Question alienware m17x r4 underperforming (RESOLVED)

May 2, 2019
8
0
10
Okay, so I've had issues with this hunk of metal since I bought it over 8 years ago and ive had problems varying from the batteries literally being drained and unusable, ssd failure, ram needing to be reseated, thermal paste needed to be replaced ect.. But the most recent issue was the computer overheating (idle temps of 200f+) so I took the liberty of buying a replacement gpu fan as I figured it could only either be the heat sink or the fan (it was the fan). I think the motor burned out, but it would make a high pitched ringing noise any time it was used. Now it runs fine but my current issue is still not fixed! My computer can run any game at 60+ fps but if the power button light isnt cycling I get 5-10 fps in any game and the computer runs just poorly overall . Literally any time I dont wait for the power button light to turn on, I start up a game and it lags horrendously. My only guess could be time for a new gpu, or maybe time for a new battery considering mine sits at 0% plugged in (not charging). I'd appreciate any bit of input, I've even considered taking the cmos battery out but I think its welded in there /:. My specs are >> i7 3610QM, 16gb of ram, and gtx 675m, bios a13 and 120hz panel. Also ive done a dell pc scan, and no issues were reported. I did notice however the bios has an update but I need my battery to be at 10% or more :) edit: also when the computer is running slow the response time dramatically increases. Like I could be at 1ms response time and it shoots up to 8-10ms of delay and stays that way. Generally its at .6 or 1 ms response time so this truly confuses me
 
Last edited:
I have seen systems run slow when the battery was not detected. Battery failure for this age of system is normal, even half of the 8 years you said you had it can cause a battery to die. No idea about the power light being on or off speed issue, power light should be on soon as you turn on the system unless it's also acting as some sort of system activity meter. How long does it take for the system to settle in and run properly with this power light thing?
 
It looks like your system uses NVidia Optimus - in its early stages, it had a lot of problems. What you described sounds like the switchable graphics aren't consistently working as they should; I'm guessing things are running off the Intel Graphics and that's where your poor performance is.

You didn't mention anything about drivers that I see, so I'd start there. Update your Intel Graphics driver, followed by your NVidia Graphics driver. This might improve the GPU auto-switching that's suspect.

The CMOS battery is almost certainly not welded to anything, so if you want to try a thorough NVRAM & BIOS/CMOS reset, go for it. The battery is either a standard CR2032 that pops out of its socket, or its the kind with the red and black leads, in which case you disconnect those from the board.
 
May 2, 2019
8
0
10
I have seen systems run slow when the battery was not detected. Battery failure for this age of system is normal, even half of the 8 years you said you had it can cause a battery to die. No idea about the power light being on or off speed issue, power light should be on soon as you turn on the system unless it's also acting as some sort of system activity meter. How long does it take for the system to settle in and run properly with this power light thing?

Well you see sometimes the power button never would turn on, and other times it would either a. Cycle colors or b.Cycle colors for a few seconds and turn black (when this would happen the pc would run slow aswell). I would unplug it and replug it back in and voila high fps and no freezes (for a short while).Now my current issue is blackscreening and green artifacts which I'm assuming dead gpu ): . I have a 660m replacement (yay) and I am currently waiting on a bracket to show up for the back of my gpu, and I have a brand new battery to go with it aswell. I'll let you know if it still gives me problems after this upgrade.
 
May 2, 2019
8
0
10
It looks like your system uses NVidia Optimus - in its early stages, it had a lot of problems. What you described sounds like the switchable graphics aren't consistently working as they should; I'm guessing things are running off the Intel Graphics and that's where your poor performance is.

You didn't mention anything about drivers that I see, so I'd start there. Update your Intel Graphics driver, followed by your NVidia Graphics driver. This might improve the GPU auto-switching that's suspect.

The CMOS battery is almost certainly not welded to anything, so if you want to try a thorough NVRAM & BIOS/CMOS reset, go for it. The battery is either a standard CR2032 that pops out of its socket, or its the kind with the red and black leads, in which case you disconnect those from the board.
I suspected optimus but any time I check my gpu is always present in both bios and nvidia control panel. It must be a dead gpu, or even a dead cmos battery lol.. I've updated the drivers and that did seem to help but alas any game I play was either unplayable (extremely low fps high latency) or would run fine for a few hours then decide to take a dump. Now I don't have to bother with the 675m but if the performance issue persists I might just do a clean install of win10 :crazy:
 
When you run a game, go to your desktop and run DxDiag. That should tell you which GPU is in use and if the auto-switching is working. System info software might also specify which GPU is in use.

A clean Windows install is a good idea if tinkering doesn't get you anywhere. If you run into issues even after that, then you might be right about the 675M being on its last leg. :(
 
May 2, 2019
8
0
10
When you run a game, go to your desktop and run DxDiag. That should tell you which GPU is in use and if the auto-switching is working. System info software might also specify which GPU is in use.

A clean Windows install is a good idea if tinkering doesn't get you anywhere. If you run into issues even after that, then you might be right about the 675M being on its last leg. :(

Did DxDiag and I got driver error code 18. Uninstalled my drivers and installed the most up-to-date ones for my 675m and ran it again with no issues. Now I still do get input lag when playing games. For example I used to play at .5 to 1ms in my game, and now any time I power on my laptop it always sits at 3-5 which is rather obnoxious. Also my battery doesn't charge, even with the new battery my charger still isnt wanting to allow me to charge it (also input lag 🤬). I uninstalled the drivers for the battery when I got rid of the graphics drivers and still no fix. Could it possibly be that my charger is a dell charger, but not the original laptop charger that came with it? If this is the source of my lag and battery issue then I will order a new charger asap. Have tried using it without the battery and it made no difference /: . Also why I think it could be the charger. Another thing I noticed after what I did was the power button light comes on and if I press it in time instead of me getting high fps low latency, I get 5 beeps and a blank screen. Held power for 30 seconds put new battery in hold fn until restart and after I turned off my system I plugged back in the charger light came on and BOOM high fps no lag. And now I just got the returning screen of "The ac power adapter wattage and type cannot be determined. The battery may not charge. THE SYSTEM WILL ADJUST PERFORMANCE TO MATCH THE POWER AVAILABLE. Now why is it I can play any game without lag but if I boot my pc regularly without waiting for this 3 second light it lags me to oblivion? What the heck man lol. My bios are a13 and I'm considering updating but it's hard to when your battery won't hold a charge! More than likely faulty charger ):
 
Last edited:
(...) Could it possibly be that my charger is a dell charger, but not the original laptop charger that came with it? ...And now I just got the returning screen of "The ac power adapter wattage and type cannot be determined. The battery may not charge.
I take it it's not the same model brick, then? If the electrical output is lower than the original charger, components that draw the most power (i.e CPU, GPU) will under-perform while under load. The issue with the battery could be related, worst case is it's the circuitry on the motherboard that's faulty and not delivering juice to the battery.

I'm not 100% about what the light indicator means. I'm guessing it's to signal when the discrete GPU is in use instead of the Intel one. If that's what's up, it would make sense why you have lag until the light changes (but why it's not reliable is less easy to guess at).

A BIOS update should usually be a last resort, but hey, it's worth a shot. Just understand that if the update goes awry, you'll likely end up with a laptop that doesn't work at all
 
May 2, 2019
8
0
10
I take it it's not the same model brick, then? If the electrical output is lower than the original charger, components that draw the most power (i.e CPU, GPU) will under-perform while under load. The issue with the battery could be related, worst case is it's the circuitry on the motherboard that's faulty and not delivering juice to the battery.

I'm not 100% about what the light indicator means. I'm guessing it's to signal when the discrete GPU is in use instead of the Intel one. If that's what's up, it would make sense why you have lag until the light changes (but why it's not reliable is less easy to guess at).

A BIOS update should usually be a last resort, but hey, it's worth a shot. Just understand that if the update goes awry, you'll likely end up with a laptop that doesn't work at all

Got an update for ya! After swapping my gpu out for the 660m I was getting error 43 and did everything (clean install, safe mode, EVERYTHING). I went to bed with a frown on my face, defeated but determined to resolve it. I plug in my laptop turn it on and dxdiag detects no issues control panel finally came back and I couldn't be more happy! Now I gotta put the battery back in and see if the performance is "up to par" but one thing I did notice was idle the temps went to 80c, but now its 73c so maybe the heat paste needed to get activated idk.. Would hate to repaste after all this work 😂. Cpu sits at 58c comfortably and my old 675m sat at 64 or so, probably because the GPU die is wider allowing more dissipation of heat. Really appreciate your help! No driver issues and I'll keep y'all posted on the performance :) . Edit: So I'm getting the black screen "you need a 240w power supply" even when it was working fine (charging, no lag) with the new gpu I moved my laptop plugged it in a bam same issue.. Only thing worth mentioning is "alienFX" is now showing in my power options which it never has before, do I need to install it for my computer to work properly? I've uninstalled it in the past and never had issues so I dont see why I would. Another thing I should mention is that control panel is completely gone again, even when the service is active :??:. Nvidia Geforce experience literally says I have my drivers, and all of them match the new version. Kinda figured maybe if I unplug it and only run off the battery I won't have issues and again I run into error 43 when I ran dxdiag. Didn't have the problem with the faulty charger so what could possibly be my issue?
 
Last edited:
May 2, 2019
8
0
10
Well I've updated to a15 and I've tried getting most the other drivers I might potentially be missing from Dell, I also installed an older version of my driver and voila! 660m is now working properly every time I reboot (yay). Now my issue goes back to the power adapter, which seemed to charge for a short time and now it doesnt want to charge at all no matter what I do. Really weird, but again probably faulty charger. My new charger is coming in Wednesday or possibly sooner so I'll give an update relatively soon. edit: for people with a similar issue, the driver I used was 352.84 win10 64 bit. Hopefully this helps a ton!
 
Error code 43 is sometimes indicative of a faulty card, keep an eye on that. As long as it's working for now and everything goes smoothly with the new charger, then congrats!

I would consider repasting and checking the fans + heatsinks for debris. 74 C at idle is way too hot for a 660M (or even something more high performance). Your CPU & GPU should sit at closer to 50 Celsius during idle.
 
May 2, 2019
8
0
10
Error 43 is gone entirely, but my issue with performance still insists even now that I have a working charger ): . You know how you press f3 f1 or whatever to continue or never see it again? I did that with my old charger and I'm thinking that could be the attributing factor, but shouldn't that performance dip from not having a original dell charge change right away the minute a charger that works is detected? Right now I'm two minutes to full charged as opposed to my other chargers (plugged in, not charging). Could it be more conflicting drivers? Dell has no drivers for me and my gpu driver is working fine with my 660, but even with my 675m I used to never get performance throttle like this i.e with my broken charger I could play fortnite 120hz, high fps. CS:GO at well over 100 fps and any other game high fps low latency. Makes no sense to me ): . One thing I haven't mentioned is I have a 3d vision laptop, but I dont see why that would make a difference considering I've installed and uninstalled those exact drivers. Edit: I installed msi afterburner and my pc is now at 83-91c idle 😡. Should I under volt, or use more thermal paste? Even then who knows if it will fix my throttling issue /:
 
Last edited:
May 2, 2019
8
0
10
So I put my older 675m in my laptop and found an older driver for it, (385.12) I also installed HWinfo and adjusted my gpu fan speeds and now my temps sit at a crispy 50c :) . No latency or lag reported with the new power adapter, but my cmos battery was in fact superglued into the slot but the clip isnt one that has a(+) wire but instead one that has a roundish piece it pushes into, and a v-like clip to hold it but I came to realize the original two prong clip isnt there, and I get beep errors when I boot and when it boots again black screen and turns off. Third time and pc turns on no issues at all, but obviously I need to glue the cmos battery back in place /: . Also it was my dad who broke the clip years ago and he just neglected to say he used gosh darn "monster glue" to hold its cmos battery in LMAO. So for future reference if anyone has this problem use monster glue! I'm running to ace hardware tomorrow to pick up some seal-all and I'm gonna use that since it has a temp threshold of like 150 and I doubt a cmos can physically get that hot unless you're gaming from oblivion.. lol. Pasted the seal-all and it dried rather fast, almost instant bond but I do suggest if you're doing this to keep pressure on the battery for a good 10-20 minutes at most to ensure it doesnt lift otherwise you will more than likely have to solder a new CMOS clip to the main board. Also note not all cmos batteries are the same, mine has no wires luckily so no soldering needed! I appreciate all the help, and I hope this helps people with my exact issue or similar. If you get black screen while playing games or tabbing out it's a simple driver incompatibility so fret not if you think your card is dead, more than likely drivers.
 
Last edited: