VOGONS


First post, by static-

User metadata
Rank Member
Rank
Member

I recently picked up a Libretto 50CT laptop. A beautiful device!! It works fine, except it has issues with the keyboard.

Most keys work, but some are "finicky".. like pressing ENTER does nothing, but if I press SHIFT+ENTER it works.
RIGHT arrow doesn't work at all

So I took it totally apart and don't see any damage, leaky battery, or bulgy capacitors.. but there is evidence of what looks like a small (drink?) spill at the top-right corner of the keyboard.

Unfortunately the dock I have doesn't come with a PS/2 port (only parallel, serial, vga), so I can't try with an external keyboard.

I cleaned the keyboards much as possible, but I could only get so far as there seems to be no way to disassemble the keyboard further.

My guess is there are some short inside that's making for some odd issues.

Does anyone have suggestions as to how I could clean this further? Is it safe to "pop" the little chicklet keys off, to get in further?

Keyboard looks like this:
https://www.worthpoint.com/worthopedia/ ... 856143492

I'd love to fix it somehow!

Thanks

Reply 1 of 12, by Horun

User metadata
Rank l33t
Rank
l33t

Odd: shift+enter works but just enter does not ??
Sounds more than a keyboard issue as each key should work and if only when multi are pressed works that is not the individual key switch but something that matrixes them.
Have you removed the ribbon cable and cleaned it end and the socket it goes into ? May not help but cannot hurt....

Hate posting a reply and then have to edit it because it made no sense 😁 First computer was an IBM 3270 workstation with CGA monitor.

Reply 2 of 12, by Thermalwrong

User metadata
Rank Oldbie
Rank
Oldbie

If you can solder thin wires inside the dock, to the dock connector pins, you can add in a PS/2 port to the dock. I've done it on mine and it's a big improvement. This site's the best, looks like it just went offline recently 🙁 https://web.archive.org/web/20190715143147/ht … k/ps2guide.html
Scroll down to the guest list for Ben Lee's solution. Each of the clock and data lines should be connected to VCC with a pull-up resistor (apparently 4.7k? works on mine). Beware taking the dock apart too, Toshiba's plastic mix from the libretto era is super fragile these days.

--------------------------

The fault in your 50CT's main keyboard is a damaged trace inside the two membrane layers. The inside of the keyboard is made up of two plastic layers, each with conductive traces. One is the columns, then the rows - if you look at the keyboard connector on the 50CT, it's made up of two separate ribbon cables. The left half should be the columns, the right half should be the rows.

I personally think that given its the right arrow and enter key, the rightmost/last column is the one that's not working or broken part-way. When you're pressing two keys, it completes the circuit that way, allowing it to read the keypresses.

I had the exact same fault with my NEC Versa keyboard a few weeks ago: Re: What retro activity did you get up to today?
It requires disassembling the keyboard, working out which keys link where on the two membrane layers, then inspecting and finding the fault. To open up the keyboard, you need to undo the plastic pegs on the back of the keyboard. The libretto appears to have loads of them though, eek! I used a soldering iron at maybe 220c to gently press on a peg, then used a pry tool to separate the keyboard layers by that peg. Held that carefully in place until the heat from the iron melted the plastic, then the plastic on that tab would lift upwards, allowing that peg to unclip. Then just kept going. Reassembly is pretty much the same too, put it back together, then gently press the iron onto the now raised plastic edges, to flatten it again, holding the layers together.

Honestly, Libretto keyboards are more common, it may be easier to get a replacement instead - Some Librettos that haven't had their RTC battery removed are starting to die from battery corrosion, so there are parts units around. The prices certainly went wild in the last couple of years though.

Reply 3 of 12, by Thermalwrong

User metadata
Rank Oldbie
Rank
Oldbie

An additional thing too, if you're working on the libretto, read the maintenance manual before doing anything - don't make the mistakes I did and pull it apart blindly. The fragile plastic parts / clips hold up a lot better if you're following the correct procedure for taking it apart. If you want to get to the motherboard, press the bits by the battery clip inwards to unclip the front, then lift up the screen / keyboard palmrest part, hinging on the rear of the laptop, where the fragile clips are.

Sadly the little 50CT doesn't support cardbus, so you can't use a cardbus USB card to get a working keyboard either.

Also, I may be wrong about the membrane? Check that the keyboard connector is clean using IPA + cotton bud to clean it. But I do suspect it's a membrane thing because it's so close to what I had with my NEC, especially since you can see where something got spilled on it.
Here's another example of a keyboard repair with a bit more detail, similar fault too: Re: What retro activity did you get up to today?

Reply 4 of 12, by static-

User metadata
Rank Member
Rank
Member

WOW... this is incredibly informative. Thank you SO much! I think I'll try what you suggest; see what I can do to take apart the keyboard and fix it. If I can't fix it, try to keep an eye out for parts model. If that fails, try this dock mod or find one with PS/2 support.

My bad keys seem to be: 8, 9, -, u, i, \, k, enter, insert(?), right

On the keyboard ribbon cable, I see two groups of pins (8pin, and 11pin)

So you were able to melt the same plastic peg back over the metal sheet?

I was able to boot the Libretto without the keyboard installed, so I assume if I did have external PS/2 that I could use that 100%.

Also do you know if the individual keys can be popped out?

Thanks again...

Reply 5 of 12, by static-

User metadata
Rank Member
Rank
Member

I followed the ifixit guide and was able to take it all apart.. you're right about the brittle plastic though, I had a small tab break off, but it all closed nicely and didn't seem to actually hook to anything.

I didn't notice any leakage, cmos battery looked fine (and actually holds charge). Caps looked fine too.

I tried cleaning the keyboard ribbon cable, black came off (kind of like a nes cart) but it made no difference... so probably the membrane as you said.

Reply 6 of 12, by static-

User metadata
Rank Member
Rank
Member

Okay so I used a soldering iron to melt the plastic bits and separated the keyboard. There was a bunch of gunk on the right side of the plastic sheets so I cleaned that all up.... but I'm having trouble putting it back together to see if it works.

I tried my best to keep the pegs intact, but isn't enough plastic to be able to bind it to the backing... some pegs kind of snap in a bit... and some I need to resize/squeeze to fit. Is there another way I could secure the keyboard to the metal backing?

It's super tricky to hold together with all the individual domes.. they dislodge really easily haha

Attachments

  • keeb.png
    Filename
    keeb.png
    File size
    721.49 KiB
    Views
    432 views
    File comment
    keeb
    File license
    Public domain

Reply 7 of 12, by Thermalwrong

User metadata
Rank Oldbie
Rank
Oldbie

Wow, well done getting that far! IT's quite unnevering to do, but it should all be fixable. Don't reassemble it just yet - best to test it first.

For testing it - with the plastic keyboard part removed, you can now reconnect the keyboard and test out which keys are working. I expect that just cleaning it won't be sufficient and you'll need to reconnect some traces using conductive paint. I watched a couple of youtube videos for membrane keyboard repair. This one explains the process really well: https://www.youtube.com/watch?v=REz928pgCII
This is also a good video: https://www.youtube.com/watch?v=7i5D8xPsGCk

To repair the traces, you just need something that's conductive and adhesive. Three methods:
1. Conductive paint applied with a syringe or needle tip

2. Using Kaptop tape, stick a single strand of copper (remove from a cut mains cable or something) over the broken point, following the trace. I just spotted this in this video

3. IF you have some copper slug tape (which you can buy in lots of stores), cut off a piece of that and shape it like the trace but a bit too long. Adhere the middle part over the broken trace part, then fold over the ends so that the copper touches at each end, then stick down with kapton tape (if you can't tell, I like kapton/polyamide tape - it's more permanent than sellotape, is quite thin and a good insulator).
There's an even better way - instead of using slug tape, use actual EMI tape, which is adhesive and conductive on both the copper and adhesive sides, then you just need to cut something in the shape of the trace and stick it down.

It took me a while to find the fault on my Toshiba's keyboard, but on both my bad keyboards, it was near the edge since that's where water ingress can most easily occur. For the Toshiba keyboard, I took a picture of both sides, mapped out which keys were bad and used that to draw lines and deduce where the fault was.
But mostly, just inspect carefully under a bright light! In both cases, the faulty part was a slightly discoloured / darkened part of the trace.

When it does come to reassembling, use some non-serrated needle pliers, or some tweezers, to carefully squeeze the plastic posts to match up with the flat parts of the holes in the aluminium. That will usually allow the pegs to go back in and increase the height / plastic amount enough that heat staking them back down should work. IF you have a larger tip for the soldering iron, I found that the larger area is easier to squash down the plastic, like the last picture of my 730CDT keyboard repair.

Also I gave you the wrong temperature for peg removal, sorry - I used 200C for my better repair, that will melt slowly allowing the plastic to deform in a more orderly fashion than it will with the higher temp. 200 or 210C should work fine for reassembly.

Reply 8 of 12, by static-

User metadata
Rank Member
Rank
Member

Of course why didn't I think to test the keyboard as is. Silly me, I'll do that first and see if the keys are still bad. Appreciate all your repair tips! I'm still worried I won't have enough plastic to secure the board. Some of it came off with the iron. If that's case, is there something else I could do? I suppose I might have to resort to glue? Or somehow add plastic? 😂

Reply 9 of 12, by Thermalwrong

User metadata
Rank Oldbie
Rank
Oldbie

You can use the soldering iron like a paintbrush to reattach the plastic while it's all clamped together. Since it's not something that will be visible when the laptop's back together. I've been using some scrap PETG 3d printed parts as 'paint' for a couple of repairs recently.

Reply 11 of 12, by static-

User metadata
Rank Member
Rank
Member

Ah-Ha... so what I've discovered is that long traces themselves seem fine, but there's those little "resistors" (thicker black lines) at each key.. right by the key's circular pad. All the bad keys seem to have much higher values.

For working keys, I'm seeing 13-20k ohm
For broken keys I'm seeing 60k-300k ohm

eg.
Broken right arrow: 672k
Broken enter: 97k
broken other: 160k, 60k, 80k, 60k, 153k,

If I rub with an 99% alcohol qtip, I see the qtip becoming back (dirt? spill? Graphite?) ... and the resistance goes down (I got right-arrow from 672k down to 300k), but I feel like I'm rubbing off the black resistor/graphite itself, I'm concerned I might wipe it off totally.

I assume I can't just use conductive paint here, as I'd guess the keyboard expects some 13-20k resistance.

Should I keep rubbing with the qtip?

You can see in the attached pic, it's starting to rub out. That measured 600k and now down to 300k

Attachments

Reply 12 of 12, by Thermalwrong

User metadata
Rank Oldbie
Rank
Oldbie

I think just patch it so it's 0 ohms to test whether that helps - I don't really understand why the Toshiba keyboards have those weird painted on resistors. My Tecra 730CDT had 10k resistors on most of the traces which should be visible in the pictures I posted.

Don't rub off all of the stuff that's on there. Cleaning it off more will just mean less trace to conduct. You did say there was a spill around that area, perhaps it dissolved the conductive paint over time for those traces.
If there's still a bit of the paint on there perhaps you could use a leaded pencil to draw on a resistor. It won't be very permanent but maybe you can get it close and put some tape over it.

I just got the ESD tape in - it works. I'll have to see if I can use it to entirely replace a trace. The last keyboard I had to fix had an entirely different fault - the angle the ribbon cable connected to the motherboard caused it to break at the connector end.