Beiträge von PhilA

    A little experiment shows that generally indicates a track is unusable. Wonderful.


    I need to find some working discs, punch the timing holes and try write them.


    I am concerned the computer will corrupt the only other working TEXT disc I have.

    FLUXCOPY says it made a good copy; I have SUCCESS! after writing the image. I used the top drive from the computer, so the alignment is good.


    However, attempting to read the disk back gives INDEX: FEHLER so it's not writing properly.


    C:\Users\phil\Desktop\FLUXCOPY>"FLUXCOPY 090 - Test.exe" /c='10' /d='1' /i='48'/f='00' /t='39' /p='S' /s='S /M='1,SHUGART,5.25,DD,39,1,1000,25,50,50,75' /M='2,SHUGART,5.25,DD,39,1,1000,25,50,50,75'


    That's what I'm invoking FLUXCOPY with - I think that's correct parameters?

    I was making myself a reference sheet last night, so I can more easily remember the commands.


    I tried saving a test file and it corrupted the disc. Great.


    I've just re-written the image with my FLUXTEEN, shall see if it worked in a minute...

    PhilA , that's great, glad you made it.

    In the meantime I found a manual in english language. If you want I can scan it sometime. It is however so that the control commands (letters for function selection) in the English variant is different to the German. Since you have the German operating system, this probably won't help you at first. How does it behave with the keyboard layout ? Some keys might be assigned incorrectly due to the different keyboard layouts.

    Hopefully you will get the US version of the operating system someday.

    I will take any data relating to the machine if it is available. I have the German version of the manual but my German is quite poor (aka I can't speak German at all...) so the rudiments of operation are quite difficult to understand. However, the way it works, even if the FUNC+key assignment is different, will be useful to understand.

    The key assignment is as you'd expect, DIN. I'm not used to the QWERTZ layout and a few of the keys are accent modifiers as you expect to find.

    I keep looking for people who possibly may have a copy- now that Fluxcopy is a thing, recreation of the discs is possible. I have hope that I'll find some.

    Until then it is nice to be able to type at all, though the most common message the machine gives me is FEHLER. Go figure.


    Disconnected keys until the keyboard started to work.

    Space, return and backspace disconnected and it allows me to type.

    That's narrowed down to 3 keys, common is the lower pin which makes the keyboard upset; too late today to carry on but that's progress.


    Phil

    I pulled one of the keyboard chips out and bypassed it, driving the "data ready" pin direct from the "key pressed" output of the keyboard decoder.


    The computer registered a key press on all but the number 7, return, backspace and space. All it did was produce the digit 7 on the screen unless SHIFT was held and it produced the digit / and if CTRL was held it produced ÷

    I don't have the datasheet for the keyboard decoders, and the keyboard is particularly strange in terms of layout. I think that is so it can detect dual key presses.


    I had a thought tonight, I think I'll rig up a counter and push 00 to FF down the keyboard bus and see what the board makes of it.


    Phil

    I removed the UART because it was holding some of the data lines low- its only task is to serialize the parallel data from the keyboard, which in this machine is not used.

    That led to the keyboard responding a little bit- it will repeat the last key pressed indefinitely, and some keys are not working. Looks like the keyboard decoder has 2 bad matrix lines unless I have shorted switches.

    I shall desolder the decoder and test, but before then I need to see if the "a key is pressed" output is being held high, and the data repeated.


    Phil

    Still away from home. A little searching is going on for my Multitech MPF-1 Z80 prototyping board, because I would like to write a little assembler and create a rom or rom emulator that'll at least test if the 8080 is initializing (nop loop, for instance).


    I'm thinking an Arduino might have enough grunt to act as a ROM?

    Electricity service is restored to the house.

    Went yesterday and the roof was continuing to leak so I added more plastic sheeting to the roof.

    At least now the air conditioner is running and acts as a dehumidifier, so that'll help somewhat.


    House isn't good to live in yet, very damp and moldy.


    Phil

    This project is on hold for a while.


    Unfortunately I got pretty hard hit by hurricane Ida at my house; water leaked in where the roof was damaged. My garage collapsed because of a tornado.

    We are still in Texas. I cannot do much to the house until the electricity is put back on.

    That's all the bad chips on the I/O board replaced.

    Result?


    No


    Edit: Slight change noticed: pulling the CPU card and powering up now causes a nice steady beep from the speaker, which is generated on the I/O board. It didn't do that before.


    CPU card next!

    I took a look at connector J5. It is not populated with a pin header, nor does it have the chips required to operate it.



    However, doing a little looking at the circuitry, my best guess is that chip 18 would be a 74LS273N and chip 10 a 74365AN.


    Reasoning for that is the connections to the bus and chip enable lines make sense, and for a line driver to be the final stage also.


    I would need to look at the connections in more detail but I think this is an unused header for LED's- drive enable, operation etc.

    The operation of these lights has been taken directly off the front panel backplane now, so that makes it redundant.


    Phil

    Floppy drive data path (read, write) multiplexors (74157).

    Both have trouble and start to go glitchy after about 300kHz.

    Pulled one off the spares board that came in today. First time it's seen power probably in decades and there it is running faultlessly at 3.5MHz.


    Anything connected to or in the path of the floppy disc drives appears to be bad.


    Phil

    Board with spare COM 2601 chip arrived today.



    Now i need to order some of the other TTL chips that are bad

    The COM 2601 has +5 and -12V supplies, and doesn't do much when tested. Hence purchase of this board. Hopefully this one isn't bad. I'm guessing it doesn't like the +5V rail to be gone for an extended period.

    Still the cheapest COM 2601 on eBay, to buy the whole board, and it does provide some spare chips.

    It's a year newer (1979), and AES have taken to socketing most of the chips. I guess they learned that lesson...


    Phil

    Recieved some 20-pin chip sockets in the mail so continued testing buffers last night. Got 6 done, another 5 on this board to go.


    However right now, I'm having a break to let some acetaminophen kick in because I've got a headache that's not good for concentration. (My advice to avoid this: If you were planning on having children, don't).


    Phil

    I started putting the index signal lines in but that got too complicated too quickly so I started over.


    Drive read data path. Working on the multiplex selector path now.



    Little wonder if wasn't working at all; the 74121 is integral to the bitstream, between it, the NAND gates and the flip-flop it forms a crude shift register to feed the UART serial data in.


    Phil