Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 23 Jul 1996 18:06:04 +1000
From:      Bruce Evans <bde@zeta.org.au>
To:        schofiel@xs4all.nl, tweten@frihet.com
Cc:        hardware@freebsd.com
Subject:   Re: The multiple COM ports discussion
Message-ID:  <199607230806.SAA24356@godzilla.zeta.org.au>

next in thread | raw e-mail | index | archive | help
>schofiel@xs4all.nl said:
>>4) In the PC/ISA scheme, interrupts are POSITIVE-GOING, EDGE triggered.
>>   In the EISA scheme, interrupts are by the default compatible to this,
>>   but can be configured to be ACTIVE LOW, LEVEL triggered.
>>5) TRistate interrupt line drivers are not neccessary in this scheme. 

>What do you mean by "this scheme?"  If you mean that "ACTIVE LOW, LEVEL 
>triggered" is "this scheme," fine.  All that is needed then is OPEN and 
>ACTIVE LOW -- two states in ths strictest sense (though people often use 
>"tristate" when all they really mean is that one state is OPEN).  If "this 
>scheme" is "POSITIVE-GOING, EDGE triggered," see my hardware concern below.

I think many of the points only apply to EISA.

>Both ISA and EISA wire all interrupt "pins" on their motherboard connectors 
>together.  The only ISA standard I have any familiarity with is the 8 MHz 
>IBM PC/AT, and the IBM-supplied boards that went with it.  Those boards 
>drove their interrupt lines with two states, ACTIVE HIGH and ACTIVE LOW.  
>There was no OPEN state.  There was no output resistor to mediate between 
>"dualing" bus driver circuits.  Therefore, there was no possibility of 
>getting predictable results from putting two boards on the same interrupt.  
>For each of the four combinations of two boards asserting or not asserting 
>an interrupt you could get board-one dominance, board-two dominance, or you 
>could get the privilege of replacing fried parts.  It all depended upon the 
>variable details of the driver circuits on both boards.  Most of the time 
>the hardware survived this kind of abuse because the board designers 
>accounted for the possibility of people accidently installing two boards at 
>the same interrupt.

I think most ISA boards do have tri-state drivers, but there is no
standard for this.  I'm only very familiar with (have a manual for :-()
pre-AT designs.  These devices in the IBM PC Technical Reference Manual
gate the IRQ through a 74LS125:

    IBM Monochrome Display And Parallel Printer Adaptor (Logic 11 of 12)
    Parallel Printer Adaptor
    Asynchronous Communications Adaptor

These devices are connected more or less directly:

    5 1/4" Diskette Drive Adaptor (Logic 4 of 6) ('765 through MC3487 to bus)

Now all these devices except the Monochrome display are usually in one
chip, but the interface hasn't changed (:-().  There would be no point
in leaving out the tri state enables which probably cost 0.1% as much
as the jumper block for manual enables.

Bruce



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199607230806.SAA24356>