werner@milano.sw.mcc.com (Werner Uhrig) (03/11/90)
In article <160700094@primerd> barry@primerd.prime.com writes: > > [ Barry reports that he has problems when trying to boot with the > powered off Maxtor in the SCSI daisy chain off his Mac SE...] your report is interesting as I experience similar problems in a very similar setup; to wit: SE with 4meg, running OS-6.0.4, internal Quantum-105S at address "0", 6+3 feet SCSI-cables to a Motorola tower containing a Newbury 4380S at address "1" and a Maxtor 4380S at address "4" at the end of the chain with onboard terminators and jumpers set to "internal power to terminators" AND "provide terminator power to bus" - the latter because I assume that it will allow external power to reach MAXTOR's terminators when it is not powered up. I use the LARGE drives just like Barry: power off when not needing them, SilverLining partitioned, SCSI-Tools CDEV to mount them on the fly, etc ... when the SE hangs and I have to reboot, I *HAD* to disconnect the external drives from the SCSI-bus (I do not recall if that was only true when I did not wait at least 5 minutes - but I will test that "soon" ...:-) until I decided to experiment further: I also have a TEAC-150meg tape drive, which I daisied after the Maxtor (and the TEAC also has onboard terminators, which did not seem to matter at all), and which I also like to keep powered off most of the time ... On an impulse, I added one of those external terminators from Apple into the cable just before the TEAC - and I have had no rebooting problems since, no matter which drives are on or off. I've tried several times to find someone (or some pointer to a write-up) to get the definite poop on terminators and related hardware concepts - but "no joy" so far.... >after the "smilely Mac" screen. In a reboot after turning off the Maxtor, >the "smilely Mac" screen often displays for 10-20 seconds, during which >time there is heavy activity on the internal hard drive. same here. One item I suspected but have since discarded was the theory that the "StartUp Device" might point to the Maxtor but not so; another theory was some INIT on drive 0 that is trying to mount a partition on the Maxtor, given that it "sees" it's address on the chain ?!? but I discarded that one also! >I've tried doing a shutdown after powering down the Maxtor followed by a >power off/on sequence on the SE. This only works if I let the Mac sit >with the power off for more than about 5 minutes (could there be a >capacitor that has to discharge?). if you have either no or a rundown battery, it takes about 5+ minutes for the PRAM-values to reset to default-settings (boot from drive 0, I believe, and 0 is normally the SE-internal drive, unless you changed some jumpers) - but I do not really suspect that to be the problem. I DO suspect that you (also) have the jumpers set on your Maxtor embedded SCSI-control circuits to power the bus, and there might be some caps there that take 5 minutes to discharge (with side-effects unknown to me) > The only other _sure_ cure seems to be >shutdown the Mac, turn off Mac power, disconnect the SCSI cable from the >SE, power on (this results in a normal boot), shutdown, power off, >reconnect the cable, power on. same here. hmm, maybe when the SE's circuits see "someone else" on the bus trying to power the terminators on the bus, it stops trying to power the bus herself and looses the internal hard drive somehow ... gee, I better stop speculating and go find a real hardware guru - I know we have some around here somewhere ... :-) btw and on the other hand, this fits what happens when the StartUp Device cannot be "seen" at boot time - boot is attempted from other drives. >I thought that there might be an issue with powering the terminator, so I >brought home the internally terminated Syquest drive I use at work and >changed the hookup to be SE <-> Maxtor <-> Syquest. In order for this >configuration to work, the Syquest drive has to be powered on. there seems to be a pattern here, doesn't there. Syquest, just like my TEAC, seems to require internal power to the terminators. > When I retried the Maxtor startup & shutdown sequences described above, > I got the same type of SE hang when I did a restart/shutdown and the same >remedies were effective. This would seem to rule out powering of the >terminator as a factor. well, not really, as my "successful remedy" might indicate (I still don't understand the electric principles of why it is working). Try what I did (hook one of Apple's external terminators into the cable just before the Syquest) and tell me/us what happens... It might require that you believe in miracles, btw, else it might not work .... :-) >The really funny thing is that I'm pretty sure (can't be positive) that >when I first started using the Maxtor in early January I was able to >reboot normally after turning off the Maxtor. I'm looking for help in >figuring out what might be causing the hang and what might have changed >since January. Why is turning off power on the Mac for less than 5 minutes >not effective whereas disconnecting the cable is? hmm, Barry, played with the jumpers in any of your hardware lately? yours in speculating (when it comes to hardware) Cheers, ---Werner PS: I am cross-posting this article to comp.sys.mac.hardware (where the original should have gone to) and the newly founded comp.periphs.scsi (where I hope to find some light for my darkness of electric ignorance) -- --------------------------> please send REPLIES to <------------------------ INTERNET: werner@cs.utexas.edu or: werner@rascal.ics.utexas.edu (Internet # 128.83.144.1) UUCP: ...<well-connected-site>!cs.utexas.edu!werner
barry@primerd.prime.com (03/14/90)
Thanks for your lengthy reply posting. I realized after I posted to
c.s.m that a better place would have been c.s.m.h and I reposted
there.
Your comment about the Maxtor powering the bus was interesting. It
was my intention to configure the Maxtor so that it was always
externally terminated and I planned to remove the jumper, if present,
that caused the Maxtor to power the bus. When I was fiddling with
the drive, I was short of jumpers, so I'm pretty sure I would have
removed any jumpers that weren't needed. Unfortunately, I can't be
sure of how the Maxtor is configured w/o removing it from the case,
which I'm loathe to do because of the hassle involved in installing
it in the case (a real exercise in "just fits" engineering). Next
time, I'll be sure to keep a written record of how the disk is
configured.
Based on your comment, I tried a test when I got home today. I
powered up the Maxtor, used it, dismounted the partitions, and then
turned the Maxtor off. I then used the SE for other things for about
5-7 minutes, and then restarted. There were no hangs or other
problems with this restart. Thus, as you surmised, the problem might
be capacitors in the Maxtor, not the SE.
So, it appears that I can reboot the SE if the Maxtor has been off
for more than about 5 minutes. The pain factor of this regimen isn't
high enough to justify removing the Maxtor from the case, inspecting
the jumpering (which I think is already set to not power the bus),
and reinstalling the Maxtor.
When I said that I thought I used to be able to restart w/o problems
after turning off the Maxtor, I probably was recalling incidents
where the Maxtor had been off long enough to avoid the problem. The
jumpers haven't changed since I configured the Maxtor.
I didn't mention this in my original posting, which was detailed
enough without this additional information, but I've got another
Maxtor 3280 that I use with my Mac IIcx at work. The configuration
there is Mac <-> Maxtor <-> Syquest. I'm 99+% sure that the Maxtor at
work is jumpered the same as the Maxtor at home (I did them both the
same afternoon). Yet, if the Maxtor at work is connected to the Mac
IIcx, the Mac IIcx won't boot unless the Maxtor is powered; this
happens both when I use the Syquest as the terminator or use an
external terminator.
So, the Maxtor at home can be off when SE boots, but Maxtor at work
must be on when IIcx boots. Is this due to difference in the Macs
or to difference in Maxtors?
Ain't SCSI fun?
Regards,
Barry
------------------------------------------------------------------------
Barry Wolman | barry@s66.prime.com
Principal Technical Consultant | 500 Old Connecticut Path
Prime Computer | Framingham, MA 01701
| 508/626-1700, ext. 1100
------------------------------------------------------------------------