[comp.sys.ibm.pc] INTEL PC 386 Inboard

radon@phoenix.Princeton.EDU (Radon Research) (05/10/88)

I recently installed an Intel Inboard PC 386 board in a real XT and
encountered some problems.  I want to know who (if anyone) else has seen
similar things.
First, my 20 meg Seagate drive generates seek errors (it NEVER did this
before)  It is so bad I'm going to replace it as soon as I complete a
"women and children first" backup.  These seek errors persist for
certain files even when the speed is bumped down or extra wait states
added to hard disk requests.  Also, the machine now fails on occasion to
boot up.  The disk whirrs, but no video (not even a cursor) comes on.
This means no memory count, no boot, - just a big desk heater.
  has anyone else got similar negative reports to make, or glowing
positive reports?  Is it going to nuke my new disk as well? (at least
this one will be nuked under warranty)

-- 
Bill Radon *Phoenix Gang*    "Ask anyone - Bill's a Gas"
Disclaimer:  If you don't like what I have to say, I will infiltrate
your home and eat your children's lungs till they puke blood.

goldberg@mit-caf.UUCP (Andrew F. Goldberg) (05/10/88)

In article <2826@phoenix.Princeton.EDU> radon@phoenix.Princeton.EDU (Radon Research) writes:
>I recently installed an Intel Inboard PC 386 board in a real XT and
>encountered some problems.  I want to know who (if anyone) else has seen
>similar things.
>First, my 20 meg Seagate drive generates seek errors (it NEVER did this
>before)  It is so bad I'm going to replace it as soon as I complete a
>"women and children first" backup.  These seek errors persist for
>certain files even when the speed is bumped down or extra wait states
>added to hard disk requests.  Also, the machine now fails on occasion to
>boot up.  The disk whirrs, but no video (not even a cursor) comes on.
>This means no memory count, no boot, - just a big desk heater.
>  has anyone else got similar negative reports to make, or glowing
>positive reports?  Is it going to nuke my new disk as well? (at least
>this one will be nuked under warranty)

I installed Intel inboards in 2 genuine IBM XT's and have only
glowing praise.  Norton index of 16.7.
At the same time I replaced the original 10MB hard
disk with a 40MB Seagate drive (from PC Connection - came with OnTrack
Disk Manager) and have had no problems, EXCEPT....

Sometimes executable programs would not run off the hard disk unless recopied
from the distribution floppy. CHKDSK was happy, and so was norton's disktest
utility in "check data" mode, but not "check file" mode, for which it reported
errors on the files I was having trouble with.

I believe the problem was with the disk cache software (ICACHE.COM)
that came with the
Inboard.  In the manual it mentions that the default mode is to read not
just the sector you need, but the entire track in which the requested sector
resides. The manual mentions that if there is a bad sector in the track the
cache software may fail. This apparently includes even known bad sectors that
DOS would otherwise tolerate.  So unless you have a disk with absolutely NO
bad sectors (I don't) this default mode will, if I am correct,
bring you guaranteed trouble.
Fortunately there is a way to turn off this "feature" (/T-) and read only
the sector that is requested.  Using the /T- switch has made this repeatable
problem go away from both XT's.

Andy Goldberg
MIT Laboratory for Electromagnetic and Electronic Systems
Cambridge, Mass.

internet: goldberg@lees.mit.edu

wnp@dcs.UUCP (Wolf N. Paul) (05/12/88)

In article <2826@phoenix.Princeton.EDU> radon@phoenix.UUCP writes:
 >I recently installed an Intel Inboard PC 386 board in a real XT and
 >encountered some problems.  I want to know who (if anyone) else has seen
 >similar things.
 >First, my 20 meg Seagate drive generates seek errors (it NEVER did this
 >before)  It is so bad I'm going to replace it as soon as I complete a
 >"women and children first" backup.  These seek errors persist for
 >certain files even when the speed is bumped down or extra wait states
 >added to hard disk requests.

  Do these errors persist when you take the Inboard board out of the machine?
  I would try that first, before replacing the hard disk.
  
 >Also, the machine now fails on occasion to
 >boot up.  The disk whirrs, but no video (not even a cursor) comes on.
 >This means no memory count, no boot, - just a big desk heater.

  Same recommendation -- check to see if it still misbehaves when you 
  "uninstall" the Inboard PC 386.
-- 
Wolf N. Paul * 3387 Sam Rayburn Run * Carrollton TX 75007 * (214) 306-9101
UUCP:     ihnp4!killer!dcs!wnp                 ESL: 62832882
INTERNET: wnp@DESEES.DAS.NET or wnp@dcs.UUCP   TLX: 910-280-0585 EES PLANO UD

jcmorris@mitre-bedford.ARPA (Joseph C. Morris) (05/12/88)

In a recent article radon@phoenix.Princeton.EDU (Radon Research) writes:

>I recently installed an Intel Inboard PC 386 board in a real XT and
>encountered some problems.  I want to know who (if anyone) else has seen
>similar things.

I'm currently trying to install an InBoard/386 in an ancient (5-slot) PC
with an expansion chassis...unsuccessfully so far.  Problems seen:

 (1) The system POD's resent the simultaneous presence of the InBoard/386,
     an IBM EGA board (with or without extra memory) and the IBM expansion
     transmitter card (with or without the expansion chassis attached).
     The EGA card works OK in character mode but has trouble addressing
     the graphics memory.  Substituting another IBM EGA board didn't help
     matters but replacing it with a QuadEGA+ board fixed *that* problem.
     Intel says that they never tested that combination.

 (2) With the QuadEGA installed everything boots up OK, but the diagnostic
     disk complains about the IBM monochrome adapter memory.  The mono
     board is in the expansion chassis.

 (3) The unit generates I/O channel errors attempting to write to the
     hard disks in the expansion chassis (WD RLL controller).  This
     may be linked with item (2) as a DMA interface problem.

Intel has these problems somewhere in its queue for its compatibility lab.
Stay tuned for further developments as they occur.  Film at 11.

wtm@neoucom.UUCP (Bill Mayhew) (05/13/88)

I just installed an Inboard 386 in the Epson Equity III that I am
using right at this very instant.  Here are some observations:

1.  Sometimes requires turning the power switch on and off several
times before the POST starts running.

2.  Will not boot at all with the BUS speed switch on the back of
the equity set to 8 MHz; has to be 6 MHz.

3.  Hangs while reading the config.sys file, if you try to use the
Intel-supplied iemm386.sys ems memory emulator.

4.  The only way to reboot is hitting the power switch.  The front
panel reset switch or the 3-finger-salute (ctrl-alt-del) won't do
it; they just hand the system.

4.  Boots up in a very slow speed of 4.4 as reported by Norton's
version 4 SI program.

5.  Running Intel-supplied "speed.exe 4 256" increases the speed as
reported by SI to 14.4.

6.  Running 386-to-the-max improves things a whole bunch!!  The
386max EMS emulator DOES work, the SI is increased to 16.7, the
3-finger-salute is restored to proper functionality.

7.  It DOES run windows 386, but much, much more slowly than my IBM
model 80.  This is surprising, since the RAM on the inboard is 32
bits wide.  Must be the fact that this equity has an EGA board,
which is much less efficient than the VGA on the model 80.  (In
this case, both '386s are 16 MHz.)

All in all, I have to give the the inboard a lukewarm reception.
It really doesn't speed normal DOS operations all that much (except
for spreadsheet recalcs, which are admittedly much faster. In fact
about 1/2 the time on the '386).  There are a lot of sort little
gremlins that might rise up and bite a neophyte user.  If you are a
power user, you'll probably know what is going on, and be able to
deal with it.  If you use the inboard, I very strongly recomend
getting 386-to-the-max as well; it fixes a lot of things that are
broken with Intel's software.

As for hardware compatibility; it seems excellent once the dang
thing is booted.  The flaky start up seems to be only a power up
problem.  The peripherals in use on this system are:

WD AT disk controller / Rodime 45 meg 3.5", 28mS drive
AST EGA 3+ / NEC Multisync
Taiwan special generic serial/parallel board
EPROM burner
Inboard with 3 megs of memory
Motherboard floppy / 1.2 meg / Archive tape backup
Motherboard serail port
Motherboard clock/cal
DOS 3.2

--Bill