Atom tapes

[Fraser]

Some files have some peculiar bits such as a missing wave or an extra bit of a wave. Does anyone know how the Atom produces the waves and if there is anything that might upset the smooth flow of waves?

Dave will you be getting round to hosting the Atom tapes that have been preserved? Is an Atom emulator in development?

[Arcadian]

Dave will you be getting round to hosting the Atom tapes that have been preserved?

Yes! Thanks to Charlie R ("sirmorris") my Atom now has a lovely MMC card bolted on the back, and this has given me the impetus to get all the Atom tapes transferred and archived online. Will be in touch soon if I have any questions ...

[CMcDougall]

my Atom now has a lovely MMC card bolted on the back

, so does that run tapes or disc images??
how the hell did DFS0.9+ work with that? is there hope for
a MMC version to work on an Elk or elk with Plus 1 or 3

[regregex]

is there hope for
a MMC version to work on an Elk or elk with Plus 1 or 3

There's a version of GoMMC for the Electron, which can plug in the expansion port or a ROM socket with enough space around it.

--Greg

[CMcDougall]

^^ yeah, i know that

but Charlie R ("sirmorris") must have changed the MM67 MMC board to work on an Atom, so wondered if he could do a elk version, as i hate GoMMC

[sirmorris]

My ears were burning

The AtoMMC board has been developed from scratch - it's nothing like MMBeeb, the hardware and software are brand new.

My Elk went missing a long time ago now

Maybe it's time to look around for one...

[CMcDougall]

^^ thats well wicked then!

canny wait if u get an elk version going

[sirmorris]

Hmm. It looks like there would have to be a plus one in there somewhere without this becoming a 'real' electronics job. I'm not a real elecronicician, I only pretend to be

[Fraser]

Is there any document about the BBCs highest cassette baud rate which is about 11000? I can't remember what it is exactly.

[regregex]

BBC Advanced User Guide, p.388:

The cassette system selects between 300 and 1200 baud using
this [ACIA] division ratio. The serial ULA is always set to 300 baud for
cassette, so division by 64 actually generates 300 baud. Division
by 16 makes it 4 times faster so 1200 baud is generated. Division
by 1 would make it a further 16 times faster, ie 19200 baud but
the cassette will not operate at this speed.

--Greg

[Fraser]

Does the long wave frequency become 19200? At 300 baud it is 1200 but theres 4 waves.

[Fraser]

Does anyone know what the Atoms fast tape add-on does? Is it needed to use the 1200 baud rate format?

[sirmorris]

There are utility ROMS which offer the 1200 baud rate. I'd have to hunt around to tell you which, though.

[regregex]

Does the long wave frequency become 19200? At 300 baud it is 1200 but theres 4 waves.

I wanted to check something about even/odd parity and decided to test this at the same time for you. The frequencies stay at 1200 and 2400 Hz (+1.6%) but the ULA switches between them faster. How much faster is debatable but is definitely not 19200 times per second. The data stream is not recoverable. See attached wave file which includes source program. VLSI behaviour is practically identical. For a suitable decompressor see e.g. here.
Regards,
--Greg

[Fraser]

I can't see any difference to a 1200 baud file. Is it recorded straight from a BBC? The waves are very regular and it easily converts to CSW with a baud rate of about 1201.

[Fraser]

Have I misunderstood you? Is the WAV file a program to produce the 19200 baud rate?

[regregex]

The 19200 baud output follows the program, as you can see it's nothing more than a security wave.

I guess that at the start of each wave the ULA latches TxData from the 6850 and emits a whole short wave or a whole long wave then latches TxData again.

I recorded it straight from the BBC using a minidisc player as a pass-through amplifier.

Here is the program listing. It copies itself to the cassette port at 19200 baud:

Code: Select all

    2 REM >CASOUT5
    4 REM Adapted from AUG p.212-3
   10 REM Fudge factor: put 2 dummy bytes in RS423 input buffer
   20 REM to allow control of RTS flag by use of buffer
   30 REM tolerance (OSBYTE &CB, 203)
   40 *FX 138,1,1
   50 *FX 138,1,1
   60 REM Enable receive interrupts to allow control of RTS
   70 *FX 2,2
   80 REM Indicate that RS423 is cassette
   90 *FX 205,64
  100 REM Select baud rates
  102 REM 19200 baud RX (not 300)
  104 REM for Ferranti and VLSI ULAs
  110 *FX 7,8
  115 REM 300 baud TX as set by MOS
  120 *FX 8,3
  130 REM Reset 6850
  140 *FX 156,3,252
  145 REM Set ACIA to divide clock by 1
  150 *FX 156,0,252
  160 REM Turn tone on
  170 *FX 203,9
  180 REM Turn motor on
  190 *MOTOR 1
  200 REM Inform user
  210 PRINT "Press record and return"
  220 DUMMY=GET
  230 REM Select output route
  240 *FX 3,3
  250 REM Send ULA synchronisation
  260 VDU &AA
  270 REM Wait (header tone)
  280 TIME=0
  290 REPEAT UNTIL TIME=500
  300 REM Send data with a '*' tape synchronisation
  310 PRINT "*HELLO THERE"
  312 FOR Z%=PAGE TO TOP-1
  314 VDU ?Z%
  316 NEXT
  320 REM Wait until buffer empty
  330 REPEAT UNTIL ADVAL(-3)>&BE
  340 REM Pause for a short period of tone
  350 TIME=0
  360 REPEAT UNTIL TIME=50
  370 REM Disconnect RS423 output
  380 *FX 3,0
  390 REM Turn off tone
  400 *FX 203,255
  410 REM Wait (for an interblock gap of silence)
  420 TIME=0
  430 REPEAT UNTIL TIME=150
  440 REM Turn motor off
  450 *MOTOR 0
  460 REM Restore RS423
  470 *FX 205,0
  480 REM Tidy up serial input
  490 *FX 2,0
  500 *FX 21,1

Greg