Spectrum +3 Keyboard issues

[Lardo Boffin]

Hi all!

I’m having troubles with a +3 reading the keyboard.

When switched on it has (once or twice) worked for a while before certain keys stopped working.

I have tried the keyboard in a +2A and it works fine. I have also tried the +2A keyboard in the +3 and that fails in the same way.

This one was a bit battered in its journey to me so it could be a physical issue maybe? E.g. a connection knocked lose.

The following keys either don’t work or stop working shortly after switch on:

3
8
E
I
D
K
X
M

I did try to look for a keyboard layout / matrix to see if this pointed to a particular connection but have not found one yet.

My next plan is to reflow all of the joints of the keyboard connectors but I did do a continuity test on each of them and they were all ok.

It does work sometimes!

[1024MAK]

There’s a fault with one of the columns. Have a look at the original rubber key layout. Each physical row is made up of a left hand set of five keys and a right hand set of five keys. Each set of horizontal “half” rows has one of the eight row lines.
And each column line does one key for each “half” row, hence there are five columns on a rubber key.

This is the same on all ZX Spectrum keyboards including the +3. The differences are in how the extra keys are handled.

Sound like you either have a dry solder joint, a cracked PCB track or some other poor connection...

There is a layout of a +2A/+2B/+3 keyboard here
[Rather confusingly, the columns (Y) are laid out horizontally].

The PCB schematics are here

Mark

[Lardo Boffin]

Thanks Mark. Column 3 by the looks of it then!

I guess I need to try and work out which physical connection on the PCB this relates to.

[Lardo Boffin]

Resoldering the socket had no effect. Worked for 30 (ish) seconds and then stopped again.

Time to investigate further...

[Lardo Boffin]

Looking at the keyboard matrix, column 3 has the issue:

Looking at the layout I assume that the top connector receives the signal from column 3?

I presume that only the top 5 pins are used? Looking at the underside of the PCB it looks like the rest just go via resistors to ground.

So my faulty one is pin 3 of the top 5 which I guess is the third from the top and goes via the joystick port?

[McKlaud]

Have a look around 40077 chip (ULA). I would not be surprised if some of soldering joints are cracked. It might be a good idea to re-flow the chip.

[PokeMon]

Maybe it is just a bad keyboard membrane.
By the way, will sell the 48k-KDLX keyboards from next week again on sellmyretro.
If you want a tactile safe clicker instead of non-responsive membranes.

By the way - fits perfectly under the upper plate/rubber mat.
https://www.sellmyretro.com/offer/details/3977

[Lardo Boffin]

Good to see your shop getting back up to speed Pokemon.

I’m pretty sure it is ULA soldering as after prodding it a bit it is currently working.

Not sure I’m brave enough to try and reflow a surface mount chip though.