NotebookCHECK - Notebook Forum

Have you tried swapping the power supplies? My best guess would be a firmware issue in the XPS 17 - it probably doesn't identify the power supply correctly as their own proprietary 130 W and instead behaves like it's an ordinary 100 W supply.

Thanks, good to see this issue finally picked up in a review.

I did wonder whether the XPS 17s have gone out with a development branch of the BIOS. There's an option in there for dynamic power tuning that has a note that it is for development only and won't be visible to customers (yet it is).

You know you wouldn't build a desktop with a bare minimum PSU so why accept these high dollar notebooks that don't supply a decent charging/power supply? Seems like the higher up in specs you go the more issues we have today with thermals, power, stability. What happened to over building hardware to last? If power through USB C isn't enough then use a proper separate port.

Can't wait for the final review! It's what will make me decide whether to get this or not

Incidentally, I have a 9700 + 1650Ti + FHD model, and I have detected the AC adapter reaching and staying at 127W at the wall when gaming.  Yet others with the RTX 2060 or Quadro 3000 can't get their system above 107W at the wall.

It seems that the issue is the AC Adapter isn't providing as much power to RTX users as it does to GTX users.

I've seen a handful of reviews about this laptop and I don't think I've seen even one where the reviewer didn't have to return it over some broken hardware issue, be it the trackpad, the speakers, or something else.

How can you get a laptop of this price so wrong?

I was considering importing it from the US and giving up on the warranty. I can kiss those ideas goodbye now. That Honor 16.1 is looking very attractive, though.

Quote from: _MT_ on July 10, 2020, 10:29:14
Have you tried swapping the power supplies? My best guess would be a firmware issue in the XPS 17 - it probably doesn't identify the power supply correctly as their own proprietary 130 W and instead behaves like it's an ordinary 100 W supply.

When plugged in, the BIOS recognized the charger at 130W. Didn't play any games or haven't fires up Premiere to check.

Allen this is an awesome report and thanks to bring up this to Dell, instead of having the users deal with Dell where from Reddit's report, Dell level 3 technical representatives is so lack of debug ability. Again, wonderful review comes from the creditable reporter! We will wait for the update.

Quote from: JohnIL on July 10, 2020, 11:45:20
You know you wouldn't build a desktop with a bare minimum PSU so why accept these high dollar notebooks that don't supply a decent charging/power supply? Seems like the higher up in specs you go the more issues we have today with thermals, power, stability. What happened to over building hardware to last? If power through USB C isn't enough then use a proper separate port.

If you know the draw, why wouldn't you? An engineer should be able to size it correctly, specify what he wants, test it, have access to detailed datasheets and engineering support. When you buy a power supply, you usually know squat about it beyond some marketing fluff and perform no characterization (for which you have no equipment). Similarly, you often don't know how much the components can draw and you don't bother to characterize them either. Sizing a supply can then resemble black magic. By you I mean a typical consumer. I'll tell you a secret. A well designed supply can supply the rated continuous power... continuously for whatever time frame it was designed to, under the conditions it was designed to. If it couldn't, it wouldn't be well designed. And more. It's going to handle some overloading (depending on duration - again, as per specification). You really don't need a 200 W supply for a 130 W load. And we're talking about laptops. More powerful supply is going to be bigger and heavier. Which is undesirable. You only oversize when the offering on the market doesn't meet your demands. Which doesn't apply when someone is building a supply for you, to your specification. Headroom should already be built into the supply. And you have control over BIOS and configuration, meaning you control consumption. When we oversize supplies, it really is more a show of our lack of knowledge and control. It's a different game.

Problem is that thinness and lightness are seen as important. The higher you go, the more importance they have. An expensive laptop has to be stylish. And these traits go inherently against overbuilding. Also, many people don't really need the outright, raw performance, especially sustained performance. So, it's something they can compromise on.

"Dell has responded to our inquiry saying that the XPS 17 "can pull more than 130 W" due to the onboard Core i7 CPU and RTX 2060 Max-Q GPU even though its PSU is rated to deliver 130 W. "

Dell completely misunderstood the bug. The "130 watt" adapter is only providing 100 watts max. This is absolutely a USB PD negotiation bug. The XPS 17, with 1650 Ti, has been measured to pull about 130 watts from adapter. It  is not capped at ~100 watts like the 2060 / 3000 trims. The trigger for the 30 watt deficit bug seems to be the 2060 / 3000 GPU and its inclusion in the power requirement request through the USB PD controller.

What Dell thinks you are asking is whether we see battery drain when adapter is providing the proper 130 watts at the wall and XPS 17 is consuming more than 130 watts. In this normal case we would expect the battery to run down. However, this is not the issue we are talking about here.

If this issue is not clearly articulated to Dell it will never be fixed. For some reason this problem is confusing people. Even within the reddit community many people misunderstand and incorrectly interpret the problem the way Dell does. Even reading the comments on this article here shows a clear misunderstanding of the issue. Please push back Allen. 

I am an engineer and I have the new XPS 17, I may run the CPU for a total of 1 hour per day 'full bore' the rest of the time it is waiting for me (I type way slower than any 5 GHz CPU), and it has never needed to discharge the battery to keep up. So for me and my usage I think the small 130 watt power supply is the proper choice (as it was with the XPS 15 / 5520 that I also use), I don't want to carry around a 250 watt brick. Everything is a tradeoff and for this one I am happy with Dell's decision. I mainly wanted a small chassis, solidly built 17" XPS 15 equivalent and dell has delivered.

Thank you Allen. Your 2pm (second) update from Dell correctly captures the bug. Please keep us posted and keep up the excellent high quality content.

I honestly don't know what people were expecting. I could see this coming miles away. It ain't rocket science. Here are some simple maths. Assuming that both the CPU and GPU perform up to their rated TDP.

Any CPU config - 45W
2060 Max-Q - 65W

That's already a whopping 110W, so the system only has a budget of 20W left from the charger. And now it also has to power the 4k screen, the motherboard, the fans, the RAM, the SSDs, the DAC/AMP, etc. Ask yourself, what the hell can it do with 20W? Of course it has no other choice but to tap into the battery.

Anyway, this is not the first time they did this either. For example, the base config Dell G5 gaming laptop with 1050Ti from 2018 only shipped with a 130W barrel plug because they decided to cut cost on that part (shame on you Dell). As a result, battery drain was quite insane when both the CPU and GPU could boost properly. The best solution for that was to shelve out extra money for a 180W barrel plug instead.

Quote from: Tony Shark on July 11, 2020, 01:27:09For example, the base config Dell G5 gaming laptop with 1050Ti from 2018 only shipped with a 130W barrel plug because they decided to cut cost on that part (shame on you Dell). As a result, battery drain was quite insane when both the CPU and GPU could boost properly. The best solution for that was to shelve out extra money for a 180W barrel plug instead.

Exactly the same thing happened with Lenovo Legion Y530 i5/1050ti powered by the 130w power adapter. I don't know how/if lenovo addressed that issue though. Higher spec configurations had 170w psu and had no issues with that.

Please try the 130W AC adapter that came with the 9500 to see if it exhibits the same behavior.  It would be nice to be able to rule out the 9700 AC adapter altogether.

Time to praise Notebookcheck.

Thanks NotebookCheck for investigating this! It'll help a lot of people and make Dell solve the issue much faster.

Quote from: Veyron on July 10, 2020, 15:34:29
I've seen a handful of reviews about this laptop and I don't think I've seen even one where the reviewer didn't have to return it over some broken hardware issue, be it the trackpad, the speakers, or something else.

...

I think the speaker problem you mentioned could be solved updating the driver, so it was not a hw problem

To be honest, if you knew the amount of power the you need when designing a system, you should have put adequate power source for it. If you knew it at first place but didn't make it right, that is an obvious engineering mistake, isn't it? If the compromises had to make, then don't provide RTX option (or other power hungry stuff) so that the total power can be confined to a given source. Even if we are not the dell engineers a simple math can get to the bottom of this issue. The engineering decision Dell made for XPS 17 9700 is a mistake.   

Again, praise for Allen, the editor in chief of Notebookcheck!

QuoteI have yet to see anyone comment or review the ports on the XPS 17 9700. I have found posts that the RTX 2060 GPU is directly hardwired to the displayport outputs. Its 9700 capable of supporting DisplayPort 1.4? Dell spec sheets show that it supports Displayport 1.2 but I'm hearing that in the BIOS you can activate the RTX to bypass to Intel GPU. Would this result in supporting 1.4? Allen are you able to determine if the 9700 supports 1.4 data transfer?

I have yet to see anyone comment or review the ports on the XPS 17 9700. I have found posts that the RTX 2060 GPU is directly hardwired to the displayport outputs. Its 9700 capable of supporting DisplayPort 1.4? Dell spec sheets show that it supports Displayport 1.2 but I'm hearing that in the BIOS you can activate the RTX to bypass to Intel GPU. Would this result in supporting 1.4? Allen are you able to determine if the 9700 supports 1.4 data transfer? Can I connect a DQHD monitor at 120 Hz?

I've just been playing Hexcells Infinite. It's a 95MB puzzle game. My battery drained whilst playing even though I was plugged in to mains. If that doesn't say something is very wrong with the XPS 17 I don't know what will.

It's obviously going to be a very niche an impractical workaround, but does using the XPS 17 with an eGPU go anyway to preventing battery discharge during sustained loads?

I'd be interested to see what difference an eGPU being output back to the XPS' own screen and also an eGPU with output to an external monitor makes regarding it's power draw.

what about the Precision 5750, the pro version, there are even higher configs possible, so I guess it was a bad idea and they should have used an extra power plug and a normal 230 W power supply instead of this customized TB power plug solution.

Another thought: this may be the reason why Dell batteries are worn out pretty fast in general? too small power supplies

However, could maybe all be fixed with a BIOS update... ;)

Quote from: Tony Shark on July 11, 2020, 01:27:09
Anyway, this is not the first time they did this either. For example, the base config Dell G5 gaming laptop with 1050Ti from 2018 only shipped with a 130W barrel plug because they decided to cut cost on that part (shame on you Dell). As a result, battery drain was quite insane when both the CPU and GPU could boost properly. The best solution for that was to shelve out extra money for a 180W barrel plug instead.

As far as I know, it's a standard feature. It has been around for at least 15 years. It allows you to, for example, take a 200+ W workstation and power it from a 12 V travel adapter rated at something like 65 W. As long as you had juice in the battery, it could use it to supplement the external power supply and provide more performance. Or to use a weaker supply in a pinch from a different laptop. Also useful for docking stations. Not to mention, laptops are typically designed to be primarily used on the go, powered by battery. Weaker supply then simply means slower charging. It's not preventing you from using a laptop as a laptop. It allows you to use whatever supply you happen to have on hand and make the best of it. This is exactly how it should be.

Yes, they sometimes leave it up to you to choose the appropriate power supply. Again, nothing new. Even 15 years ago, there were options to buy more powerful "bricks" for many laptops, even high end models (like top of the range Latitudes, Precisions). I believe my first Latitude came as standard with a 65 W supply and I opted for the optional 90 W one. It wasn't much and it was nicer (it was nice even by today's standards, designed so that you can nicely wrap the cables around it with its curves rather than edges, double insulated with a two pin plug rather than the annoying three pin, it had an integrated elastic cable strap). I believe there was an even more powerful supply listed in the configurator, but it was more of a brick.

In this case, I believe they really want to use USB-C for charging. It's meant to be a MacBook with Windows kind of a machine. Form over function. And they have to live with the limitation. But that's not really the issue here. The issue is that a 130 W supply isn't providing 130 W. That's not intentional.

Quote from: Sterlinger on July 13, 2020, 08:26:11
what about the Precision 5750, the pro version, there are even higher configs possible, so I guess it was a bad idea and they should have used an extra power plug and a normal 230 W power supply instead of this customized TB power plug solution.

Another thought: this may be the reason why Dell batteries are worn out pretty fast in general? too small power supplies

However, could maybe all be fixed with a BIOS update... ;)

It's not a full size workstation. It's an ultrabook on steroids. We'll see how it performs but I expect it to be quite thermally constrained. There is little point in bundling a 200 W supply if it can't cool 200 W of heat. Not to mention that it's funny when you have a laptop that's supposed to be very thin, very light and then you have a beast of a brick to go with it. If you don't mind traveling with a brick, you can buy a full size system.

I assume there has been no update from Dell since 11th? I'm beginning to assume that they hope that this thing just blows over with time...

I want to order one with the 2060, but i will w8 to see what dell says after reading this. Don't think i can leave with this when i pay this premium.
CHecking notebookchat daily to see if there has been any updates.

This is a deal breaker for me... I was so exsited about a proprer 16:10 screen... Its a pitty that the 7750 doesn't have the samr 16:10 screen

A user on Reddit supposedly has a unit that does not have this issue. Apparently I can't post links, but you can start with the Dell subreddit and then add this string to the end of the URL:

/comments/hsi2yd/dell_xps_17_battery_drain_must_read_now/

However, Dell's own support page on this issue, updated today, makes no mention of the wattage discrepancy. Again, can't post links. but the article number is sln322189. Searching their support site will take you right there.

I cancelled an order over this. I would love if they would make some kind of official statement about the results of the investigation in their last update.

Can you, Allen or Notebookcheck, confirm that you have not heard from Dell since July 11th - and no official statement from Dell?

I was informed by Dell yesterday, after two weeks of escalation and sharing this article and my own findings (my unit only pulls 106W at the outlet, uses an average of 125-130W under load, drains about 15 % an hour), that the laptop is operating as designed and there is nothing they can do. I don't accept this response.

That is, the limited akku-cycles are counted down.

Quote from: voltoptimus on July 19, 2020, 19:05:13
I was informed by Dell yesterday, after two weeks of escalation and sharing this article and my own findings (my unit only pulls 106W at the outlet, uses an average of 125-130W under load, drains about 15 % an hour), that the laptop is operating as designed and there is nothing they can do. I don't accept this response.

Ask Dell why all xps 15 and some 17 is pulling 130-135 watts from the adapter?

Does anybody know if the Dell thunderbold Dock WD19TB with 180w power is a solution? Or is the USB C Port limiting ?

is it a Dual-C Thunderbolt Dock? I have no idea, is it even possible to use two TB ports parallel for charging / as power supply?

To reply to Dennis07 : I have the WD19TB - 180w power adapter and the battery drain is also present with this USB-C Docking station.
Also plug two adapters the WD19TB and the 130w default adapter don't change anything because only the first plugged adapter is recognized.

Quote from: _MT_ on July 10, 2020, 10:29:14
Have you tried swapping the power supplies? My best guess would be a firmware issue in the XPS 17 - it probably doesn't identify the power supply correctly as their own proprietary 130 W and instead behaves like it's an ordinary 100 W supply.

definitely has to be firmware related, unlike the BS i just read at pcworld: "Just got off the phone with support. CONFIRMED: the XPS 17 battery drain is a HARDWARE issue and NOT a firmware issue," HistoGraham wrote. "

have a 5576 here, replaced 130W with 240W, overclocked GPU pulls the most. no way to get more than 140W from wall with battery in, if i detach battery it goes up to 220W from wall.

in case of most dell laptops it is bq24780S in charge and there are suppose to be SMBus way of disabling hybrid power mode. at least one thing i could do via some dell battery software is prevent it from charging until next reboot - sure other commands that charger IC specifies in its datasheet would have an effect too

Won't allowing the battery to continually discharge also impact the longevity of the battery itself? 

I have these issues with the Dell XPS 17 9700. I got a Dell technician to change my motherboard, but the issue has not been solved.
There is a next step that seems to replace the motherboard with another one that will provide 25W for the processor instead of the original 45w, lowering the power of the processor to deal with tasks... unacceptable.