04-03-2017, 05:36 PM
Every so often I have a moment of madness.
I decide that a particular day is a play day and get a couple of computers out of storage along with a CRT and have an afternoon wallowing in nostalgia whilst trying to educate my kids as to the joys of 8 bit computing.
The kids wryly look at dad as an old fart, to be tolerated.
This particular afternoon in question was one of those play days that mutated into an afternoon of diagnostics. The computer in question was the Spectrum + model that I had obtained a few years previously. It worked well with tape based programs but I could never get a microdrive to operate correctly with it. I had chatted a long while back with other users about the issue and a chap who was more up to speed on the dark arts of CPU timing suggested that the CPU may be at fault, in this instance the M1 control line.
So after a few rounds of Jet Set Willy it was off to the workbench with the computer, back off and in with something I didn't have the last time I looked into the issue, a DSO. The Z80 M1 (machine cycle 1) control signal is required along with IORQ to correctly control any device plugged onto the edge strip. It should be a clean square wave switching signal, so I was hoping to capture an alternating square wave 0/+5v evenly squared image. What I did get however was an image that looked like the electronic equivalent of wire wool, and after further examination of the CPU it looked like the brains of the unit deffo had an internal fault.
Getting the old CPU out is semi easy. Cut the legs and remove the main body, then a good hour or so with a solder sucker to remove the remaining legs and clean the holes out. As there was plenty of space inside the case I elected to fit a 40 pin DIL turned pin socket, then plug the replacement CPU in. Fingers crossed, power up and all worked with the Sinclair K at the bottom of the screen.
A quick scope around showed a nice clean signal at pin 27 (M1), turn off, fit interface 1 and a microdrive, power on and I could access and load off the MD cartridge. Case back on and another unit that works as the designers intended.
I decide that a particular day is a play day and get a couple of computers out of storage along with a CRT and have an afternoon wallowing in nostalgia whilst trying to educate my kids as to the joys of 8 bit computing.
The kids wryly look at dad as an old fart, to be tolerated.
This particular afternoon in question was one of those play days that mutated into an afternoon of diagnostics. The computer in question was the Spectrum + model that I had obtained a few years previously. It worked well with tape based programs but I could never get a microdrive to operate correctly with it. I had chatted a long while back with other users about the issue and a chap who was more up to speed on the dark arts of CPU timing suggested that the CPU may be at fault, in this instance the M1 control line.
So after a few rounds of Jet Set Willy it was off to the workbench with the computer, back off and in with something I didn't have the last time I looked into the issue, a DSO. The Z80 M1 (machine cycle 1) control signal is required along with IORQ to correctly control any device plugged onto the edge strip. It should be a clean square wave switching signal, so I was hoping to capture an alternating square wave 0/+5v evenly squared image. What I did get however was an image that looked like the electronic equivalent of wire wool, and after further examination of the CPU it looked like the brains of the unit deffo had an internal fault.
Getting the old CPU out is semi easy. Cut the legs and remove the main body, then a good hour or so with a solder sucker to remove the remaining legs and clean the holes out. As there was plenty of space inside the case I elected to fit a 40 pin DIL turned pin socket, then plug the replacement CPU in. Fingers crossed, power up and all worked with the Sinclair K at the bottom of the screen.
A quick scope around showed a nice clean signal at pin 27 (M1), turn off, fit interface 1 and a microdrive, power on and I could access and load off the MD cartridge. Case back on and another unit that works as the designers intended.
