Video: Nexus 7 touchscreen defect | Hacker News Reader

Video: Nexus 7 touchscreen defect

geek.com

74 pointsukdm51 comments

I've also had issues with a slightly unresponsive touchscreen on my Nexus 7. Touches that (based on smartphone experience) SHOULD have registered, don't.

Separately, and more seriously, this morning I flipped the screen on (after using it for a while with no issues), and the screen was completely washed out to the point where you almost can't see what's on the screen [1]. I obviously tried adjusting brightness right away, powering off and on, etc. Nothing. This seems fairly widespread [2] (lots of reports on forums etc., then again a lot of people bought it and only people with issues complain). I will be trying to tighten up the screws inside, and making sure any connectors are seated properly, but sending it back for a replacement if it that doesn't work.

[1] like this: http://www.nexustablets.net/forum/nexus-tablet-site-news/650...

[2] more reports here: http://forum.xda-developers.com/showthread.php?t=1749849

I have noticed the problem with the edge of the screen not being responsive since I've had my pre-order one since 17th.

My own gripe is the ghosting/flikering that was demonstrated earlier this week or last last week. When I turn the screen on and start reading Currents, as soon as I scroll to a new page I will see the screen flicker a little bit. This is after the screen gets refreshed with new content.

Umm reading through what I wrote makes me thing it is the auto-brightness changing.

The Nexus One used to have touchscreen problems as well: http://code.google.com/p/android/issues/detail?id=6296

Hopefully Google/ASUS will respond well with fixes. It really does sound like a heat-related issue affecting the digitizer.

I don't think this is heat related as switching the screen off and on again clears the issue.

Looks like a software defect which will hopefully be rectified with a hastily developed update.

A software defect the affects the same portion of screen regardless of orientation and that is temporarily fixed when physically turning the screen off? I'm not convinced.

Could be software in the driver layer that doesn't care about logical orientation, and data structures that are reset when the screen is reset.

>I don't think this is heat related as switching the screen off and on again clears the issue.

I'm talking based on a very rough understanding of how capacitive multi-touch works, but it seems plausible that on restart, the screen calibrates itself for the current environment, "zeroing" itself according to the capacitance it is currently seeing, and that that zero value is no longer accurate once the temperature changes significantly.

As someone that works on projected cap screens, I can say that you're pretty much on the mark. You normally take a "baseline" or background read of the sensor's capacitance and then use the signal level above that baseline to determine the presence of a finger.

Now if you don't keep dynamically adjusting that baseline to the environment, and there are many different approaches on how to do this, you could start seeing false touches or no touch at all.

That being said, every modern manufacturer has algorithms like this in place. So another place to look is construction. When you see problems on the edge, is the flexing of the screen causing your glass to delaminate from the sensor line? Perhaps power cycling fixes it (for now, because of dynamic baseline recalibration), but eventually the circuit could completely fail from a mechanical point of view. I'm curious to see if the problem grows and if it eventually results in failed units that software can't recover.

{edit} Looking at the teardown, Nexus7 is using an Elan controller. Out of all the teardowns I've seen this is the first I've seen from them in a major product as opposed to TI, Broadcom, Atmel, or Cypress. This might get interesting.

Interesting - I have an iPhone 4S, and I've noticed that if it's in my pocket, the touchscreen is really unresponsive. I get a call, pull my phone out, and it takes several tries for the "swipe to answer" gesture to succeed (usually the iPhone will drop my swipe about halfway through).

The incoming call probably makes this worse, when I am not in a call I don't recall ever having trouble unlocking my phone with the same gesture.

Was the screen facing your body when it was in your pocket? You just had a very large capacitive body next to the screen for an extended period. What you described would make sense if the iPhone's screen recalibrates itself.

Yes, I always keep it facing inward to protect it should I do something like walk into table corner.

It's a bit of an annoyance, but thankfully is pretty temporary. I've gotten close to missing calls before, but haven't actually missed one.

I have this exact issue with my HTC Incredible (great phone!) occasionally. I always keep the phone in my pocket with the screen towards my leg.

This appears true. I have a Nexus S and I have seen some times that the touchscreen freezes after I finish the call. It usually happens on a hot day and if there is sweat on the screen. When I press the power button and unlock again, things are back to normal. Have checked with some of my friends who use iPhone and they do not seem to have experienced such issues.

Digitizer as in you could use pressure sensitive stylus to draw rich sketches.

I know I can't post on geek.net, but since apparently one data point is enough to ruffle feathers, my iPad had a non working rocker switch on the side ( the one that you can use to lock the screen or mute the tablet ). I had to send it back to get replaced. At the time I told everyone they should be concerned, clearly one failure indicates a real problem with the device.

I have this same issue. I think it's related to CPU and runaway processes in the background. After installing a CPU monitor I can pretty easily see the correlation between CPU usage and degradation in the touch recognition.

I'm not familiar with Android's architecture, but it seems like they need better safeguards against runaway processes.

Interesting. Can you come up with a hypothesis that would explain why only a specific portion of the screen would be effected?

Edit: anothermachine has a workable story for how it could be a software problem elsewhere on this page

From the video I thought it was going to be more of a hardware issue - eg the hot chips are on that side of the device, and the heat interferes with the touchscreen.

I wouldn't expect, though, that turning the screen off & back on would significantly lower the heat within. Leaving it off for a minute might, but the quick cycle he demonstrated in the video wouldn't.

I'm not sure I buy the heat hypothesis either. I encountered this issue very quickly when running a Tegra game: Tegra games aren't particularly demanding, but rather they exploit additional shader features. Other games and activities that heated up the unit substantially, however, did not cause these touchscreen issues.

Yeah that's when I changed my mind to "I don't know if it's a hardware or software issue now"...

> I think it's related to CPU and runaway processes in the background.

I don't think it's that.

I got my Nexus 7 at Google I/O. The first one I got suffered from this same issue right from the very beginning. I could barely even get past the welcome screens, because I couldn't type certain letters on the keyboard. It was worst at the bottom of the screen (which would be the right side in landscape as the video points out).

The device help desk replaced it for me the next day. Unfortunately, the second one had a different problem -- the display was badly corrupted, and the left half of the screen (in portrait mode) was entirely red.

The third one seems to be working fine, apart from one annoying dead pixel.

I got my preorder tablet in the mail last night, and didn't see anything like this over 4-5 hours of broad, "try everything" use. It might be an assembly or component issue with early tablets. Both the linked article and the grandparent seem to be arguing that it's heat related. Mine certainly got hot, though the sample size is too low to make a determination.

I'm with you. Got mine a few days ago and have noticed zero issues like this, even while running games with the tablet flat on a bed.

Seriously? From a single data point you deçide to cry wolf? I say this as an iPad user.

I'm not an iPad user and I really love my N7 but I notice this lag occur a few times yesterday. I suspected that it was due to high CPU load but didn't have the time to investigate it.

It actually occurred while I was showing off the N7 on the DC Metro to an interested rider. I feared that this out-of-love iOS user may have been deterred by the momentary lag even though the tablet is otherwise awesome.

I hope that this gets fixed in 4.1.2; Project Butter has made Android significantly smoother and I think this is just a rogue element in Google's aforementioned "War on laginess".

Seriously? Given video proof, and without bothering to check for multiple supporting testimony from other affected users, you make cheap irrelevant accusations?

Sir, did you read the article? The author makes no mention of even anecdotal evidence of problems beyond his single Nexus 7. You don't write an article about a single defect instance. A post in a forum maybe, and after there is evidence of the problem being widespread you write an article. Perhaps I should write an article about the rotten apple I bought yesterday and title it "Whole Foods Apples Rotten".

The evidence that appears herein is after the fact, and even so is hearsay and subject to selection bias.

It's funny how well your comment applies to the article author, yet you attack me.

I get the feeling that you only read the article and didn't watch the video because in about the third sentence he says "There have, however, been some sporadic reports of problems with the touch screen, and it turns out, the unit that I have received does seem to be affected. So, I thought I would show you what some of these problems look like."

He is obviously seeing other reports of problems and is presenting his own as a representative instance.

As a former Android user, I say that a history of responsiveness issues on Android devices makes me particularly sensitive to issues like this popping up on the latest hot new Android device.

Responsiveness is a big feature that pushed me onto iOS devices. Kind of ironic, given that it's the other way around on the desktop for me: my Linux machines are nice and responsive even on modest hardware, while my former MacBook Pro beach-balled and beach-balled. It took going entirely solid-state storage on the MacBook Air for OS X to stop driving me insane with responsiveness issues.

If turning the screen off and back on rectifies the problem, then hopefully this could be resolved with a patch.

The nexus one (or at least some) had a similar issue. After a while the touchscreen became almost unusable. Turning the screen off and on again fixed the issue, but it was annoying. Seemed like a hardware defect, because it never got fixed by software. I hope this is not the same problem.

I had the same problem with the Nexus S

A similar thing happens to a Motorola Xoom that own. Turning the screen on and off always fixes the issue. It's usually rare, once every month or so, so it's not that annoying, but when it happens, I end up swiping multiple times and then realize, oh I've run into that problem. I can live with it, but would be nice if it was fixed.

I have no problem what so ever with my Nexus. It's the most freaking awesome tablet I have had.

Author and other people might have got a bad piece.

There's yet another similar issue with the Galaxy Nexus: http://code.google.com/p/android/issues/detail?can=2&sta... (video included!)

So... I just let mine bake for 30 minutes on the same game he did. The device was hot by the end of it. No problems with touch/multitouch recognition afterwards.

Had the same sort of problem with my Nexus One. Different OEM too. Pretty strange. Maybe the open source version of Android doesn't have anyway to keep the touch sensor calibrated beyond turning off and back on. I can recall ThinkPad's where the trackpoint would get stuck going a certain direction after some use as well. So continuously calibrating input devices is certainly something that can be required.

If it's an overflow/waterfall type issue(for lack of a better descriptive term) then turning the screen off and on will fix it temporarily until the trigger (such as heat) sets it off again.

It's less likely a calibration issue because the rest of the screen is performing accurately, and the behaviour is inconsistent.

I just tried his same test on my 7 with the same game and the same multitouch tester app and I was unable to replicate the bug. Must be a manufacturing issue that affects only some devices.

Hmm. No such issue with mine. Might be a literal manufacturing defect.

I wonder if OP bothered to try and reproduce this with other models, or if they decided to make a mountain out of a molehill.

Have you used Tegra-specific software?

I grabbed SHADOWGUN THD (a Tegra-optimized game) this morning and experienced exactly what the submitter experience: The game is very quickly completely unplayable because you have to control movement on both sides of the screen, which you quickly discover isn't possible.

I am a huge Android booster, but I have to confess to being quite disappointed at this point. The screen has the left side defect where it raises from the frame. There are bizarre lag glitches and pauses (start a Netflix movie and for the first minute or so of play there are multi-second long video pauses).

I have a feeling this is going to turn into a giant PR disaster for Google/Asus.

Yes.

I wasn't willing to pay for Shadowgun just to test this, but I played Fruit Ninja THD for a good 10 minutes with no problems. I also watched a full Netflix show with no appreciable lag.

I've noted very bad screen flicker on the Nexus 7 when browsing the web (large white background). This is a Google IO unit.

I've been skeptical of Google-branded hardware ever since my Nexus One stopped working due to a non-functioning power button. I find myself associating "beta" with their stuff - that's fine for software but not so much for physical goods.

I think this is most likely a software problem. There were similar problems with multitouch in the bottom right corner of the Galaxy Nexus which was fixed in the first OTA update, just a few weeks later.

If it's heat from the hardware, that can still be fixed with either more optimised drivers, or outright capping performance.

Any ideas if this is a software or hardware defect? If the former then a patch may be on the way.

In most cases software or firmware patching is sufficient, even in the case of some heat issues. If it can not be fixed via patch or only affecting a small batch of users, then returns are a better approach to distributing firmware.

If it's widespread then google didn't do enough intervention testing(QA processes). Which wouldn't be unusual since even mattel are guilty of letting intervention testing slide. (It slows production and pushes up costs.)

FWIW, I just played Shadowgun for 20 minutes on my n7 and I'm not having any touchscreen issues.

i experienced this when I was registering, keyboard input would not register until after 5-10 presses, then buffer filled up and spit it all out with multiple key presses in the input box.

Reply on news.ycombinator.com