Latest BIOS 1.31 for Aspire E5-553-1786 Makes Sudden Reboots Worse

FerchogtX

Hi all.

Just to let you know that, I had to downgrade my BIOS, because the latest version available for my model cripples the system, seems the code rewrite wasn't as effective... I'm currently on version 1.19, as I was before this BIOS 1.31 release.

Fortunately, seems I'm figuring out why this model of laptop has sudden reboots (for me, once after a month or so) and seems the BIOS is having some sort of failure, specially with the HyperTransport (Bus) part, that causes a WHEA Error Code 20 and a bugcheck 0x124. The funny part is that this happens even when idling, intense gaming or work doesn't trigger a reboot (temps are OK)

Oh, taking advantage of this post, seems this latest BIOS upgrade enables TPM 2.0 on this system (which you need to disable in downgrades or you get error codes 10 with a resource already in use message in Device Manager), which is funny, because that is never used on a system tied to a home edition Windows... (Bitlocker is exclusive to Pro editions)

I hope Acer team can do a better job on coding a BIOS next time, I can't even see if this is an official release, because for Meltdown or Spectre vulnerabilities, this model appears as "In progress", so, I ignore what this update does. Is there anyone that can tell specifics about what is changed in every BIOS release? A small changelog would be nice, even AMD does that, and it's rather hectic to guess what each release does...

2 Years of battling with this laptop... at least I don't get bored of just working on it like any normal being XD

Thanks in advance for reading.

JackE

@FerchogtX >>>Thanks in advance for reading.>>>

Yes, thanks for the informative read. It reminds me of my own three BIOS firmware rules of thumb. (1) If it doesn't seem busted, I don't fix it. (2) If I really can't figure out what an update is supposed to do, I don't try it. (3) If I get hyper about overhyped vulnerabilities, I try to calm down.    Jack E/NJ 

FerchogtX

JackE said:

@FerchogtX >>>Thanks in advance for reading.>>>

Yes, thanks for the informative read. It reminds me of my own three BIOS firmware rules of thumb. (1) If it doesn't seem busted, I don't fix it. (2) If I really can't figure out what an update is supposed to do, I don't try it. (3) If I get hyper about overhyped vulnerabilities, I try to calm down.    Jack E/NJ 

Normally, I never upgrade BIOS, in my whole life using computers I never had the need to do it, but this laptop is another story XD.

Since the first day, I had to send it back in warranty because the BIOS lacked support for displaying colors properly (So they released 1.13 with that fix...), and from that day, I'm battling with unexpected reboots, which seem to be worse or better depending on the BIOS released (for example, 1.19 makes it less frequent, but 1.31 completely borks it, making those reboots at last once per 2 days...) or the version of Windows installed (1511 was instant to reboot, while 1703 seems to try to stand still until it can't anymore and reboots)

I really hope some coders inside Acer read this, because, due to the nature of this reboots, I resist to believe this is hardware related (A failure or physical damage), many think it is, but in that case, the reboots should be random and in any moment, not that perfectly scheduled. And yeah, I made the classic testings: Memtest86, Stress the system, the HDD, the VGA, the CPU... nothing triggers even a BSOD either for failures or high temperatures (this system is rather cool, 30°C on idle, sometimes 28°C, Max 65°C if I max out the processor and/or VGA)

So... I'm assuming the BIOS has some sort of bad coding in the Hypertransport handling part... I mean, if they missed, since the beginning, necessary code for making color display smooth in this LCD, it's safe to think they missed coding to handle HyperTransport and bus management correctly... (Not to mention they missed a small bit to have a little switch to enable virtualization in a more than capable system)

Now... I'll see if any others with this model pop up and have any opinions in this matter...

JackE

I'm not sure if this particular model, like a few other E series models, has a mainboard safety interlock switch that gets inadvertantly activated by an ill-fitting case or keyboard. Usually triggers a random shutdown or reboot by physically moving or disturbing the machine such as by using the keyboard or touchpad. If you care to read thru these threads, you will find some who claim the fix is duct tape and spacers. Jack E/NJ

https://community.acer.com/en/discussion/comment/546665/#Comment_546665
https://community.acer.com/en/discussion/comment/549720/#Comment_549720
https://community.acer.com/en/discussion/comment/572977/#Comment_572977

FerchogtX

Thanks for that, Jack!

In fact, as the warranty is over, I opened this laptop, and I just found the battery reset button and a probably CMOS switch, I leave a picture I took when I was about to remove the whole case:

As you can see, it's right-top from the HDD space, between this and the RAM modules.

The curious part, is that I seldom use the laptop with the battery only, and when I did, I never had a single reboot... All my issues happen when I'm on the AC Line.

Also... I don't know if happens to others, but I recently noticed that utilities like GPU-Z tend to report bad speeds on the GPU, I ignore if it's just me, but, for example, it tends to report 56 Mhz for the GPU clock instead of 100 idle or the 756 Mhz when doing something, AIDA64 seems to report the correct values and even the changes, but, as soon as you try to stress the GPU, after 5 minutes or so, speed goes down by 200 Mhz (756 drops to 600 or 580 Mhz). I notice also temps raising, but not by much, since the CPU, that is an APU is the only thing detected, is the only temps I can monitor, but never passes from 60C°, but if it goes above that, like 62 or 63C°, it drops speed.

Another funny thing is that, if you try to stress the CPU and the GPU at the same time, Task Manager drops the processor speed completely, leaving it to 750 Mhz and the Vcurrent multiplier drops to 0.8V aprox (normally on idle its 0.9 @1.25 Ghz, due to Windows power savings, on high performance it's always on 1.18V@2.90/3.00Ghz more or less, supposedly it can go up to 1.3, but it rarely does that) I'm adding to this that I have been streaming with this machine (Using AMD encoders instead of CPU based ones), and even played games on the Dolphin emulator, but I don't have performance drops or lag, not even high temperatures (Max 65C°) I still use a cooling pad when streaming tho.

This just leaves me rather confused about the system configuration this laptop has XD

Any inputs in all of this?

Cheers!

JackE

>>>All my issues happen when I'm on the AC Line.>>>

That's usually when the CPU & GPU are unthrottled and try to go full-tilt if/when called upon to do so. It's possible the charger isn't quite up to snuff and not supplying the watts needed to run them at full tilt plus keep the battery happy. The stress temps don't seem very high. Jack E/NJ  

FerchogtX

Just an addendum: Latest Versions of CPU-Z and GPU-Z seem to finally detect speeds and temperatures correctly, still, that CPU speed drop happens when stressing the GPU. It's curious it sets itself at the same speed the GPU Core runs... so I'm about to assume this is an intended behavior for this APU.

Speeds for GPU also drops after a while when stressing, just like I pointed in my past post, but no performance drop is shown... it's quite interesting in fact, it's like it's doing some throttling, but in fact it never happens.

I wonder if other units have this features...

FerchogtX

I'm using the original charger that came with this laptop... any suggestions in this regard?

Thanks for your answers, Jack

rosesbyrosex

I have the same model e5 553 1786 with 1.31, virtualization enabled, installed 240 gb ssd + 1tb caddy and everything works fine, if you have problems with hibernation or sleep, reset the bios setings and power plans on windows to default.

JackE

What happens when you unplug the battery and try to run on the charger alone? Jack E/NJ

FerchogtX

Thanks for the answers everyone!

I have reset the BIOS a lot of times, specially after installing a new BIOS, but that doesn't seem to fix the issues.

And, unplugging the battery just makes this run a bit cooler (but is minimal really) seems the system is stable, and all the stress tests work as expected.

By now, after a week of downgrading to 1.19, I haven't had a single crash or BSOD or unexpected reboot... If this continues like this at least for a month, it will be a clue that the BIOS has something bad that makes Hypertransport not work as expected...

JackE

>>>All my issues happen when I'm on the AC Line.>>>And, unplugging the battery just makes this run a bit cooler (but is minimal really) seems the system is stable, and all the stress tests work as expected.>>>

Seems like AC problems only when the battery's plugged in. Not clear that the BIOS firmware has anything to do with this??? Jack E/NJ

FerchogtX

If the AC has something to do... how do I fix it? Changing the charger? (Sounds a bit risky)

Besides, it worked just the same with the battery and without...

Just to be sure...

Event ID 20 - WHEA-Logger

A fatal hardware error has occurred.

Component: AMD Northbridge Error Source: Machine Check Exception Error Type: HyperTransport Watchdog Timeout Error Processor APIC ID: 0 The details view of this entry contains further information.

And a bug check 124... is this having to do with the AC line, bad BIOS or the processor? Because when it reboots, leaves this en event viewer. And just happens after a somewhat fixed time (month or 2 months), in the newest BIOS it was systematically happening after 2 days.

JackE

>>>And, unplugging the battery just makes this run a bit cooler (but is minimal really) seems the system is stable, and all the stress tests work as expected.>>>

No, not the AC charger but the battery. I thought you meant by the above statement that the system is stable and all the stress tests work if the battery is unplugged.  Jack E/NJ

FerchogtX

It just worked the same to be honest, no changes.

any idea why I get those WHEA errors?

JackE

If you haven't done so already, please try to this step-by-step WHEA troubleshooting guide. https://support.microsoft.com/en-us/help/14238/windows-10-troubleshoot-blue-screen-errors

Jack E/NJ

FerchogtX

Well... the only thing I haven't tried is upgrading to the latest Windows Version, but Im waiting until they sort out some bugs that are a bit nasty to deal with...

By now, 2 weeks and not a single reboot since I downgraded the BIOS and returned to build 15063.994 (I was on 1011 or so...) I don't know if either one or both were the culrprit at this point, but I don't wanna investigate either XD.

One question, I tend to stream, and I use the AMD encoder to make things faster, of course, this generates some heat, and I notice that, when the GPU is in intense use like this (and only in this exact moments) the battery seems to discharge while being on the AC line, tryes to recharge, but it does it very slowly, I assume the heat is too much for the batt to recharge at it's normal rate, which can be a problem and can degrade it quickly. Obviously, the disrcharge rate would be faster if I used the battery alone.

I find it like a design mistake Acer did this battery internal only, because it suffers from this a lot, unless, again, I'm the only one...

Any experiences with this particular?

Cya!

JackE

>>>the battery seems to discharge while being on the AC line, tryes to recharge, but it does it very slowly, I assume the heat is too much for the batt to recharge at it's normal rate,>>>

More likely a higher-power output (watts) charger would address this issue. Jack E/NJ

FerchogtX

Ok, that could probably explain those weird reboots (if it has to do with the AC line...)

Do you know how much watts the original charger is supposed to deliver? (Supposedly, it rates 2.2 A, and like 45 Watts, but I'm unsure in this last, since I can's read it from the label)

If the AC charger is not enough, which one would be enough for this laptop to work like normal?

Thanks for the answers!

JackE

Sounds like you got the lower power 45 watter. The 65 watter would be better. Google 65-watt part numbers KP.06501.003, KP.0650H.004 or KP.06503.013. You can also search for the replacement ac adapter using your E5-553 model number for generic adapter which are probably cheaper but also compatible. Jack E/NJ

FerchogtX

I'll look for it, hopefully, that will solve my problem. Thank you!

That means other users that got this same model, have the 64 watt charger instead of this one? I wonder why this one was packed with this particular AC charger XD