home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!cs.utexas.edu!zaphod.mps.ohio-state.edu!uwm.edu!spool.mu.edu!sgiblab!bridge2!gilligan!camerons
- From: camerons@NAD.3Com.COM (Cameron Spitzer 764-6339)
- Newsgroups: comp.os.linux
- Subject: sharing IRQs, was Re: BUG in 0.99[p1] kernel config script a
- Summary: don't short ISA IRQ's!
- Message-ID: <1437@bridge2.NSD.3Com.COM>
- Date: 31 Dec 92 22:37:53 GMT
- References: <1heirvINN3gr@gap.caltech.edu>
- Sender: news@bridge2.NSD.3Com.COM
- Reply-To: camerons@NAD.3Com.COM
- Lines: 22
- Nntp-Posting-Host: gilligan.nad.3com.com
-
-
- In article 1heirvINN3gr@gap.caltech.edu, nyet@cco.caltech.edu (n liu) writes:
- >
- >Cards CAN be connected to the same IRQ.. its just that your system won't
- >know which one called the interrupt and won't service the right card all
- >of the time...
-
- This advice is dangerously wrong. ISA boards drive their interrupt
- request lines high and low. If two boards disagree about which way to drive
- the IRQ wire they share, too much current will flow from the high one to the
- low one. If the boards are not short-circuit protected, the weaker of the
- two can be permanently damaged. Even if nothing burns out, the resulting
- signal may not be a valid logic level, and may therefore be vulnerable
- to noise, creating spurious interrupts. Some boards may seem to work,
- for a while, but they will sustain unknown damage.
- Please see my earlier post about how to share ISA interrupt request lines.
-
- EISA solved this problem. EISA boards drive IRQs low only, and a resistor on
- the motherboard drives high. The resistor is too weak (8.2K) to hurt anything.
-
- Cameron
-
-