48K's and bad DivIDE compatibility
I have had some 48K boards and complete machines in the past which didn't work fully with a DivIDE interface.
The reason for that always was a Z80 processor having problems with the signals.
I replaced the Z80 with spare ones, and the problems were always solved.
But now I have about 5 different 48K boards with sort of the same situation, but after replacing the Z80 I still have problems.
I tried several Z80 models, some of which I'm sure they work with DivIDE in other machines.
I also tried other ULA's, but the problem is still there.
When I power the 48K the DivIDE boots, detects the card and pressing a key boots the Spectrum rom.
But the NMI button does not fire up the NMI menu; instead it does strange things, like letting the 48K beep or reset.
So I'm looking for what parts could be failing.
I have these categories:
1. the NMI circuitry is failing
2. There is some other component failing, a chip or a discrete part
Maybe there are more causes to think of.
I added the pin 38 -> pin 20 wiring on the CF card adapter for 5V on the CF card.
Any ideas where to look?
The reason for that always was a Z80 processor having problems with the signals.
I replaced the Z80 with spare ones, and the problems were always solved.
But now I have about 5 different 48K boards with sort of the same situation, but after replacing the Z80 I still have problems.
I tried several Z80 models, some of which I'm sure they work with DivIDE in other machines.
I also tried other ULA's, but the problem is still there.
When I power the 48K the DivIDE boots, detects the card and pressing a key boots the Spectrum rom.
But the NMI button does not fire up the NMI menu; instead it does strange things, like letting the 48K beep or reset.
So I'm looking for what parts could be failing.
I have these categories:
1. the NMI circuitry is failing
2. There is some other component failing, a chip or a discrete part
Maybe there are more causes to think of.
I added the pin 38 -> pin 20 wiring on the CF card adapter for 5V on the CF card.
Any ideas where to look?
Post edited by bverstee on
Comments
I changed the Z80 and the problem was not solved.
Maybe a bad RAM chip?
I have a +3e that has this problem I am reviewing the requirements to repair it / and the sound and the audio input / output socket LOL!
gotta love eBay!
I replaced the remaning 74LS157's.
I replaced the 1uF capacitor for the reset.
I checked all resistors for NMI, RESET and other signals.
I replaced the 7805.
Still the problem remains.
The active parts remaining:
- 4116 ram ic's
- all transistors
- clock circuitry
Any ideas?
If not, I'm going to have a real lot of spare parts very soon...
I'm starting to dislike the 48K boards, so many components fail nowadays.
Is it aging or are the parts bad from the beginning?
I'd say its a little of both, if you are sure the RESET line is not at fault then I am a little stumped - I am sure you have checked all the Z80 Expansion socket pins for continuities and shorts...
All you have left is RAM's and Transistors.... I suppose atleast you would get to the bottom of it - once everything is replaced... the benift there is to know what problem / faulty part was that could be affecting the others.
update
I would also suggest trying external power for the CF adapter if you haven't already.
Crappy design :)
Plenty failed when they were new. Sinclair famously put price before quality :p
If you are sure you have good CPU, then issue may be with distorted signals (weak, delayed) due to poor PCB contacts. Proper timings are critical for divIDE to work. Maybe ROM checking if it disconnects properly.
Ingo.