD myself with XP SP2 after I've updated to Eurocom's BIOS 2.04a - same CPU = Clawhammer 3.2, nothing else changed.
My 4750 isn't that old, just over a year, but I'm still running the old keyboard BIOS as discussed before.

After the thírd BIOS update in a row I didn't have BSOD anymore, but Prime95 stopped after the first self test, which wasn't the case before. I went back to Medion BIOS 1.05 for now which was okay with DMA5, and is reliable according Prime95.

My biggest surprise came when I noticed that I have RAM speed now at 200 MHz with the same old 1.05 BIOS as before, which was set to 166 MHz previously, for sure. I have 1 GB with 2 x 512 MB Apacer DIMMs.

Anyway, the 1.05 Eurocom BIOS is still on their website. If anybody wants the Medion 1.05 (=015) BIOS, I can forward this - I had been taken off their website a while ago. The Medion has two more options (WLAN = last state or off at boot AND touchpad on or off at boot) plus a diff. start up screen (showing only AMD64 without further info).

I will prob. do further testing with changes to par. / ser. port settings the next days ...
Do we already have sb running 2.04a on xp sp2 sucessfully with the orig. CPU? If yes, can we have results if it is okay with Prime95 (torture test, 2nd option from top=much heat), please?
Also - CPU voltage used for max speed and RAM config + speed.

Thanks for all the support here - Steve

Edit: my CPUz info attached

UPDATE:
A couple of days and a couple of reboots later I'm back running my mermory at 166 Mhz instead of the 200 shown above (same BIOS) - strange behaviour ...
LL
LL
LL
LL
LL
post #97 of 197
Quote:
Originally Posted by steve byte
Do we already have sb running 2.04a on xp sp2 sucessfully with the orig. CPU? If yes, can we have results if it is okay with Prime95 (torture test, 2nd option from top=much heat), please?
Also - CPU voltage used for max speed and RAM config + speed.

Thanks for all the support here - Steve

Edit: my CPUz info attached
I don't think it is an issue with SP2. I was getting BSODs with both the Sager and Eurocom 2.04 bios' while trying to do a fresh Windows install. I think Ken maybe on the right track about minor hardware revisions.
post #98 of 197

Anyone with an "old" 4750 get the Turion upgrade to work?

Hi there,

I read this thread with much interest--as I've been planning on doing just such an upgrade since news of the Turion chip came out. However, I too have had my 4750 since 2004 (?) and am worried that the upgrade might not work for me. Is there anyone that has succeeded with an older 4750? Also, is there anywhere I can look that would indicate how old mine is? Or revision numbers we can compare?

Thanks!

Abe

p.s. It occurs to me that I might be able to tell how well my 4750 would take the upgrade by flashing to the 2.04a BIOS and then benchmarking it. If it runs the benchmarks stabily then it'll probably take the new processor okay. It seems that those who failed to install the Turion chip successfully had problems with even just the 2.04 BIOS--or am I wrong?
post #99 of 197
Thread Starter 
[quote=nullvision]Also, is there anywhere I can look that would indicate how old mine is? Or revision numbers we can compare?
QUOTE]

That is a good question. I have been using mainly software tools that typically read the mainboard revision. However, these tools do not return any info on my D470K machine. So I suspect that this method of determining the revision is of no use to all of us.

The only remaining approach is to actually physically expose the mainboard and locate a revision print (or sticker) on the PCB. Doing so may violate the warranty for some of us as it forces us to open the chassis shell and remove the warranty sticker.

Regards;

MS
post #100 of 197
I've posted my mainboard info (with the Medion BIOS) from CPUz on the previous post, pic # 3. It shows Rev. 01, but unless sb comes up with anything different, you have to look somewhere else.
Can anybody from Eurocom / Sager confirm where to look for the revision on the board (place of sticker etc.)?
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sager & Clevo Notebook Reviews
NotebookForums.com › Forums › Notebook Manufacturers › Sager & Clevo Notebook Forums › Sager & Clevo Notebook Reviews › Turion CPU upgrade - Sager 4750V (D470K)