Keyboard woes

[sorcerer]

I hate asking for help before doing any research myself first but I've never seen this issue before and, to me, it seems so bizzare that I can't even formulate a suitable search string for Google so I hope you can forgive me this time.

Customer brought in an HP 17-x047na laptop saying that some keys on the keyboard aren't working, so I think ok, new keyboard, but then he goes on to say that the on-screen keyboard is also showing the same symptoms but not necessarily with the same keys - and so is an external USB keyboard when he connects that!

It happens in all aspects, ie, when using Gmail, MS Word, Notepad or any other program that requires text typing.

I've uninstalled/reinstalled keyboard drivers.

I've tested the hard drive and it's all good.

I did a couple of virus and malware scans that found maybe half a dozen things, cleaned it up, and had some improvement in that an external USB keyboard now seems to work alright on all keys, although the laptop's own mechanical and onscreen keyboards are still playing up.

It came in with Windows 10 v1903 so I've upgraded it to v2004 in the hopes that that would solve the problem but no joy.

I opened a blank Word document and typed out some text using each keyboard in turn, videoing it in an attempt to illustrate the problems. You'll notice that as I get to the onscreen keyboard I manage to type out the alphabet in lower case but when I click the Enter key to move down a line it highlights rows of keys in blue and I've no idea what that's all about. I'll upload it here if I can but I'm unsure of file size limits so if it all goes a bit Pete Tong then you can view it from my Google drive here

Any ideas folks?

 

[britechguy]

Check to see if Fast Startup is enabled (and it is by default) and if it is, disable it. [Control Panel, Power Options, What Power button does, options not available link].

I have seen some of the most bizarre issues in my life that come from the corruption of the contents of the hiberfile.sys that gets used as part of the form of hibernation used for Fast Startup.

At least it's a starting point.

 

[sorcerer]

Check to see if Fast Startup is enabled (and it is by default) and if it is, disable it. [Control Panel, Power Options, What Power button does, options not available link].

I have seen some of the most bizarre issues in my life that come from the corruption of the contents of the hiberfile.sys that gets used as part of the form of hibernation used for Fast Startup.

At least it's a starting point.

Thanks Brian, will do.

 

[JoelM]

Did you try using a spare hard drive you have with a fresh install of Windows 10? This is the only true way of eliminating a software issue as a possibility.

 

[Computer Bloke]

1. Try it with Linux. If the problem's still there it's almost certainly hardware.
2. Disconnect the internal keyboard and test with an external keyboard. If the problem's fixed then you have either a faulty keyboard or a stuck modifier key.
3. If you're still seeing the problem then the next things to look at are the keyboard language/layout and accessibility options.
4. If none of that helps it's time to sacrifice a virgin, if you can find one. I've been to Preston...

(Edited to highlight "a stuck modifier key" because that's what my money would be on, based on the fact that the OSK is affected too.)

 

[sorcerer]

Sorry about the delayed reply, just had a hospital appointment this morning.

Thanks for the suggestions, I'll give them a try and let you know how I get on.

Cheers.
PS Computer Bloke - you do indeed know Preston well

 

[sorcerer]

I can now say that it's definitely hardware as the problem was still evident whilst running from a Linux live CD.

I'm leaning towards faulty RAM because after booting from UBCD it crashes out every time I try to run Memtest86+

Thanks again guys.

 

[sorcerer]

Sorry again for the delay, this lockdown has got me all strung out.

Of course, as Computer Bloke suggested, the problem turned out to be the physical keyboard. Although there was no outward indication, no stuck keys and all keys physically felt the same in operation, something has obviously gone wrong at the membrane level.

I ordered a new keyboard for it, which arrived today, and I stripped it down because it's one of these silly designs where everything has to be removed to get at the keyboard. Even sillier, it seems to be mounted behind a plate that is fixed down with loads of little plastic pop rivets rather than screws!

The options seem to be a) drill out the pop rivets and then replace the plate with some glue of some sort, or b) buy a whole new top cover with keyboard which would 1) make it more expensive for the customer and 2) make the one that I've already bought, surplus to requirements.

Or is there a third option that I don't know about? What do you do when confronted with this?

 

[Porthos]

I ordered a new keyboard for it, which arrived today, and I stripped it down because it's one of these silly designs where everything has to be removed to get at the keyboard. Even sillier, it seems to be mounted behind a plate that is fixed down with loads of little plastic pop rivets rather than screws!

No help because I refuse to do those without a full replacement but welcome to the new normal when it comes to keyboards.

 

[britechguy]

And I don't think those are rivets. I have faced this situation more than once. What appears to be "the fastening method" is dropping the keyboard on to strategically placed plastic posts that fit through the holes in certain spots then "melt welding" the tops of the posts such that they form something a bit like a screw or nail head holding the keyboard firmly in place.

When I've had to deal with this, I have used a rotary tool with a cut disc, very carefully cutting those "melt heads" off. The keyboard can then be lifted out. For reaffixing, I have used epoxy glue to do exactly what the plastic had been doing. A large drop (but not too large) atop each of the posts (which should still be very visible and, in fact, still go through the keyboard at least close to level with it) does pretty much the same thing as the original. Haven't had one fail yet. [And I, like others have already said, try to avoid having to do this, period.]

 

[sorcerer]

Thanks Brian, good idea.