S-XPS 1645 AC Power Throttle Issue Investigation

Hardware replacement got some +++

For £1300 I'm planning to complain about every single problem that I will have (like dust spot under the monitor - replaced), noisy mainboard (will be replaced) hot power supply (will call later) and the most important one - Throttle Issue.

Each time I will ask engineer to come and fix the problem as I've spend £160 extra for the extra warranty.

Bonuses are that you laptop will have new parts every complain, sometimes if your complain skills are very good you can get a new model

 

No clue how a mobo can be noisy. No moving parts in it, only moving parts are ODD and HDD.

Hot power supply: they normally get hot when charging.

 

Ah, I see. Why must Dell do things backwards? I should've went with a Toshiba Qosmio >_< I chose the SXPS16 for a aesthetic appeal + performance but I'm not really getting either. The computer is a fingerprint magnet and makes it look fugly unless you wipe it down frequently, and the throttling issue is killing performance when you need it most.

 

Hi All,
I’ve been following this forum since the days of the “Original Owners” thread. I’m still reluctant to purchase the machine as I’m assuming it’s not “Just Gaming” issue. I’m mainly going to be editing HD Video’s and hence my need for the combined power of the i7 CPU & GPU acceleration (both supported by software). Has anyone tested this? I’m assuming it’s easier to note issues while gaming (fps). However, I’m assuming this issue might (or most likely) exist in intensive HD Video Editing.
Am I wrong / correct? Would love to receive some feedback on this with an emphasis on HD Editing and DVD / Blu-Ray authoring.
Happy Holidays.

Edit: I forgot to mention I do need raw speed. With that in mind, I was considering the 820 CPU, 8GB memory, RGB Screen and an SSD.

 

I think Dell rushed the i7 version of the SXPS 16 as a modified layout of the 1640 and didn't account for the extra power usage under load. In any case I'd choose this laptop over Toshiba any day (no offence) since Toshibas, like Acers, seem to be bigger to get the same raw performance.

 

that's why I want to cancel my order. The longer it takes for them to come up with a fix, the more it looks like a hardware design issue. And this is something they could fix silently with a new mobo revision. So best case scenario you'd have to beg them to send you a replacement machine.

 

You look like DELL's worst nightmare I hear you...Been there done that.

 

I'm working as IT support and I've never saw so hot laptop power supply.
I cabn bet I would be able to make fried eggs on it.
Also I'm not charging my battery as I've disabled tat option in BIOS.
My mobo is making high pitching noise when it's idle.

 

I agree and the designs are just kinda ehh but I'd choose performance over a beautiful machine that doesn't do what its supposed to. Here's hoping for a fix. The Qosimo is massive and that's half the reason I chose Dell.

 

I won't be begging. They are giving me a replacement free.

Just like how I did not beg for this free 1645 worth $3400 CAD when my original laptop was a 1640 worth $2400.

 

Wow that does soound weird.

 

If the hardware does indeed need to be replaced, i have a feeling we might be waiting a while. As hardware revisions probably dont happen over night. Then the associated testing that goes with it after a revision is done, then forwarding to manufacture. Could be a while if that truly is the case.

Im not sure what to think, it seems like dell knows more then their letting on. At least thats what i get from talking with them etc. I cant seem to shake the feeling theyve known about this from the start, and possibly the delays they had shipping this laptop is somehow related. Maybe thats just my paranoia but its not all adding up, as the throttling seems very deliberate

 

So I suppose the "safest" thing to do right now is to use the 90W adapter or run off the battery.

Does anyone know whether the 90W adapter gets extremely hot if the battery is removed? I'm talking about normal usage, not stress testing.

 

No. Even if battery is in it, if the battery is fully charged, the adapter doesn't get warm to the touch.

I've found my adapter only gets warm when it's charging the batt.

 

No, you can't upgrade video card, simply because it's not mechanically possible. However, if you're in the US, there's a law prohibiting the manufacturer from voiding the warranty if you put your own parts in the laptop.

 

Never heard of that law, what is it? If it existed it would probably be at the state level anyway.

 

The reason you can't upgrade the video card in the XPS 16 is because there isn't one. The Radeon 4670 is integrated onto the motherboard; there's no removeable card or slot available. Some other laptops use MXM slots and cards that allow replacement and upgrades.

 

Really, i am sure mine is always scorching hot!! Anyone else notice adapter heats up only when charging the battery?!

 

All you people posting "I want to cancel, should I do it?" and the conspiracy theories are really killing / have killed this thread completely.

I have zero issues with this laptop with the 130watt and lowered brightness (which is my desired brightness level anyways) so I'm happy to wait for a fix or new revision.

If you're going to cancel, do it. Quit talking about it and just do it. The people posting theories as to what the fix is (Motherboard revisions and so on) arn't helping either, as its not anything that we can do, test, or prove. You're not contributing to the thread at all with these posts so either post some benchmarks or new information regarding actual facts or don't post at all.

Thanks.

 

yeah its a bit to late everything is just buried quick by offtopic comments.. I do want to see someone run the intel burn benchmark and see if they have a 12x multiplier. As thats just a cpu test and even my system goes down to 7x unless i dim the screen(pulls 90+ watts), where on battery its full 12x. Seems that test may be more stressful then the prime95 and really shows that the cpu alone can put this system at the limit through the adapter.

ya really should put a lot of the results (power/bench wise) on the first page if ya can change that, or we just need to remake a whole new thread as the results are so long buried no ones looking at them.

 

Yeah the front page isn't very informative. Should create separate sections highlighting fact and fiction, i.e. our test results and the theories about why it's not working and conversations with Dell reps. And then talk about various ways you can workaround the issue. Also we should encourage people to call up Dell and report the issue so we can get this prioritized higher.

 

Wasting your effort guys. This thread is trashed. A new results thread will also get trashed. Shame, have followed it for quite some time now.

 

Amen to that as I feel the same way and am very happy with the SXPS 1645. As far as I am concerned I also have zero issues as well.

 

Out of curiosity, what are you guys doing with your laptops to conduct so much heat on your system and/or power supply? I can run Dragon Age or WoW for 5 hours+ straight at mid-high to high settings and I'm certainly I can't fry an egg on mine.

I have the standard set-up with the 720 processor (I know some have the higher models now, this was not available at the time of my purchase nor do I desire it at this time) although one differentiation appears to be that I have the WLED versus RGBLED. Again, this was a financial issue and I've been more than happy with this screen versus my old Toshiba.

I do notice the throttling, and I do notice FPS/performance increase with lowered brightness on my 90W. I just received my 130w today so I'll see if there's any improvements. On that given note, does anyone if I need to return the 90w? The instructions aren't clear on paper.

 

Yes, you do need to return the 90W Adapter or they will charge you for the 130W adapter after 10 days. That is what it is saying in the shaded area at the bottom of the invoice, or send back the 130W.

You can try to speak with tech support to see if they will allow you to keep both as you will be testing it until there is a final resolve, maybe they will allow you to keep both, its worth a try.

 

I'm surprised they expect you to return the 90w adapter. My XPS M1710 psu died a few months back (just before the warranty expired); they just sent me a new one and didn't want the old one back.

Back onto the main topic though... from what I've seen the Studio 1747 isn't having this issue, and I would have thought that the mobo is fairly similar in that system (I know it's got a 4650, but it is meant to draw the same amount of power; 35 watts). That system also has a 90 watt PSU. In addition, I would have thought that a 17.3" screen would draw more power than a 16" one also... perhaps the issue does exist there, I've just not seen any reports of it so far, but I'm throwing this in as it seems rather strange.

 

not even going to bother. the facts are out, everything else is pure speculation.

I called. I got my 130watt adapter. I filed my BBB complaint and received satisfactory compensation for the hours and hours i spent on the phone and the hassle i went through and so on. Theres almost 100 pages of information that people can show Dell to help them get support. There were only 10 pages when i got my adapter mailed to me. Some people are gonna just have to try again and not accept no as an answer.

The larger power brick helps a little bit, but the meat and potatoes of the problem is going to be some sort of fix that only dell can provide. Threads done

 

said the poster with 1 post

 

I agree, Heres to hoping DELL fixes this quickly *holds up beer*

 

We would have to see the same tests run on our machines vs theirs. its possible they arnt having issues, but they probably are if its the same 90w we suffer from. I have evidence even on my low power system (wled etc) the 90w isnt enough to to keep up the cpu alone when at 100% using the intel burn test. Its been said before but the rated wattage isnt exactly what it will use, i would suspect the 4670 would probably still use more power in the long run, but pure speculation. Anyway they need to confirm their not downclocking like the 1645's are, very likley they may be and just not as apparent as the 1645's.

 

@ atlstlang

You seem to have the same configuration as i have ordered. How badly is the throttling affect you?

 

The gaming works decent, there is throttling going on though. It depends on the game a lot i think but for the most part its playable, unlike other systems. Like ive said before the laptop other wise is fantastic and its only under those high loads that any problems actually arise. The adapter does get mighty hot when charging/gaming for bit.

 

Can somebody confirm that with the new BIOS they didn't down clock the CPU?
Plus, are there any new tests after the update and how do they compare to the older Bios version?

 

Do it yourself, seeing as you have a 1645.

If you don't like the BIOS just flash A02 back.

 

It doesn't seem to have any effect on CPU throttling, but there are lot of changes in this BIOS (from a basic diff between A02 and A03). I might get hardcore later and disassemble it and look at the difference between the two.

 

I don't know if I should answer you but anyway I would be happy if I have the laptop. Its supposed to be delivered on Monday.

 

Oh, my bad! lol

I just saw your sig and thought you already had your system in hand

 

dumb question:
kinda @ atlstlang since we have similar builds

How much do you have to turn the brightness down to stop the throttling? Like ALL the way down or just like 2 notches?

 

Hey ppl, Just saw that the new BIOS A03 is out. Can anyone confirm does it solve anything? I still dont have the system so I cant and I am just curious

 

I dont get major throttling like others more like half throttling, but its like in the gray area while playing certain games. You can tell because every time it throttles all 8 multipliers switch to like 9.00 at the exact same time and stay there before going up or down another 1.000 before returning to normal operation. While when under normal operation each thread is all over the place with like 13.534 etc.

Now if i run the stress test it pulls so much power it doesnt matter, but i just run the cpu stress test (intel burn test) it will throttle at max brightness, and be just fine at 12x on min brightness. Due to it being at the very edge. so its highly dependant on what you do.

The wled doesnt draw as much power as the rgbled, and seems to have little effect, through the adapter i see 5w maybe difference between bright and low. IMO very little difference between bright and low, as when its throttling its probably to far over the limit anyway. so i usually just run max brightness.

edit: bipul i did quick benchmarks it doesnt solve the issue, but it is unkown if it may have helped some although it doesnt seem the case from feedback thus far

 

Hi all,
This is sort of off topic but I was trying to install the new bios update (even if it doesn't fix anything) and I keep getting: access denied: error code 5.

Has this happened to anyone?
I am using an administrator account and tried shutting down and retrying but always same error.

A second question actually on the issue of this form - is it worth getting the 130W adapter or might it actually damage the system as some people seem to be indicating on the form.

thanks for any help!
_________
studio xps 16: i7 720qm, ati 4670, 1080p R G Bled, 4gb 1333mhz ram, 320 gb HD, BD/DVD combo drive

 

bios installs just like normal here.

IMO the 130w wont really hurt or do anything besides run cooler then the 90w. I think its just due to fear from the dell reps as its not spec'ed for the system. As the voltages etc are identical, and the system seems to limit the power draw to 90w regardless if a 130w adapter is used. 130w may be able to help some stability for some, but as mentioned it doesnt draw over 90w because of the system limiting itself so doesnt really solve the problem we are seeing. what is nice about it, it wont light your house on fire from the laptop charging normally, it stays rather cool. (the laptop tends to draw the full 90w when charging)

 

Think about it this way : when on battery, the system has enough power, meaning that it uses more or less than 130w, meaning that the system has been designed to support this amount of power. So why an AC adapter would be different if it delivers the same power ?

Plus the fact that components such as the mobo/cpu/gpu respects some standard specifications that define the maximum power consumption of such components so thus they may be designed to support such a power.

If the system don't support such power, Dell would already have replaced every batteries.

As said, I think the reps have been breifed to say it may damage a system if an unlisted component is used. It's just a security to prevent from advicing an end user to do so and potentialy void its warranty by following a Dell rep.
They prefer to say this instead of "we'r not sure".


So, I think that unless you feed your system with an overvolted AC adapter, you can't dammaged it and the more powerfull the AC adapter is, the more cooler it will run.

 

I'm jumping in late to this coversation, so please forgive my ignorance. First, I was wondering does this effect both 1640 and 1645? Also, if it is, then how do I determine the difference between the two adapters? Finally, I see that the latest discussion seems to be so-so on actual improvement and possible heat issues. Is the risk worth the reward?

 

@Helmut: I don't think anyone of us understand the workings of the motherboard well enough to identify that in fact the AC power draw is limited strictly by software or firmware. The possibility that the 90w limiter exists at the point of connection, or between that point and the motherboard, points to an alternate scenario to the one you pose, in which a hardware limiter is causing the draw to be capped., while the battery may benefit from a different concersion apparatus, or may not even have to pass through one to supply power, and therefore offers the laptop all the power it requires.

To simply write this off as a coverup or conspiracy does no justice to the actual cause of the issue, which may simply be an inopportune power limiter on the AC plug in point. It could also be the power converter (DC to AC) as well, which could explain why some European users have seen better reults with the 130w.

 

i think some configs require more power than others, who knows why. i know, that i switched to 130W (was the first to get it), and my framerate drops to 1fps, and audio stuttering, went away asap when i unplugged the 90W, and also is not present with the 130W either.

 

have you checked to see if you still suffer from throttling though?

 

i don't recall if i'e tested that. i'm a wall and ceiling contractor, and i got really busy all of the sudden. i just ran prime95 while i was downloading furmark, and my multiplier was changing from 12 to 13 which is 1.6 - 1.73....which sucks, since i have the 820qm, and 1.73 is supposed to be my base.

so when i run furmark, with prime95 running, my multiplier in i7turbo/RT goes stays at about 9, and it doesn't go back up until i close both apps, and then restart furmark. then it goes back up to 13 (my lowest supposedly) and then down to around 12 or so.

that is with the 90W, im going to go grab my 130W from the living room right now, and see if it fixed my throttling issues.

 

try running the intel burn test, like the same thing as prime 95, even on the 130w my system couldnt handle it that intel test, unless i dimmed the screen it would work. thats with the 720qm.

But yeah the big problems come when the gpu and cpu is going at the same time. I run game in a window and see like half throttling going on when playing games, around 8-10x mult, when on battery its around 18-20x.

 

TriangleLaptops doesn't seem to have the 130W anymore. Any other options? I don't want to get rapped by ordering from Dell.