[Coco] Q: Burke & Burke hard disk controller

[Steven Hirsch]

On Tue, 30 Sep 2008, Gene Heskett wrote:

> On Tuesday 30 September 2008, Steven Hirsch wrote:
>> Dumb question (haven't started trying to set things up yet, so forgive any
>> naivety):
>>
>> If the expansion box has a selector to make exactly one slot active, and
>> the hard disk interface is plugged into one of them, how exactly does one
>> gain access to the floppy disk?  Are the disk interfaces always active
>> regardless of the slot selector?
>>
>> Steve
>
> No.  The drivers for the hardware are responsible for handling that switching,
> and restoring it when finished, usually to the default slot 3 where the
> floppy controller normally lives.  Most however, read it, save it, and put it
> back to where they found it.
>
> Note that due to a brain fart (they were possibly thinking like gamers only),
> they also switch the IRQ paths, meaning a device not currently selected can
> exert an IRQ in vain, causing loss of incoming data in a comm proggy for
> instance.
>
> The std fix for that that I have been promoting for many years is to jumper
> all 4 slots together at their pin 8, and remove 3 of the 4 pullup resistors
> located along the front edge of the MPI's pcb, otherwise the device trying to
> exert the IRQ pulldown may have the abilities of its output device taxed
> excessively.
>
> Once this jumpering is in place, the IRQ will get through, and the os9 IRQ
> scan will find it, call the driver to service it, which will switch it, do
> its thing, and switch it back when the IRQ has been serviced.  End of lost
> data, even at 9600 baud.

Ok, so the CoCo uses level-triggered interrupts?  Sounds easy enough to 
do.  Thanks VERY much for the informative response!

Steve


--