This is it:



So I decided (with some help from the community ;) ) to stick with the 48K/128K/+2(grey) models.
If someone finds a solution for the +2A/B/3, I'm happy to adapt it to my schematic, but most likely you'll have to adapt your +2A/B/3 internally.

I think I will have some ready in one and a half week.

I replaced the 74LS541 that drives the datalines with a CMOS type 74HCT541 on the prototype.
Unfortunately the interface does not function at all with this type of driver, so I still have no solution for the +2A/B/3 ZX Spectrums...

But.. Help me to choose:

1. I will produce the interface for the 48/128K/+2(grey) ZX Spectrum, and in the mean time try to solve the problem for the +2A/B/3 models; there may be no solution for those models

2. I will continue to try to fix the interface for the +2A/B/3 and wait with the production till that's done, so you can be certain your purchase will work on all Spectrums - but again, maybe the problem can't be solved

3. I will not focus on the +2A/B/3 anymore, because they already have a descent keyboard, and only start working on the production now; 'com'on Ben, we're waiting!'

I know there might be people like me who also want to keep the possibility open to use a +2A/B/3 with this interface, but I'm only having so little precious time... So I hope 'you people' can help me out; my wish is option 3.



p.s. the cause of the problem may be fixed by placing resistors between the ULA and the Z80 of the +2A/B/3 models, but that would mean you have to solder the parts in yourself.
That way you can use the interface I have ready at the moment.
If anyone is interested, I can test this and create a document as manual to do it yourself!

I pulled out my wreck-of-a-ZX-Spectrum-+2B out of a box with spare parts.
Fortunately this one was working, although I had used it before to experiment with (case was in 3 pieces, taperecorder disconnected, and more).

So I inserted the PC keyboard interface, and yes!!
...

No... not all keys worked.. pity.......

It seemed dataline D0 did not function correctly.
But I also noticed the internal keyboard did work simultaniously, although that should not be possible with my current design of the interface.

I contacted the very happy Miguel from Spain again (you may know 'they' won the European soccer-competition yesterday!) to confirm our thoughts that the +2A/B/3 ULA has a much stronger coupling with the Z80.

This is caused by a lack of resistors on D0..D7 so that external interfaces almost can't compete with the data supplied by the ULA.
The only attempt of solving this problem without modifiing your +2A/B/3 is to use a databus driver on the interface that has a very low impedance, or shortcircuits the datalines to 0V when needed.

But.. this is not without the risk the +2A/B/3 ULA is damaged.

Our first attempt to solve this problem with cause is to replace the 74LS541 by a 74HCT541 which has the same function, but uses MOSFET transistors with very low impedance.
This might work well!

I ordered the pieces an hour ago, but it may take up to two weeks to arrive :( SORRY!!!

This will be the PCB design for production:



As you can see, I rotated the PS/2 connector upwards, so the keyboard connector is inserted from top instead of from the right.

This way there is no hassle with the power connector from the Spectrum.


+2A/B/3 testing will be next! (this evening maybe)
After that the production will start!

I must have a complete partlist with prices somewhere...
Let's .. check... ahh yes, here it is:

If I build 25 pieces, the price would be only 25 euro .
I'm not sure if that's with silk screen finishing on the PCB or not..

(Check check... )

Ok, so that INCLUDES a silk screen, which is not on the prototype (the pictures in previous messages show the bare PCB without silk screen and coating).

If you are already interested, mail me and I will put you on my list.