M6800 Owners' Thread

Hello folks

I'm struggling with a very strange issue with an M6800 that I bought used a couple of weeks ago. Maybe it's something you'll have seen but it's a first for me.

Until I opened it up, it powered up and down as you would expect. But during some Windows driver tweaking work, I got strange and inconsistent information from the BIOS the device manager about the graphics card(s). The seller hadn't mentioned anything about a separate card inside - it was sold on an integrated graphics only basis. Initially I could see the card in Device Manager and the BIOS, then BIOS only and then it disappeared entirely. Since this all happened in a very short space of time, I wanted to be sure I wasn't dealing with ghosts in the machine - and given the age of the M6800 it seemed a good time to reapply thermal paste to the processor and clean the fans.

So I opened it up and it turns out there's a K4100M card inside. I cleaned up and re-applied thermal paste to both the CPU and GPU. Put everything back together and . . . the machine wouldn't power on. The battery charge light came on, but there was no power to/through the power switch. Tore it down again and got it to start after having removed the power supply, battery and CMOS. Ran the diagnostics checker (F5) and everything checked out just fine. The machine restarted and booted: it seems happy enough to restart from Windows 7 (the only OS I had on hand at the time).

But if I shut the machine down and try to power it on again, I get nothing. I've tried draining flea power and all the other soft methods I could find to fix the issue but to no avail. The battery light will illuminate if the battery needs charge and I've tested the switch and some other components with a multimeter and they all seem to check out. Right now, the only way I can get it to power up from the off state is to pull the power, main battery and CMOS. Put that lot back in and it fires up and stays running until I tell it otherwise. If it's any help the charger is one of the 240w monsters.

I've never had anything like this happen before and after checking and rechecking every cable and connector I can find, I'm stumped. So I guess my question at this point is what am I missing? This just doesn't make any sense . . .

 

On a quick glance it sounds like you bios might need to be reflashed. Even if it's up to date, it never hurts. Boot in to Windows and flash the latest bios over the existing one.

 

Yeah, do a bios reflash. It doesn't hurt, after all.

But my money is on GPU. Maybe seller didn't say anything about dedicated graphics because they were faulty, so it won't hurt to try to run laptop without K4100M and see the behavior.

 

Thanks people - that's kind of you.

I must confess that I don't know anything much about motherboard power architecture, but I would have thought that something like the BIOS wouldn't affect power to the power button; I could see it screwing things up once there was any sort of power but not before. Same goes for the graphics card - I need to get any sort of power in before that would throw a fit.

TBH I have a horrible feeling there's a busted component on the motherboard. I have no real idea what, but it's almost as though part of the power system isn't registering that there's power input or resistance across something is too high/isn't being overcome by the power going in - and that by pulling everything out and plugging it in again, it's forcing a one-time high power pulse that causes a switching component to latch.

I'll try the BIOS reflash and get a copy of Win 10 installed and see if that makes any difference.

Meantime, any hints or tips on building the OS onto the mini SSD card only? I've already realised the other drives need to be removed after some trial and error, but I'm still getting messages saying there's an invalid partition before I press enter and it boots up. That doesn't strike me as being right . . .

 

Not sure what's going on there. The mSATA drive is just like any other drive as far as the system is concerned, I've never had issues getting Windows loaded on it.
Tips:
- Remove other drives so that only one drive is present when you install Windows. (Or you can just disable the extra drives in the BIOS.) This way you can be sure that the boot loader and Windows are going to the same drive. Windows setup might split them onto different drives if it notices a previous Windows install or boot partition on a different drive.
- Make extra sure that you have UEFI boot selected in the BIOS, and that you are using UEFI boot to load the Windows install media. If you do not have secure boot enabled then the F12 boot menu will offer both UEFI and legacy boot choices. Windows will set the partitions up depending on which one you use to boot the install media. If you mis-match then you will get failure to boot after the first phase of the install.

 

Thanks Aaron -that's great. I think I was struggling with UEFI not being selected. But all that might be kinda academic now . . .

I have a really bad feeling that I've managed to brick the BIOS. Looking into the suggestion about the BIOS maybe needing to be reset, I flashed the new-old (V04?) BIOS to the machine. Part way through it said something Intel had failed but it was moving so fast I didn't register what. Anyways it continued its run and shortly after I got the BIOS flash completed successfully notification. Went back into the BIOS setup to check that it had taken and it reported that we're back on the older version . . .

. . . and then the fun really started.

When I put power into it now, it fires up to a black screen. The HDD indicator light flickers and periodically the fans spin up. And that's about it - it repeats that cycle until you yank the power. I managed to get it to boot one time after that; the BIOS screen reported the older version and nothing was obviously out of whack; it registered the new RAM fine, too. But then everything went dark and outside of the cycling behaviour, it's not letting me back in. I'm beginning to think the Intel message may have had something to do with the onboard graphics, although if that really had flat out failed I don't think it would have let me see the BIOS that one time.

I think it's safe to say that at this point I'm well beyond the scope of my competence and the next step is likely to involve some sort of repair technician. I'd maybe look at a replacement motherboard if they weren't so expensive to ship . . .

Arse.

 

Let's try just resetting the BIOS to default settings, if you haven't tried that yet ...
Remove the bottom panel, unplug the coin-cell battery, wait a few seconds, plug it back in and fire up the machine. Then try to access BIOS setup with F2 (if you get anything on the display)... you'll have to set everything up how you like it again.

 

Thanks again. I wish it was that simple, but sadly, that's not doing it.

I've had to remove the CMOS battery to get the machine to fire up at all. I should have mentioned in my previous post that I got to the point of being convinced the CMOS (or something related) was preventing power from getting to the starting circuitry. Symptoms were that the machine appeared completely lifeless with a fully charged battery until the CMOS was removed, at which point it showed signs of life (power switch lighting up). This may have something to do with the original CMOS battery connector falling apart when I disassembled everything the first time. I've since replaced it with another identical battery that I've tested and is good but I'm now wondering whether the polarity is some how reversed.

Anyway, the long and short of it is that the 'science' I tried earlier was all pretty much done without the CMOS being connected. Once removed, the device used to allow me to select either F2 (BIOS), F5 (diagnostics) and F12 (boot options) and use any of those utilities. F5 checked out without the CMOS and the only issue I found when I could see in the BIOS was that without the CMOS was the clock kept going out. But at least I could see the BIOS or utility display screens. After flashing the second BIOS flash, I couldn't see any of that - either with or without the CMOS. As I said, the machine just cycles with no display while the RAM heats up.

Given that the diagnostic tests went through without any sort of trouble the first time, I don't think it's a major component that's causing all this trouble. But I now think I've got two issues:

- either something non-CMOS related is preventing the machine from booting to BIOS or preventing the monitor from displaying anything; and

- something really isn't right with the CMOS battery polarity, connections or something local to it on the motherboard. It's almost as though the power stops when it hits the CMOS on its routing from the main power source to the power switch.

The sequence of events in getting to this point seem so unlikely that I really can't imagine many other people have encountered this situation. I've noticed that HP have a force BIOS recovery option that doesn't include flashing an offboard BIOS file (it's already resident) and I'm praying that Dell have hidden something like that . . . but honestly, I'm not holding out much hope.

 

My M6800 is system clock is jumping ahead 2 hours every time I power it down. I'm assuming it's the BIOS battery. Before I replace the battery, is there a good way to backup all the BIOS settings so I can quickly/easily re-load them once the new battery is installed?

Thanks

 

There's no import/export for the BIOS settings. You'll have to trawl around the screens and write down everything...