+- Discussion Forum for all things Microbee (https://microbeetechnology.com.au/forum)
+-- Forum: Microbee Forum (https://microbeetechnology.com.au/forum/forum-1.html)
+--- Forum: Microbee Hardware (https://microbeetechnology.com.au/forum/forum-6.html)
+--- Thread: New Keyboard replacement kits (/thread-211.html)
RE: New Keyboard replacement kits - g67wjjk - 07-01-2021
Managed to fix a few keys now.
It seems that a nicely working key switch has about 200Ω resistance when pressed, open circuit when not working at all, and a higher resistance (600-1800Ω?) when you really need to mash it down to register, with the resistance decreasing the harder you press it down.
RE: New Keyboard replacement kits - someone - 08-01-2021
(07-01-2021, 01:11 PM)g67wjjk Wrote: I have tried to open one up, not having much luck.Your photos show mid generation key switches. They become lousy after repeated use because their elasticity is lost and the carbon pill doesn't make a good contact - particularly on the shift and enter keys because they're pressed hard often.
Okay, just tried again - trick seems to be to gently lever the middle part of the base plate. Did one side, then did the other and ... the spring is now somewhere in my garage! But I can now open them up.
The later generation with the vertically mounted elastic pills have the opposite effect and become hard, don't flex and eventually form stress cracks thus their carbon pills don't make any contact. The carbon compound on these pills lose their conductivity over time irrespective if they've ever been used or not.
The first generation keyswitches used metal contacts but they're a rarity.
The surface of the carbon pills is rather slick and most things slide or break off. Someone has been experimenting with cheap adhesive aluminium tape, nickel paint, carbon paint and others. The aluminium tape seems to work well. The nickel paint seems to crack off and the carbon paint's resistance seems too high. Silver would be an option but that would tarnish over time. Even with the successful repairs, the feel of the keyswitches is horrible.
Replacing RN1? - g67wjjk - 18-01-2021
I was reading in the HWNB and found a section about improving key switch reliability - one thing it mentions is increasing RN1 to 33kΩ ... Section A, Chapter 5.
In my 32K it is 4.7kΩ ... worth replacing? There's still one or two I'm not happy about, and I have to type quite slowly in general - just assumed it was because it was slow (2MHz and BASIC 5.10) ... ?
To improve the reliability of the keyswitches on the microbee, a resistor network, RN1, has to be changed.
This resistor network is in front of IC4 under the keyboard frame. It is used as a pull-up resistor on the input lines of IC4 and does not appear on the old circuit diagram. It is on the 1248-4, the latest circuit diagram.
The resistance value of this network is 3k3 or 4k7, which is too low. The resistance has to be increased to 33k (or near to this value); a 33k network or eight 33k resistors will do.
To check the value before removing, measure the resistance with a multimeter. From the bottom of the base board, measure between pins 1 and 10 at each end of the network.
<snip about replacement procedure>
If you have a machine with faulty keyswitches, it is advisable to check to see if the correct value is fitted for RN1.
RE: New Keyboard replacement kits - someone - 19-01-2021
Don't believe that article!! It just suggests bodges to a fundamental issue with the key switches.
The resistor value used depends upon two main factors:
- The keyboard logic family used for the keyboard scanning circuit
- The resistance of the keyswitches
The different logic families have different logic threshold voltages thus the combination of the keyboard switch resistance and and pull up resistor resistance needs to be compatible. There have been 3 families used 74LS, 74HC and 74HCT and each of them have their own nuances.
Irrespective of this, the most critical factor is the key switch resistance.
RE: New Keyboard replacement kits - g67wjjk - 22-01-2021
Ahhh, okay, thanks. It did look like a real PITA to replace, having to remove two key switches ... although I did consider just cutting the legs and then soldering the new one on the underside.
I've certainly found checking the resistance is a good way to confirm whether a switch is working or not before soldering a prospective spare in and then finding it's the same - the Series 2 PC I'm currently selling had a failed left SHIFT key - I've got a bunch of spares from the Educator I gutted so was able to check the resistance on a replacement BEFORE soldering it in and testing.