Acer Travelmate 5530G CPU upgrade to Turion Ultra ZM-87 - thermal throttling

[Blasku]

Hi!
As someone with an aging laptop, I'm trying to lengthen my TravelMate 5530G's life-cycle a bit. I've succeeded in upgrading pretty much everything by now except the CPU. My laptop originally came with a Turion RM-70, so I expected a considerable increase in speed (of compiling programs, and it did so quite nicely ) when upgrading to a ZM-87.
So the computer boots with the stronger ZM-87 and there aren't any problems with it but one: whenever I'm doing something really processor-intensive (e.g. play STALKER) I noticed that as the temperature goes a little higher (near 75 °C) the thermal trip point is reached and it immediately switches to the lowest p-state.
What's interesting, is that I can bypass not being able to warm the processor to higher than 75 °C by running something that only puts a 60-70% load on the system, and in this case, the temperature steadily increases beyond 75, until it reaches 78 degrees without throttling!
And 78 degrees is as high as it gets, thanks to my modification of the cooling system (all the thermal paste substituted with MX-2, thermal pads replaced with copper plates) and my notebook cooler (Cooler Master Notepal U2). So there's really no danger of an electrical fire . It's funny that the trip point would only have to be 5 degrees higher, and I wouldn't have a problem...
At first, I opened a thread at the notebookreview forums, where kizwan eventually modded (corrected actually) my DSDT table, setting the trip point to 95 degrees. The problem is that it didn't do any good . I've been told that I should start a thread here to see if any of you guys could update my CPU microcode, since that's the only remaining explanation.
Do you think that this is possible? I have a dual-booting system with Debian Linux and Windows 7, so any diagnostic program you might throw at me, I have a good chance of running it in case you need the info.

[Blasku]

I've already posted this bit of information at notebookreview.com, but I'll also write it down here: my new CPU surely isn't defective because it has been tested earlier in an HP Pavilion DV5 and it didn't down-throttle at all, not even at 90 °C.
Having read this subforum's readme, I think I'll link in a BIOS of that model, hopefully that way my thread might get a response.
So I choose the HP DV5-1150EN, some of those were shipped with Turion Ultras: download
And here's my BIOS, latest version (1.26): download. Although I doubt that is a full BIOS image there.
I've also attached the output of dmidecode.

P.S.: I haven't checked, but I think the HP BIOS is for machines with an AMD M780G chipset that is also in my Acer.

---

OK, here's a better one: HP tx2500z BIOS. These were shipped with ZM-86 CPUs and have the M780G chipset.

[Blasku]

I've just realized that both HP BIOS-es I've linked are Insyde, and not Phoenix.
I'll try to find a Phoenix BIOS that surely handles the Turion Ultras, but that won't be so easy...

---

I've just got an idea that the Fujitsu Siemens Amilo Sa3650 can most definitely handle the ZM-86, and has a Phoenix BIOS. The only question that remains is why didn't I buy the Sa3650 in the first place, instead of raping my cheap Travelmate with GPU upgrades etc. ?

[Blasku]

Almost forgot about this: if I'm going to have to overwrite my BIOS with a custom one, could you also include the DSDT tables that were corrected by kizwan? Then we'd kill two birds with one stone .

[TheWiz]

Sorry for the delay in replying to your thread. I will be sure to take a look at your BIOS and the work kizwan initially did as soon as possible. I am curious though if kizwan ever tried to see if the active trip point was a hidden option in your BIOS as it is quite common as a hidden feature in the CMOS setup that may get the ball rolling for you. It is possible the CMOS is overwriting the manual entry kizwan made in the ACPI table.

Thanks,
TheWiz

[Blasku]

Sorry for the delay in replying to your thread. I will be sure to take a look at your BIOS and the work kizwan initially did as soon as possible. I am curious though if kizwan ever tried to see if the active trip point was a hidden option in your BIOS as it is quite common as a hidden feature in the CMOS setup that may get the ball rolling for you. It is possible the CMOS is overwriting the manual entry kizwan made in the ACPI table.

Thanks,
TheWiz

No problem , I'm just glad that you're now looking into this.
I don't recall kizwan mentioning anything about the CMOS setup, he only modded my DSDT which I checked with asl.exe under Windows (check out the notebookreview forum topic I've sneakily hyperlinked in my first post here).

[TheWiz]

Hi Blasku

I can include the custom DSDT tables, but it would be easier if you could post a link to the bios with the DSDT already inclucded and then I will run a CMOS check on the corrected DSDT BIOS revision.

Thanks,
TheWiz

[Blasku]

Sure thing, but I didn't patch my BIOS permanently with the custom DSDT yet, I just checked it out with `asl.exe /loadtable` command on one boot of my Win7 (which I've reinstalled since). So saving the BIOS with the help of my DOS pendrive wouldn't help much.
Do you know any program (sorry if it's mentioned in a FAQ around here already) that could save the already loaded and patched BIOS from the RAM correctly, instead of messing around with the ROM? That would do the trick, wouldn't it?

EDIT: I managed to insert the DSDT via PBE and the build was successful. So the following 2 posts can now be ignored. The BIOS can be found attached to post #12 of this thread!

[Blasku]

Uh, did I say something naughty for not getting any more replies?
I seriously would be grateful if this could be solved in the next few weeks, instead of another 2 months. I'd even pay for not being limited in what I can use my laptop for any longer.
If you need anything else that could help, then don't hold it back!
One more question, is this a complex problem to solve or you didn't have time to look into it yet?

[Blasku]

I started to look into updating the microcode in the BIOS by myself, and getting the WPH file out of the installer exe for the Amilo Sa3650 was a 2-hour (Google's relevant results FTW ) process in itself. Actually the installer is impossible to work with, but thankfully a bootable ISO image was also available on the website. That image was created with Nero's hard drive emulation option, with the actual boot image hidden. I sorted that out with the help of this post.
Next, I had to trim the two WPH files to 0x200000 in WinHex in order to get the ROM files that PBE finally deigns to open.
After that it was easy to extract the UPDATE0.ROM out of both BIOSes with the TEMP folder trick. Comparing these two files I found that the original is larger by a few bytes (I guess that'll make replacing it easier with some padding). There are a lot of other differences in their contents too, so this might actually make a difference.
Before I proceed in replacing the UPDATE0.ROM as according to this guide, I'd like to ask one simple question: just off the top of your head, do you think updating the CPU microcode will really help my problem?
Also, updating the DSDT is done simply by replacing the ACPI0.AML module in PBE, right?
I think I'll wait a few days for your reply, after that I'll try to do these steps by myself and will post the results here.

P.S.: I'll attach the FuSi's BIOS to make things easier for you in case I chicken out .
P.P.S.: BIOS with updated DSDT attached.