Re: D64 image transfer

   From: John Iannetta <73510.2152@CompuServe.COM>
   Reply to: [1]John Iannetta
   Date: Sun, 22 Nov 1998 15:36:12 -0500
   Organization: CompuServe, Inc. (1-800-689-0736)
   References:
          [2]<3655854D.1CEC@chemie.uni-konstanz.de>
    Nicolas Welte said,

"Still a little mistake in there: It's pin 17 instead of pin 15,..."

    That was just a typo, but I'm glad that you mentioned it.  I've been making
entirely too many mistakes lately.  So once more (for the X1541 cable) going
from the 1541 to the PC printer port, it's 3 to 1, 4 to 14, 5 to 17, and 6 to
16.

    Nicolas continued,

"This number makes no sense to me, because it would mean that you could
not connect more than three serial CBM devices together, since each of
them draws 5mA. We all know that more devices work properly and I think
even Commodore allowed more than two external devices at once. And I
think they should know the interface specs better than we do.".

    and,

"I also didn't quite understand this noise margin. Does it mean that the
7406 driver can maintain a LOW level voltage of 100mV if it is loaded
with 40mA, but this voltage rises to 400mV if a load of only 16mA is
used? This does not make sense to me, I would have thought that the LOW
level voltage rises if you apply a greater load and not a smaller one.".

    The noise margin is the difference between the ouput voltage level and the
input threshold voltage; the absolute value is used.  Normally, standard TTL
devices assume a 0.4 V noise margin.  A typical totem-pole gate will have a
fan-out of ten.  The maximum low output sinking 16 mA will be 0.4 V, 0.4 V less
than the 0.8 V input low threshold voltage.  And the minimum high voltage
sourcing 400 uA will be 2.4 V, 0.4 V more than the input high threshold voltage
of 2.0 V.  The 7406 spec for sinking 16 mA is 0.4 V.  But it will also sink 40
mA at an ouput voltage of 0.7 V (a 100 mV noise margin).  So if five 1541's are
connected to a C-64, its CLK and DATA 7406 inverters may each be sinking 30 mA.
mA.  That gives a lower noise margin than if only one drive were connected, and
makes a well-shielded cable important.

    Nicolas added,

"...But how many users can determine from which technology their LPT port is
made and if it is safe to use the simplified XE1541 cable?".

    The simplified XE1541 cable (diodes shorted out) is exactly the same as the
X1541 cable as far as any possible damage is concerned.  Perhaps Joe Forster
should think about that, if he is concerned about being blamed for damages.  I
just checked "SC.TXT" (version 0.81) and couldn't find any warning not to use
an X1541 cable if you are not certain of the printer driver circuitry.  The
XE1541 cable seems to be completely safe, but suppose that an X1541 cable were
connected to PC that used a printer driver that had VERY great current sourcing
capability.  Let's assume that an ideal SPDT switch were used.  When bit 3 of
register 2 is clear, the switch connects the -DATA line (SELECT IN, pin 17) to
+5 VDC.  When the bit is set, the line is connected to ground.  When a command
is sent to the drive, the -ATN line (-STROBE, pin 1) is taken low; that makes
the drive turn on its DATA inverter.  So there is a direct connection (through
the cable) from +5 VDC to the collector of a saturated transistor to ground.

--
I pressed RETURN, and the message said, "Missing GOSUB".

                          John

References

   1. mailto:73510.2152@CompuServe.COM
   2. news:3655854D.1CEC@chemie.uni-konstanz.de