Dell E6400 Login Blocked So Swing the Lid

I probably banged my laptop too hard, since I broke my hard plastic water bottle at the same time. The laptop was working fine earlier in the day on battery power and now was asleep. But the next time I wanted to use it, my password wouldn’t wake it up. It’s a Dell Latitude E4300 (much of this also applies to the E6400).

I no longer have either of these laptops, so I can’t go back and check anything on them.

My USB keyboard lets me log in, but I don’t want to carry the bulky USB keyboard around all day, especially with the relatively delicate key caps unprotected, so I wanted to fix the internal keyboard or whatever was broken.

(Disassembly instructions by the manufacturer are available online, so I won’t repeat them here. Look for the service manual for either of these laptop models.)

Testing revealed that, when the problem is present, some key caps were not getting responses any more, and generally the same key caps (but not the same across the two machines). Most key caps remained okay. No entire row has the problem, although the problem may be concentrated in one row or certain rows.

The temperature of the keyboard probably does not matter. The length of time since a power-up or a warm reboot does not correlate with functionality or failure.

An OS update on the E4300, which runs Ubuntu 20.04.2 LTS Linux, seems to have solved the problem with all keys. The E6400, which runs Fedora 34 Linux, which is kept evergreen, has not experienced that benefit so far. On the E4300, switching between English (US) and Italian keyboard layouts seemed to make no difference about which key caps failed or succeeded even though resulting characters differed as specified by the layouts. (Keyboard layouts for languages using the same Roman alphabet often vary in the placement of nonalphanumeric keys.) I wanted to do further testing, but the update seemed to end the failures.

Reseating the keyboard did not solve the failures when they were occurring, and I’m beginning to suspect that reseating it too many times, which may be a small number, can harm the keyboard connection and may contribute to internal keyboard performance degradation for the affected key caps (although not other key caps, so maybe that’s irrelevant).

I found out why reseating failed even when done only once. I took the keyboard out again and this time I looked more closely at the electrical connector on the edge near the user. For the E4300, there are about 37 conductors on it (on the E6400, I think 37 or 39, but I forgot which). It looked like the distal tips on a whole bunch of them that were adjacent to each other were bent slightly upward, enough to show up under a flashlight and a magnifying lens. (The distal end of something is the end away from a center of a larger object. The distal end of your arm is the end away from your shoulder and away from your chest. It’s a medical term.) Maybe reseating risks bending the connector tips.

We should be careful in fixing that. They likely can break. Uninsulated wire conductors on an integrated chip (granted this isn’t about an IC) might be made to withstand only three bendings, since a factory process should not need more than three to install it once. If one of those breaks, repairing it may require hyperlocal soldering, because regular soldering introduces more heat and that might destroy something near the tiny conductor, and hyperlocal soldering may be costly and hard, so you may have to replace the unit, which can be pricey. I don’t know if that’s true for this keyboard connector, but I don’t intend to test to destruction.

So, basically, I wiped the conductors with a fingertip, with a light-to-medium touch, toward the distal end (away from the key caps), with the fewest strokes needed to straighten them. A small flashlight and a magnifier helped me check.

On the E6400, the distal tips looked fine and I didn’t even try adjusting most of them. However, I wiped that connector crosswise a few times with a paper towel soaked with 50% isopropyl alcohol (rubbing alcohol), which you likely can buy cheap in dollar stores and drug stores (it’s for cleaning and medical use, not for drinking or smelling).

A problem I did not solve is that the connector, basically a plug, goes into a receptacle which I can hardly see into and can’t do much about. (I should have a dental mirror, which is handheld and can see behind teeth.) The receptacle is small and hard to reach into. It’s possible there’s dirt inside the receptacle. It’s possible that conductors in the receptacle are damaged and are damaged in a way that leads to intermittent failures. I could only hope for the best.

I slid the keyboard back in, without too much pressure, and, to make sure the keyboard was lying flat in position, used my open hands to press across many key caps at once.

It’s not necessary to bolt the keyboard in or to snap the LED cover strip back in yet, provided you’ll be careful not to touch anything that could be electrically conductive yet, including any metal that could carry leaked ground current. I plugged in the AC power and tested all the key caps, using a few apps since not every app need use every key cap, especially all the function (Fn) keys.

A couple of keys still didn’t work on the E4300. Advice somewhere on the Web said it could be due to dust or crumbs, and maybe I should have taken the keyboard out again. But I had already held the keyboard up in the air and pressed every key, tapped edges, and flexed it carefully. I didn’t want to unseat and reseat the keyboard this time, in case reinsertion could damage the keyboard’s connector. So I hoped for the best, replaced the three bolts and the LED cover strip, reconnected power, rebooted, and tested it.

After one repair attempt, the machine still needed improvement but it was much better than it was before that. Every key cap, at least those that I knew worked before the new problem, began working again. A few intermittently hesitated to be responsive and a few intermittently made double characters with a single strike even when accessibility repeatability was off. I’d have to proofread a lot more, but at least the computer was good again for use during travel. I may have to get another computer but this may be good enough for a while.

The rubbing alcohol didn’t work for the E6400 for the first few keystrokes that had recently failed, but then all was fine. That’s odd; but I hoped it stays that way. A previous attempt with rubbing alcohol had not made a lasting cure. But after sleep overnight, success was briefly inconsistent until failure was consistent. Powering off for a couple of hours helped, but a while later the problem resumed. Maybe heat in the machine contributes to the problem.