My Klyball 600D build

[bxdanny]

Congratulations!

Remember, you have to press <space> to take the computer out of LOAD mode. It won't respond to any other keys until you do. If that's what you thought was the machine "hanging", it wasn't.

[ronin47]

Congratulations!

Remember, you have to press <space> to take the computer out of LOAD mode. It won't respond to any other keys until you do. If that's what you thought was the machine "hanging", it wasn't.

I'd forgotten about pressing space, I'll check that tomorrow, thanks

[bxdanny]

Hey, I just realized something. Your screenshot shows

?SN ERROR

without the N being replaced by a graphic symbol, as it typically is. Could there be a problem with bit 7 of the video RAM? Or with the character generator? Or does your copy of the BASIC1 ROM have the error messages fixed? What is the value at $A167, is it $4E (N) or $CE?

[ronin47]

First off, yes you were right, I'd forgotten to press space to stop load - D'oh!

Second, If I've read it right, the value of $A167 is $CE. I reflashed ROMs 1&3 with files that state 'fixed' after them, however with the fixed ROM 1 the machine resets itself almost as soon as you arrive at the BASIC prompt and keeps doing so.

However I've left it with just the fixed ROM 3 and I now see this instead:

IMG20240611103208.jpg (5.76 MiB) Viewed 6525 times

Which I am assuming is correct based on what you've said.

I did actually manage to save and almost load a basic program this morning (One line was corrupted) using a cassette deck, so now I know it's nearly there. I suspect saving is fine and now I just need to tweak R57 until loading works as expected.

It's good progress and that 150Hz routine certainly made troubleshooting easier, I appreciate all your help @bxdanny.

[ronin47]

A quick glimpse of the cassette test program:

IMG20240611111200.jpg (5.46 MiB) Viewed 6523 times

No, it's not working properly, but it is an improvement on not working at all.

I'm going to start off by setting the pulse width at U69 pin 5 to 500uS and will slowly increment to see if things improve.

[bxdanny]

That is the way a syntax error is usually displayed by OSI BASIC, yes. Although I'm not sure why you changed it from the way it was, I prefer seeing the letters that Microsoft actually intended. For the record, to "fix" it again, the odd-numbered locations in the ROM from $A!65 through $A185 should have bit 7 cleared: $Cx changed to $4x and $Dx changed to $5x.

As for tape errors, I know Sideburn found that switching to better-quality tapes solved the problem for him, so you might want to try that. Perhaps also keep the tape recorder away from the switching power supply (if that's what you're using), to avoid magnetic interference?

How are you feeding that "cassette test" into the machine? From an actual tape, or from a sound output of the PC? If the latter, adjust the volume for best results. Presumably, the input signal at J2 pin 10 should have about the same amplitude as it did in the test yesterday. And you can try adjusting R57 while the tape (or audio file) is playing, as Thomas suggested, and see if that helps. But I would have thought that getting that square wave reproduced meant that things were adjusted right.

You could also try generating your own copy of that test tape with this program:

10 SAVE
20 NM$=CHR$(15)+CHR$(64)+CHR$(15)
30 FOR N=65 TO 96
40 PRINT TAB(8);"! ";
50 PRINT CHR$(N);" !";NM$
60 NEXT N
70 GOTO 30

[Note: The resulting tape will differ slightly from Thomas's in the number and placement of nulls, but should look and act the same.]

Record the output from that for as long as you like (it will loop until you stop it), then try playing it back.

[ronin47]

That is the way a syntax error is usually displayed by OSI BASIC, yes. Although I'm not sure why you changed it from the way it was, I prefer seeing the letters that Microsoft actually intended. For the record, to "fix" it again, the odd-numbered locations in the ROM from $A!65 through $A185 should have bit 7 cleared: $Cx changed to $4x and $Dx changed to $5x.

I assumed you were suggesting I should change it, I just used the ROM files that I had. I had no idea it was wrong until you mentioned it and I burned a new ROM. IIRC I just used the original non-fixed ROMS in the hope I could just make it work.

As for tape errors, I know Sideburn found that switching to better-quality tapes solved the problem for him, so you might want to try that. Perhaps also keep the tape recorder away from the switching power supply (if that's what you're using), to avoid magnetic interference?

The tape deck I'm using, and the tape could easily be an issue, however I don't have any better tapes at the moment, so I switched to the PC, however I do have a second tape deck that I could try and I could try to see if I have any better tapes hanging around.

How are you feeding that "cassette test" into the machine? From an actual tape, or from a sound output of the PC? If the latter, adjust the volume for best results. Presumably, the input signal at J2 pin 10 should have about the same amplitude as it did in the test yesterday. And you can try adjusting R57 while the tape (or audio file) is playing, as Thomas suggested, and see if that helps. But I would have thought that getting that square wave reproduced meant that things were adjusted right.

I am currently trying to load from PC and adjusting the volume, the amplitudes all look fine and the same as yesterday. The main problem with the PC is that it doesn't like mono cables and the 600D doesn't like stereo cables, so I need to figure out something better, and that was partly why I was using the tape decks as both ends are then mono. I have set the pulse width to the minimum and am going to try adjusting whilst I try and load, I just need to confirm which direction I need to turn whilst doing that.

You could also try generating your own copy of that test tape with this program:

10 SAVE
20 NM$=CHR$(15)+CHR$(64)+CHR$(15)
30 FOR N=65 TO 96
40 PRINT TAB(8);"! ";
50 PRINT CHR$(N);" !";NM$
60 NEXT N
70 GOTO 30

[Note: The resulting tape will differ slightly from Thomas's in the number and placement of nulls, but should look and act the same.]

I'll try it as soon as I can, thanks.

Record the output from that for as long as you like (it will loop until you stop it), then try playing it back.

Understood, thanks.

[ronin47]

OK, so I hope people are sitting down.

I have managed to find a 'better' tape in a pile of old Atari tapes, its a TDK high resolution chrome oxide tape. It's the best one I have in the house...

I also moved the tape deck away from some other things to limit interference.

I wrote a tiny 2 line basic program and saved it. I've now been able to load it 3 or 4 times consistently.

Next I will try your tape test program @bxdanny.

Will report back on the results.

[ronin47]

So, I typed in the following:

10 SAVE
20 NM$=CHR$(15)+CHR$(64)+CHR$(15)
30 FOR N=65 TO 96
40 PRINT TAB(8);"! ";
50 PRINT CHR$(N);" !";NM$
60 NEXT N
70 GOTO 30

And ran it, it seemed to be all good, but first time I didn't record it. I then did a list to save it to tape and it worked just fine, as after a reset I was able to load it back from tape

Then I ran it and let the tape record. This is the result when playing back

IMG20240611153126.jpg (6.36 MiB) Viewed 6507 times

I call that a result. It isn't 100% perfect as occasionally it skips a line, but the fact I can save and load programs with reasonable reliability.

It's interesting the difference a better tape made too.

[ronin47]

Just a quick addition, I just also managed to load a basic program I saved earlier onto the digital voice recorder, so that is potentially helpful given that finding decent quality tapes is getting harder.