[Coco] DLOAD (was: Re: Stupid Cassette Port Tricks, (was: Re: Coco and .WAV files))

Gene Heskett gene.heskett at verizon.net
Sat Feb 2 13:29:21 EST 2008


On Saturday 02 February 2008, jdaggett at gate.net wrote:
>On 2 Feb 2008 at 5:37, wdg3rd at comcast.net wrote:
>> Aside from other trivial factors, the cassette port was two and a half
>> times faster than you could reliably shove through the bit-banger.
>> Sad but true, the bit-banger in those days (improvements have been
>> made, thanks Mark and Boise) was a slow and unreliable interface.
>> Couldn't be trusted above 600 bps.
>
>I never used the internal serial port on the Coco1 or the Coco3. To much
>overhead having a software UART. I used the MPI and the serial cartridge.
>Had good reliability up to 9600 baud. Just wished the serial printers of
> that time supported faster speeds.

I have, on a coco2 running OS9L1, VERY reliably at 4800 baud, using the 
aciapak driver through /t1.  That is how my imitation EDisk talked to that 
GVG video mixer for nearly 15 years.

In that time, if the data got scrambled, it was always in a custom made fifo 
chip in the mixer that was 4 bits wide and 16 nibbles deep, and which were 
used in the GVG for an 8 bit wide bus transfer by using two in parallel with  
common clocking on each side.  The major failure was that the two chips would 
get out of step with each other. GVG wanted $178 each back in 1988, but I 
found a place that had an AMD made jedec numbered version for about $3 & had 
the foresight to buy two 24 count sticks of them after the second failure in 
2 weeks.  I had 3 chips left when it was finally pulled out of service in 
2002.  Interestingly, I never had to replace one of those AMD chips again.  
IIRC there were something like 48 of them in that mixer, so in that mixers 20 
years of life (we bought two of them, a corroded by NYC's air piece of junk 
from the J.C.Penny's production house that I couldn't stay ahead of, and 
KTLA's pullout about 4 years later, which was in much better condition), the 
failure rate was about 95%.  <1% would have been the rate for every other 
digital chip in it.

The analog stuff in it was another horror story and toward the end I was 
replacing their $1800 video speed op-amps built on ceramic postage stamps 
with one of National Semi's $1.37 op-amps, but they were _much_ faster and it 
gradually got out of color phase, about 8 degrees for every one I had to 
replace.  That's what you get for replacing a 5mhz part with a 300mhz part 
that used 5% of the power their hybrid used.  Technology marches on at a very 
rapid pace in that genre of chips.  I should have replaced them all, then no 
phasing problems but cleaning up that 4 layer pcb to get all the through 
holes cleaned out was a PITA.  I would have had to do that 24 times to get 
all of them.

-- 
Cheers, Gene
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Often statistics are used as a drunken man uses lampposts -- for support
rather than illumination.



More information about the Coco mailing list