ovdluhe@ethz.UUCP (Oskar von der Luehe) (07/18/90)
In article <1671@yenta.alb.nm.us> dt@yenta.alb.nm.us (David B. Thomas) writes: >dougm@lakesys.lakesys.com (Doug McIntyre) writes: [discussion on odd/bad blocks deleted ...] >> Second. I picked up a WD2010A from a mail order place for the 3085 >>so I can use tracks 1025-1170.. The WD2010A doesn't work with my system.. I keep hearing this message; i. e., a WD2010 not working in a 7300/3B1. How about exchanging experiences on this topic? Maybe we find out about what's wrong. So here goes my story: System: Unix-PC 7300, 512kB on-board RAM, 20MB Miniscribe 3425 The motherboard has a "Rev. G" sticker on it, and every- thing seems to indicate that the system was assembled in Summer 1985. HD Controller: WDC'84 WD 2010 B - PL used as replacement for the ori- ginal WD1010. Symptoms: HD works fine, but access to Floppy is gone!!! *ANY* access to the floppy crashes the kernel "panic: gfpwrite: Can't verify write to WD2797" The diagnostic disk won't boot. Replacing the WD1010 HD controller fixes things immediately. Sure enough, the *FLOPPY* controller is a WDC '82 WD2797 A - PL. What's happening? Don't the 2010B and the 2797A go along with each other? I don't have schematics, and my hardware knowledge goes back to times when vacuum tubes were still common. Anybody an idea? What are the symptoms of failure of the rest of the world? >Do anyone know if the 3.51/3.51m distinction makes a difference here? I doubt it! > David Oskar -- Oskar von der Luehe UUCP: "...!unido!cernvax!ethz!ovdluhe" Institut fuer Astronomie SPAN: vdluhe@20641.span ETH-Zentrum DECNET: 20641::VDLUHE CH-8092 Zuerich, Switzerland
thad@cup.portal.com (Thad P Floryan) (07/20/90)
chad@cup.portal.com (Chad The-Walrus Netzer) in <31882@cup.portal.com> writes:
I have this same problem (ie. WD2010 chip doesn't work) and I'm also
using the Microbotics 3085, but I have a 2010(B) chip (Not A). I've
tried it with two different "B" chips, but neither worked correctly.
I thought it may be due to a motherboard incompatibility (I have a
Rev. B board) but Thad Floryan has said that he has a 3085 running
perfectly with a Rev. B Motherboard using a WD2010B, so I am at a
loss. What motherboard revision do you have? I want to test the chip
and drive on my revision C motherboard to see if that has any effect,
and I'll let you know what happens.
Anybody else have any ideas or experiences with the WD2010 and/or the
Microbotics 3085. I'm particularly interested to know what
motherboard revisions people have and haven't gotten it to work with.
``Microbotics''?? Wrong newsgroup! :-) The drive in question is "Miniscribe."
But, seriously, there DO appear to be a number of systems that are unable to
function with the WD2010 and, since Chad lives near me and is working with me
at NIS this Summer, we WILL attempt to find and post the solution.
The "Fire and Water Damage Restoration Team" should have my office back in
shape by this weekend and I'll then be able to consult the ol' schematics.
There ARE some differences in the disk driver circuitry flanking the 26LS31
and 26LS32 chips (the RS-422 disk data driver and receiver chips) and perhaps
those differences source these problems; hopefully we'll know soon.
The UNIXPC breeds some s-t-r-a-n-g-e problems. For example, a few weeks ago a
friend bemoaned his 3B1 reporting a "parity error" and going belly-up EVERY
time he requested a screen-dump-print or a long listing. "Yeah, sure," I said,
"printer parity errors. Uh huh." After thinking about it for a few minutes, I
consulted the schematics and instantly noticed that a spare buffer in the
74LS245 servicing the parallel port was ALSO used to coax the PERR* signal on
its merry motherboard meanderings. Turns out that previously the printer cable
underwent the "joys" of live insertion (i.e. was plugged in with both the
printer and the system powered up) and a short ensued ... thus zapping the
74LS245 and the PERR* gate such that anytime the printer requested flow control
it would affect the PERR* buffer gate and halt the system. Replacing the
74LS245 solved the problem.
There ARE enough differences between the various 3B1 motherboard revisions to
suggest possible interaction with a different disk controller chip. Who knows,
perhaps the impedances of the WD2010 differ from the WD1010? Chad and I will
be trying some things and will report the results.
Thad Floryan [ thad@cup.portal.com (OR) ..!sun!portal!cup.portal.com!thad ]thad@cup.portal.com (Thad P Floryan) (07/20/90)
chad@cup.portal.com (Chad The-Walrus Netzer) in <31882@cup.portal.com> writes:
... but Thad Floryan has said that he has a 3085 running perfectly
with a Rev. B Motherboard using a WD2010B, so I am at a loss.
Hmmm, I don't remember saying "Rev.B", but averaging only 2 hrs sleep/night
since the ol' fire I'm surprised I even remember my name. :-) The systems
aren't where I can check them right now, but my recollection is a higher
rev level. Will know by this weekend.
Thad Floryan [ thad@cup.portal.com (OR) ..!sun!portal!cup.portal.com!thad ]ovdluhe@ethz.UUCP (Oskar von der Luehe) (07/20/90)
In article <31882@cup.portal.com> chad@cup.portal.com (Chad The-Walrus Netzer) writes: >In a previous article (Doug McIntyre) writes: >) Second. I picked up a WD2010A from a mail order place for the 3085 >)so I can use tracks 1025-1170.. The WD2010A doesn't work with my system.. >)Does the B version work better with these systems? Can I pick up a B version >)from someone? Is there anything that can modified to make this chip work? >)I thought the 2010 was about 99% compatible with the existing 1010.. > > I have this same problem (ie. WD2010 chip doesn't work) and I'm also >using the Microbotics 3085, but I have a 2010(B) chip (Not A). I've tried it >with two different "B" chips, but neither worked correctly. I thought it may >be due to a motherboard incompatibility (I have a Rev. B board) but Thad >Floryan has said that he has a 3085 running perfectly with a Rev. B >Motherboard using a WD2010B, so I am at a loss. What motherboard revision do >you have? I want to test the chip and drive on my revision C motherboard to >see if that has any effect, and I'll let you know what happens. How about collecting the experiences of those folks who have problems with the WD2010 in their Unix-PCs? WHAT EXACTLY are the symptoms of failure? Here is my story, for example. System: Unix PC 7300 (that is, *not* the 3B1) with 512 kB on main board, and a 20MB Miniscribe 3425 harddisk. There were no hardware mods at that time, other than replacing the original WD1010 HD controller with the 2010. The motherboard has a little sticker saying that it's "Rev. G". Date stamps etc. point at an assembly time in summer 1985. Running vanilla 3.51 at the time of interest (I have 3.51m now). HD controller: WDC '84 WD 2010 B - PL, replacing the original WD1010. Problem: System boots fine, no problems with the harddisk drive, but the *FLOPP* is gone!!! Any attempt to access the floppy drive while running Unix results in a kernel crash: "panic: gfpwrite: Can't verify write to WD2797" The diagnostic floppy won't boot. Needless to say, that the WD2797 A -PL is the Floppy controller, which is located near to the HD controller. My first reaction was that I might have zapped something during the HD controller replacement. But the problem went away when replacing the 2010 with the 1010, so I suspect that the 2010 is the culprit i.e., not as pin-compatible as advertised. I have no Hardware manual and only little means to sort out a HW problem by myself. Maybe somebody knows about a fix. -- Oskar von der Luehe UUCP: "...!unido!cernvax!ethz!ovdluhe" Institut fuer Astronomie SPAN: vdluhe@20641.span ETH-Zentrum DECNET: 20641::VDLUHE CH-8092 Zuerich, Switzerland
scs@lokkur.dexter.mi.us (Steve Simmons) (07/21/90)
ovdluhe@ethz.UUCP (Oskar von der Luehe) writes: >How about collecting the experiences of those folks who have problems >with the WD2010 in their Unix-PCs? WHAT EXACTLY are the symptoms of >failure? Here is my story, for example. (whow, deja vu) AT*T 7300, .5MB motherboard untouched since original manufacture almost 5 years ago. Using a Fujitsu 71MB drive and 3b1 top. Drive ran fine with 1010. Inserted WB-2010B, and immediately started hearing recal sounds from the disk every couple of minutes. It seemed that large numbers of disk block had suddenly become bad. Replace the 2010, and the blocks were good again.