Jump to content
Welcome to InsanelyMac Forum

Register now to gain access to all of our features. Once registered and logged in, you will be able to contribute to this site by submitting your own content or replying to existing content. You'll be able to customize your profile, receive reputation points as a reward for submitting content, while also communicating with other members via your own private inbox, plus much more! This message will be removed once you have signed in.


  • Content count

  • Joined

  • Last visited

About CharredPC

  • Rank
    InsanelyMac Sage
  1. Hey all! Blast from the past, huh? Just wanted to bring a conclusion to this thread (not that it needed one, but... ); the hot-rod Acertosh finally gave up the ghost a little while back. I never determined exactly what the cause death was other than it was an electrical surge to the motherboard. With many things going on in my life at the time, I grudgingly retired the project and moved on. If anyone still has one of these and has followed the progress here, I've kept most of the parts since so much work and documentation went into them (screen assembly with iSight webcam replacement / modded dual-link cable / high res LCD; extended battery; higher-amp charger; etc). It was a lot of fun rereading this years later and made me miss helping pioneer these hacks. Hopefully I'll have the time and money for similar projects in the future!
  2. 10.6.7 Update released

    Updated my Dell Mini 9 (Atom netbook with Intel chipset) from 10.6.6 to 10.6.7 using Software Update. I use the latest NetbookInstaller. Everything went smooth and works perfectly. No difference noted. Took a while to reboot as NI automagically patched everything needed, but patience is a virtue. And a song by Guns n Roses.
  3. DeviceMergeNub for _DSM style injection

    Let me test my understanding here... In theory, this is an editable kext that matches/edits/modifies/adds information to hardware profiles. I have a GMA965 / X3100 laptop with random EDID detection issues. The only fix so far has been a per-OS manual install of a Display Override, to force the EDID. It sounds as though maybe I could make a kext which does the job instead. If all that is true, a few questions- would it work from E/E, or need to be in S/L/E? In a dual boot 10.5/10.6 setup, would there need to be two versions of the kext in E/E, and would they conflict? Can anyone give a quick example of how this could be written/implemented? Many thanks in advance.
  4. The only real problem in SL now is... audio! I hate VoodooHDA. It has the internal mic problem, but also is totally useless after sleep for me as well. I'm working with Slice to get an alternate solution working, and will update here once we finalize a solution. If the SSD drive were small enough, it might even fit inside the 5620 casing (where the MXM card goes in different models). There's even another SATA port to tap into on our boards... maybe once I have the money I'll make that another project
  5. Huge props for this guys- much appreciated! I was forced to not use RC4, as it introduced a nasty "stuck at SystemVersion.plist" bug. Not only did this version fix that, but it also successfully enabled native cpu throttling, made my cpu and memory detect properly, and gave me working deep sleep! Major upgrade I can finally transition from 10.5.8 to 10.6.4, now that everything's fixed. It also booted my SL setup partition with no problem, making for a super easy install. Great work! (Acer Extensa 5620 laptop, with T9500 2.6GHz cpu, 4GB DDR2 (2 x 2GB) RAM, X3100 / GMA965 chipset)
  6. Well, good to hear that I have some time. Today I'm reinstalling Snow Leopard from scratch, and will make it work no matter what it takes. By the time you get your laptop revived, I should have instructions and a new set of drivers. Already done some tests, and I have working sleep, deep sleep (hibernate), and native video and speed step support under 10.6.4 with all updates. I think I'll finally make the move from 10.5.8 to 10.6.4 now that all that is solved... it's a much faster OS
  7. Never a bother! I'd consider myself lucky, and view this as a happy accident. Our boards were designed to have MXM slots (even my 5620 has a space for one), but they didn't implement it. I even considered buying one! In theory, it should be a drop-in replacement... make sure the LCD connector is the same, and I say run with it. You'll have a seriously pimped out Acer, and nearly any MXM card you toss in there will probably run circles around an onboard X3100
  8. I see no harm in trying a 1:1 match of my cable (since it works well- no shorts here!). I was pretty nervous the first time I tried new wiring too. Just triple-check everything, keep notes, and take your time. Not included in the pictures, I also decided to add the missing ground wires to my cable as well, back when I was trying to pin down the EDID issue. As far as I can tell, it made no difference either way... but the specs call for it, so it couldn't hurt. As long as you go strictly by the standard wiring code map, there shouldn't be any more sparks I'll try and dig up my latest DSDT tonight (one that isn't embedded in my BIOS!). I also highly recommend the latest Chameleon RC5 rev428 build package- it solved a few important glitches here (fixes dual boot RC4 problem; enables native speedstep- so no more need for Coolbook- yay!; recognizes ram & cpu properly; a few other random little things). Not sure how closely your hardware matches mine, but I'd be happy to post my current driver selection as well. As I said, I prefer Leopard on this machine because sleep is important to me. To be honest, once things were 100% stable and "just worked," I started using my 'Acertosh' instead of tweaking and fixing it! Which I guess means it was a success...
  9. Glad it was helpful! I'll take a look at what you've done tonight- sorry I've been away for a bit. Offhand, I wanted to say a few things first: -The 30-pin LCD pinout configuration is -technically- standardized. Between our two (very similar) Acer models, the pinout should be exactly the same. After studying my original cable, and the standard WSXGA pinout, I came to the conclusion that half the data channels were simply missing- luckily, no need for much trial and error! -The blank screen is normal! No need to panic. In order to get higher resolutions in Mac on OSX using the X3100, yes, you have to add the "dual-link" DSDT mod. I've put mine into the BIOS itself, to make things more permanent I can even explain how to do that too, if the LCD project works out. It can be really helpful when installing retail OSX (otherwise, no DSDT patch is loaded, and the setup goes 'blank screen' as well). -That picture may indicate a few things. You might just need more ground wires, or (if we're really lucky) it's simply a Ubuntu driver issue. I haven't tried it, but it may very well be that Ubuntu needs a "dual link" type fix as well. The important thing is that it DOES have picture! Was it just the images that looked poor, or was the text fuzzy as well? Sounds like a good game plan would be to do the dual-link DSDT mod with things "stock" (it should have no effect on your original screen, so no harm in adding it). Then put the new screen on and boot into OSX. And troubleshoot from there. These days I triple boot between 10.5.8, 10.6.x, and Windows 7. I'm usually in Leopard, though, since sleep remains broken in SL and I prefer my old drivers. It's worth mentioning that Win7 had no problem seeing and using the new LCD / resolutions.
  10. Mac mini and mini atx motherboards

    Hey there, long time no chat It sounds like maybe what you want to look at is the motherboards that Aopen uses in it's miniPC. I have an MP945-D here (I know, older than you want), and I think the motherboard would be just the size you're looking for. They made many newer revisions to this line, and I think they kept the basic form factor. It may be worth looking into getting one of these boards, or even just buying and gutting a miniPC for your project. Even my lame GMA950-chipset based one works excellent for a home theater pc running Windows 7, with full 1920 x 1080 HD output via HDMI to our 40" Sony HDTV.
  11. If the display hardware itself isn't being detected, I don't think this fix will work. We have to be able to reference it somehow, even if it's just by using the default 'generic' override (used when the EDID isn't picked up) of /System/Library/Displays/Overrides/DisplayVendorID-756e6b6e/DisplayProductID-717. If no screen is found, then it won't look for or use any specific override. Does that make sense? Think of the overrride as a driver. To make the driver load, we have to make sure it matches the hardware somehow. So if the hardware isn't being detected at all by OSX, then there's no real way to get the driver (or override) to 'catch.' Unless by "not being detected," you just mean that the display doesn't turn on / isn't being used. In that case, it might be in fact outputting, but at very wrong settings, and the default /System/Library/Displays/Overrides/DisplayVendorID-756e6b6e/DisplayProductID-717 location could be seeded with the full EDID. But this is just speculation without knowing more details about your specific setup
  12. Is the LCD monitor not detected at all, or detected incorrectly? If the actual hardware isn't being seen/used (no DisplayVendorID at all listed in ioreg), I'm not sure if this fix will work. I'd be happy to look at your ioreg output if it would be helpful
  13. Good tip about avoiding the truncating $! Do you know if your videocard / driver can output resolutions that high? Even if your display is capable, it's only half of the equation. The videocard still needs to be able to drive it. Glad this guide helped a bit.
  14. AnVAL (ACPI Loader)

    I'm sure all the OS's are syncing time with an online server; that's normal behavior. My point is, RC3 worked great and never had this problem. This bug started with RC4 (and continues in your build) so I view it as 'something got broken.' I wouldn't even budge off of RC3 if I didn't want hibernation ability with my 4GB of RAM....
  15. AnVAL (ACPI Loader)

    A couple issues... as mentioned, after installing, default (and only) theme on EFI partition isn't used. I could live with that, though, if for ****'s sake, some build of Chameleon somewhere would solve the infernal hanging at SystemVersion.plist introduced in RC4. It only happens when trying to boot 10.5.8, and can be solved by fully booting into SL, restarting, then trying 10.5.8 again. It often happens after booting to Windows 7, which leads me to believe it's perhaps a bios-time thing... which makes sense, if booting into SL resets it back. What bugs me is this problem was NOT there in RC3, and no one can manage to fix it in any later release (though it's been reported, discussed and attempted multiple times). valv, if you can get this glitch solved, you'll be my hero. I could care less if my ram and cpu are reported right, or even what my startup looks like. I just want it to be able to boot OSX... after all, isn't that what Chameleon was made for?