Hack to get 5V for keyboard

[mrtinb]

I have a custom keyboard for my ZX81 with a 14 pin ribbon cable. 13 pins are the keyboard lines (8+5) which are connected to the keyboard matrix.

I have would like to add some 4066 ICs to have some of the extra keys on the keyboard to do Shift together with a key. However that will require 5V and GND. But I only have 1 wire left in the cable.

So I started to think a way to hack this. I thought that the wire could be used for GND. And since at least one of the address line has HIGH 5V at any given time, I could put them all into an OR-gate, and the output would be constant 5V. Then the 5V for the OR-gate would be fed from itself. Then maybe add a capacitor, to ensure a constant voltage.

Electronics is just a hobby to me, so forgive me if this is nonsense. Is this possible at all or is it just in my dreams?

(Edit: The keyboard is not connected directly to the motherboard but to a external keyboard interface designed by Wilf Rigter, similar to a Memotech keyboard.)

[Paul]

You could have a closer look at the 48p-kdlx from Pokémon. He enables extra keys without extra lines (just 8+5) and makes energy harvesting from the keyboard lines. He uses low power ICs for this. There is a Schematic in the tlienhard forum.
Kind regards Paul
Edit: name correction and specify the proper forum.

[1024MAK]

What you want, from your description, is a ‘phantom’ power feed.

Use silicon Schottky diodes, e.g. BAT41, BAT42, BAT43, BAT46, or similar.

Connect the anode of each diode to each of the five column lines. Then all the cathodes connect together, and go to the +5V / VCC / VDD pin of the 4066 ICs. You also need two capacitors connected in parallel between the +5V / VCC / VDD pin and the 0V / GND pin. One should be a ceramic 100nF / 0.1uF, the second should be a 4.7uF 10V (or higher voltage) electrolytic.

No need for an OR logic chip. The diodes ‘OR’ the power from whichever line(s) are high. But keep in mind that the amount of current via the 5.1KΩ resistors (if you have used the same value as shown in Wilf’s schematic, which I presume is this one http://www.user.dccnet.com/wrigter/inde ... XKBDv3.htm ) is rather limited. So it will only power CMOS chips.

Mark

[mrtinb]

Thank you very much for your help Paul and Mark.

Yes. Pokemon's schematic and your description Mark are quite similar.

He uses BAT54 and only one 10uF capacitor.

[XavSnap]

Hi,
I would like to know if in this case, the power of the assembly could be externalized with a common mass?
There are many cheap food that can be recovered for this assembly.
Are there risks of overloads or problems of separation of electrical voltages on TTL?
… just to avoid the main 7405 overload (overheat) with ram… expansion cards… or others.

[1024MAK]

An external 5V power supply can be used. The requirements are that the 0V / GND connection of the external power supply is connected to the computers 0V / GND connection. The external power supply +5V MUST NOT be connected to the computers +5V connection. The external power supply should be a regulated type.

Although if the only thing that is being powered is some low power CMOS chips, the 7805 voltage regulator in the Zeddy will not even notice the very small extra current. So will not make any practical difference to the heat from the heatsink.

Mark

[mrtinb]

Connect the anode of each diode to each of the five column lines. Then all the cathodes connect together, and go to the +5V / VCC / VDD pin of the 4066 ICs. You also need two capacitors connected in parallel between the +5V / VCC / VDD pin and the 0V / GND pin. One should be a ceramic 100nF / 0.1uF, the second should be a 4.7uF 10V (or higher voltage) electrolytic.

Thanks for your reply, Mark.

I've added the capacitors to my schematic. I don't know much about capacitors. Are one of them polarized?

phantom.png (8.75 KiB) Viewed 2966 times

[1024MAK]

Connect the anode of each diode to each of the five column lines. Then all the cathodes connect together, and go to the +5V / VCC / VDD pin of the 4066 ICs. You also need two capacitors connected in parallel between the +5V / VCC / VDD pin and the 0V / GND pin. One should be a ceramic 100nF / 0.1uF, the second should be a 4.7uF 10V (or higher voltage) electrolytic.

Thanks for your reply, Mark.

I've added the capacitors to my schematic. I don't know much about capacitors. Are one of them polarized?

Yes, the 4.7uF electrolytic type is polarised. Wikipedia shows the symbols here. I prefer the European symbol which uses two rectangular blocks, one filled in (negative) and one not filled in (positive). The left hand one in this image:

On the actual component, with the more common radial type (two leads at one end), one side will have a stripe with minus symbols, the lead/leg on this side is the negative lead.
Wikipedia has this section on it
This page is a bit easier to read and includes a photo showing a real capacitor including polarity marking.

Ceramic 100nF / 0.1uF capacitors are not polarised, and can be connected either way round.

Mark

[mrtinb]

As SMD I didn’t find electrolyte capacitors, but they had tantalum capacitors. I have never heard of tantalum capacitors before, but it looks like it could replace the electrolyte capacitors.

[Moggy]

As SMD I didn’t find electrolyte capacitors, but they had tantalum capacitors. I have never heard of tantalum capacitors before, but it looks like it could replace the electrolyte capacitors.

They are polarised like electrolytic caps but are a bit more prone to damage when a reverse polarity is applied and they are usually more accurate in their specification I have found as well as being a bit more expensive.