Partial VICTORY on the Cifer! After swapping dead chips, probing with a meter, probing with a logic probe, and probing with an oscilloscope, I'd almost reached the end with the graphics card.
But the tech manual pointed me in the direction of a variable capacitor on the board. After 30 seconds of tweaking I got the garbage on screen to be stable and then hit the reset button.
And it works. YES! A properly operating, Tektronix 4010 emulating, graphics board!
Think I might have to give up on the Cifer. It works as a single floppy CP/M machine but the hard disk controller, Tektronix 4010 emulation card and 68000 CPU cards just won't work..
I've checked as many of the chips as I can, replaced a load of them and spent far more money and time than any sensible person should.
I did manage to get it to boot and fixed the keyboard, so that's two small victories. I might come back to it at some point but I need a break from it...
Finally, some small progress with the Cifer. Managed to get the 68k board partially working by removing a 74LS166 which allows the 68k CPU to start. This activates the IEEE488 bus. But if the SASI controller is connected this stops the bus working. So there's a fault on the SASI board somewhere. But at least the system is now looking for a hard disk even if there's no card installed... #retrocomputing #cifer #abouttime
I am so mad at myself. Stupid. Stupid. Turned on the Cifer after chip swapping/testing and found the keyboard didn't activate the usual disk message with ctrl+disc keys. And then the smell hit me... A burning smokey dying electronics smell.
Didn't take long to find it. Chip ML93 which is a 74LS126, was the temperature of the sun. I have destroyed that chip by making the dumbass mistake of putting it in the wrong way around. FFS.
Chip swapping. Still need a 7425 and a couple of 74LS01. The original batch of 74LS01s I bought from eBay were all faulty. Fortunately, the seller refunded me straight away citing a potential bad batch. So I'm now waiting for new ones (and the 7425) and we can get testing.
Bizarrely, a second check of the dram from the Cifer 68k board now has at least four chips that seemed faulty yesterday to be working today. To be fair, it may be the socket as I don't have a zif socket to use.
I managed to find the moment in the test where one chip failed. Two pics, the first shows my kit oscilloscope reading back zeros. Second pic should be the same but the two spikes are the chip spitting out '1's when they should be '0's.
I'm in pieces, bits and pieces...
Cifer power supply has an issue. Between 5v and ground there is a resistance of 12.5ohms which my (cheapie) meter beeps as a short. I'm not a power supply expert. Wonder if a Meanwell would fit in the gap? 🤔
The power distribution board is literally just a bare board with connectors so no easily replaced tantalum caps there. Boo!
A bit of an update on the Cifer 1887.
https://crashedfiesta.blogspot.com/2023/03/the-wheels-on-bus-go.html
Hmmm. The Cifer is still working on its floppy but I cannot get the hard disk to do anything. My issue is, there is the chance the drives are dead, but I can't be sure as I think the SASI/IEEE488 controller (mounted on top of a Xebec 1410 Winchester controller) has an issue.
I wonder if there's a DIP switch that needs to be swapped somewhere... 🤔
I also re-made three of the cables as they were stupidly short and made it impossible to get back together easily.
Cifer 1887 board photos. Posted here, primarily so I don't forget which is which... The IEEE bus is still not working. Humph.
Stripping down the old Winchester disk. Circuit board off. Let's get some pics. Start at that end, great. Now the main circuit board. On to the other en... F***!!
First sign of life on the hard disk front... Sort of anyway, at least it's checking the addresses but it only did it once. I repaired the broken leg on the Z80 on the disk controller board but, do you know what? I can't remember if I reconnected all the cables... Doh. #cifer #retrocomputing
Dangnabit. Unexpected socket replacement in the workshop area.
The Cifer tells me the IEEE bus is not active so thought I'd try and clean up that board. A pin broke off a Z80 while I was reseating the chips.
YES! Cifer 1887 keyboard now working. A faulty CD4011 at position 5 on the PCB.
Managed to get a CP/M floppy to boot. Computing like it's 1983! Ah-haaaaa!
So, I'm assuming that because the Cifer keyboard springs into life (if spouting garbage) when I touch the pins of IC6 or IC7 I'm gambling one of these has a problem... There are new chips on the way.
Also, there is a weird pattern on screen when it's cold. After it's warmed up, those lines disappear and the 'garbage' text from my poking becomes proper characters.. Hmmm.
Sitting, trying to get my head around how the Cifer keyboard works. I get nothing when keys are pressed. If I pick up the board and my fingers touch around the legs of IC7 or IC6 on the underside of the board then the serial output goes nuts and sends garbage to the screen. Hmmm. I'm coupling something together... 🤔
Looked at the Cifer keyboard briefly. While I was connecting up the sillyscope I got sparks from the ground clip. Turns out, to be compatible with the older keyboards this one is driven by 0v to -12v, with -12v connected to 'ground' and 0v at the 12v point. So I suppose that means connecting the ground clip to 'ground' essentially put a dead short across the -12v rail. That would explain the slightly odd noise that came from the power supply when I connected it up initially... FFS. 🙄
For anyone who missed my Cifer escapades yesterday, it starts up but had garbage on the screen. Tracked down a broken 4164 RAM chip on the VDU board and that fixed it. Yay!
Next issue is that none of the keys seem to work on the keyboard, so it's either locked up (but still had a blinking cursor), the keyboard just isn't working or there is some other problem hidden on one of the many other boards.
One thing at a time. 😁
Another Cider update. The garbage on screen at startup stays there even when the 68k board, graphics board, processor board and floppy/hard disk boards are removed. So it's something on the VDU board.
I then swapped:
Z80 CPU
Z80 PIO
Z80 CTC
No difference.
I tried swapping the static RAM chips around (there's five at 2kb each) but no difference.
There is 64kb of dynamic RAM made up of 8 x 4164s so my money is on the fault being there...😬🤔
