[Coco] Deal to first 3 responders

[Gene Heskett]

On Tuesday 25 August 2009, Christian Lesage wrote:
>Gene Heskett wrote:
>> That was the name of an article in Rainbow, on how to install a normal
>> acia (sy6551) serial comm chip, in a std coco2, and do it in such a way
>> that under os9, it shared the regular serial ports connections, but with
>> I believe, better flow controls.
>
>Nice mod! The only drawback I can see is that it still uses only 4 wires
>to communicate with the outside world, so true hardware flow control is
>out of question.

If you read carefully, I believe the ready line was treated as a wired or, so 
either side could pull it down and stop the transfer at the end of the byte.

How well that worked I have no idea cuz I never ran it faster than 2400 baud, 
with $88 byte packets, the contents of one buses control register, so there 
was a dead clock pulse or 3 in between the GVG handing it data, or vice 
versa, as some registers in the GVG were cascaded as much as 5 packets deep.

The only dropped data problems I ever had with it in that application were 
the result of a badly designed and very high failure rate 4 bit wide , 16 
nibbles  deep fifo that GVG used a pair of as a rubber buffer between parts 
of the circuit, clocking pulse failures internally.  I finally found an AMD 
chip that was almost a std jedec number, and about 10,000 times more 
dependable.  I bought 2 sticks for $1.84 ea, compared to the $30+ Grass 
wanted for their old stock, and I believe there was two OEM chips in it yet 
when we pulled the GVG in favor of about 80 grand worth of a new switcher 
from echolab that ran on 5% of the Grasses power, and had 10x the bells and 
whistles the Grass had.

>I've been looking for doing something similar, but I
>was thinking about putting the 6551 inside the Multi-Pak interface
>instead. Yes, it would imply doing address decoding and RS-232 level
>translation, while this mod merely uses one additional chip. And yes, I
>know Roger sells RS-232 paks, but my goal is to free an additional slot
>on the Multi-Pak interface.

That we could use.  But far more utility would be made by simply doing full 
decoding of the i/o ports used .  Wasting a $20 wide chip enable on a 4 
address wide i/o port, and doing it 4 times if you count the disk controller 
is inexcusable, and IMO purposely hamstrung the machine.  Whoever did that 
ought to have been water boarded, repeatedly, until he got religion.  Yes, it 
could be fixed, but it would be an individual fix to every coco that wanted 
its Achilles tendons whole.  On the mobo, at the design stage, cost might 
have been a quarter dollar!  I boil over every time I think about it.

-- 
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)
The NRA is offering FREE Associate memberships to anyone who wants them.
<https://www.nrahq.org/nrabonus/accept-membership.asp>

<Knghtbrd> hardcopy is for wussies
<Topher> computer program listings....next, on HardCopy