[comp.sys.amiga] 2090 Problems

bmacintyre@watsol.waterloo.edu (Blair MacIntyre) (06/03/88)

I'm posting this for a friend ....

"We run our club BBS on a B2000 with a 2090 card.  During the day, the computer
is used by a local dealer.  Whenever this local dealer runs an overscan demo
we have hard disk problems.  I seem to remember that this bug was dealt with,
however, Commodore Canada says we have the most recent driver software
available.  Could someone email us the newest driver (to this account)?"


--
Blair MacIntyre (bmacintyre@watsol.waterloo.edu)| My spelling is my |Amiga  ///
 "Help, help ... I'm being repressed"-M. Python | own, so I try to  |Power!/// 
                                                | keep is unique!!  |  \\\///  
 ( This space intentionally left blank ... )                            \XX/

bobb@tekfdi.TEK.COM (Robert Bales) (06/16/88)

In article <7192@watdragon.waterloo.edu> bmacintyre@watsol.waterloo.edu (Blair
MacIntyre) writes:

>"We run our club BBS on a B2000 with a 2090 card.  During the day, the computer
>is used by a local dealer.  Whenever this local dealer runs an overscan demo
>we have hard disk problems.  I seem to remember that this bug was dealt with,
>however, Commodore Canada says we have the most recent driver software
>available.  Could someone email us the newest driver (to this account)?"

I have the same hardware setup, same problem. When I run a program that I
wrote which uses overscan, I get the read/write error requester. No permanent
harm, except that to the owner which inevitably accompanies the first error
which occurs on his hard disk. The date on my driver corresponds to that given
as the correct version by Commodore. Apparently, the problem has not been
totally fixed.

   Bob Bales
   Tektronix, Inc.

I help Tektronix make their instruments. They don't help me make my opinions.

Doug_B_Erdely@cup.portal.com (06/16/88)

 I am having some severe problems using my 2090 controler and Miniscribe
40 Meg ST506 drive. It seems that the Format command that is used does not
report any errors to the 2090 controller. What is commodore doing with the
extra 2 cylinders on the Hard Drive anyway? I have run into about four nasty
little errors and am getting disappointed/pissed about the lack of a low-
level formatter and the garbage manual that came with the controller.

 I used the PREP command to get rid of the two that came documented with
the drive but not sure that worked either. Is there a program to tell you
where the error is (Cyl/Head) and what the hell is a block offset and how
do I find out what they are. Frankly PREP sucks people! I like my HD and my
Amy controller but am disgusted with the 2090 Docs and formatting and error
checking routines.     PS. I am using my friends account.....Greg Kanis
       **** Why do people always lick the centers out of Oreo's First? *****

cjp@antique.UUCP (Charles Poirier) (06/17/88)

In article <2137@tekfdi.TEK.COM> bobb@tekfdi.UUCP (Robert Bales) writes:
>In article <7192@watdragon.waterloo.edu> bmacintyre@watsol.waterloo.edu (Blair
>MacIntyre) writes:
>
>>Whenever this local dealer runs an overscan demo
>>we have hard disk problems.  I seem to remember that this bug was dealt with,
>>however, Commodore Canada says we have the most recent driver software
>>available.
>
>I have the same hardware setup, same problem. When I run a program that I
>wrote which uses overscan, I get the read/write error requester. No permanent

I don't use overscan.  But whenever I have a high-resolution (640x200
or 640x400) by four-bitplane screen in front, I get a read/write
requester 100% of the time for hard disk reads or writes.  Yes,
absolutely 100% of the time.  With any other type of screen in front,
not a single problem.  I guess you could say I'm annoyed.  I use the
2090 controller and Quantum Q280 SCSI drive.  I have the "November"
hddisk device, and rev. 4.2 motherboard.

-- 
	Charles Poirier   (decvax,ucbvax,mcnc,attmail)!vax135!cjp

   "Docking complete...       Docking complete...       Docking complete..."

matt@beatnix.UUCP (Matthew D. Shaver) (06/19/88)

In article <2137@tekfdi.TEK.COM> bobb@tekfdi.UUCP (Robert Bales) writes:
>I have the same hardware setup, same problem. When I run a program that I
>wrote which uses overscan, I get the read/write error requester. No permanent
>harm, except that to the owner which inevitably accompanies the first error

I have a similar problem if Itry to use Paint from my hard disk.  If I
am using hi-res with or without interlace I get errors when I try bringing
up the load or save requestor or when I actually try to actually load or
save a picture.  It has been my experience that clicking retry on the
system requestor always recovers the operation but it has completely
discouraged me from using DPaint from my hard disk.
I had similar problems trying to use a demo from the fish disks which
displays CT scan pictures.  It too was using hi-res mode.
==-==-==-==-==-==-==-==-==-==-==-==-==-==-==-==-==-==-==-==-==-==-==-==-==-==-==
Matthew D. Shaver --- UUCP: ..!{sun|lll-tis}!elxsi!beatnix!matt
                      USPS: ELXSI Ltd., 2334 Lundy Pl., San Jose, CA 95131
                      BELL: (408) 942-0900

schabacker@frambo.dec.com (CB, posted by generous Tim) (06/28/88)

> [Various people complaining about problems with hard disk access
> while having overscan/hires (deeper than 3 bitplanes) screens in
> front]

Well, as mentioned in my introducing message, I have a 2090 with a
70MB Micropolis and a 20MB Epson hard disk connected to it. And as
you people might know, we have the PAL video norm over here, giving
you a MINIMAL vertical overscan of 56/128 lines compared to NTSC. 
I experienced the problems descriped, too. First let me say that this
problem is only present when working with 4 bitplane (or deeper <grin>) med-res/hi-res (min. 640 pixels wide) screen in front, which
has some or sometimes NO overscan, due to Chip-DMA (Agnus, that ole
lady) colliding with the DMA from the 2090. Well, it's really the
hddisk.device in ChipMem not getting enough CPU cycles to do it's
job, but read on.

Here is what I did to get rid of it:

1. First make sure that your hddisk.device doesn't get into ChipMem or
pseudo ChipMem ($C00000-$C80000) on A500/B2000's.
So place a "fastmemfirst" or "slowmemlast" BEFORE the "binddrivers"
in your Startup-Sequence. (This was mentioned some time ago by Andy
Finkel)

2. Then you might do some trick that I found to work most of the time,
even BEFORE I did #1. Change your preferences, so that your Workbench
starts at the leftmost position. This seems to help by some magic to
relieve some of DMA strain. Don't ask me why (Jay?).
This works only with programs that "clone" the Workbench screen like
ProfPage, SuperBase Professional, Shakespeare, etc.

The 2090 is the fastest Amiga HD controller I've seen so far, really
nice design, no comparison to C-Ltd. I have a long list of diskperfs
available, if inquiring minds wanna know. :-)
My Micropolis tops ie. at 238312 bytes/sec read and 174762 bytes/sec
write (FFS 34.76 on gamma7)  . And that's an ST-506 drive!!!

- <CB>

P.S.
Could someone please e-mail me the newest version of hddisk(.device)?
Mine is Version 33 (rev 37) (Mon Oct 26 11:06:14 PST 1987), a little
bit out of date, no?

--   __  __    
  / /   |  \ \    <CB> aka Christian Balzer - The Software Brewery -        //
 / /    |   \ \   UUCP : decwrl!{frambo|fra03|fra04}.dec.com!schabacker    //
< <     |---<  >    or : schabacker@{frambo|fra03|fra04}.dec.com       \\ //
 \ \    |   / /   CIS  : 71001,210 (Be brief!) | Phone: +49 6150 4151   \X/
  \ \__ |__/ /    Snail: Im Wingertsberg 45, D-6108 Weiterstadt, F.R.G
                  "If there is something more important than my ego, 
                   I want it rounded up and executed."

If you think this .signature should shrink your're right!