;)
Hi all.
Due to a major computer disaster here at work I
haven't been reading list mail and probably won't
for a few more days. So, I'm not ignoring anyone
and if anything is really important e-mail me
directly (bill(a)booster.u.washington.edu).
And, if anyone can tell me why 11 NT servers in
a single domain would crash one at a time every
4 minutes and 20 seconds until the whole network
is gone - please let me know ;).
Bill
> What should be on display considering:
> 1) space limitation
> 2) Mainframes, minis, micros
> 3) What would most attract visitors
> 4) What would most attract funding :)
Three things which have changed radically over the years.are
- 01 Power (including speed)
- 10 Cost
- 11 Display capabilities (on affordable systems, that is)
You might want to consider comparing the above parameters system-by-system.
Software has become less klunky over the years. Try having an ancient word
processor running so people can see, say, the number of steps required to
open a file.
Don't forget output technology...compare an old daisy wheel to a laser, a
superannuated 9 pin to a modern 24 pin, get a thermal printer, and a
thinkjet (really raunchy output, but FAST).
I'm collecting historically significant micros (such as the C64 -- they're
cheap, so everyone got one.) If you can get the dirt on why, for example,
CBM shot themselves in the foot, that would be interesting. Why aren't
Apples more common, for example?
At 01:18 PM 7/15/97 +0000, you wrote:
>simply too small try first: under the width of drive cage, seal the
>grille holes on metal sheet and on case cover on right side, seal
>that underside holes, all with a good tape. On that XT PSU, how much
>airflow blowing out back, weak? Look for a replacement fan that does
>better than this el cheapo fans. Try AT type case if you can, they
>has bigger, powerful fan.
Ok, I checked it out, and it appears that there already is a vinyl plastic
glued to the front, with a 5160 sticker on it. I also put a cardboard cover
in the empty bay, and this seems to work. the PSU is actually not IBM made,
it is an aftermarket brand that has a strong fan in it, stronger than what I
find on NEW Yackard Hells! <G>
>XT case design especially in cooling department is bit goofy, blamed
>by IBM for this ill. :)
It continues to amaze me on just WHY whatever IBM made, as flawed as it is,
the public jumped on it and supported it... look at winsucks 95.....
>> there is good air flow, just they seem to run this
>> hot.
>More power to air flow is needed, see the tips..
>> I never ran these drives before and have no idea how hot they should
>> be. as for power cycling, they get NONE. this system has been running for 2
>> months continous, without any powerdowns since they were installed, so for 2
>> months now, the thermal stability has (hopefully) stayed the same.
>> thanx
>Hey, no problem!
>Jason D.
>
At 09:34 AM 7/15/97 -0400, you wrote:
>First off why bother withg the st225s? They are slow (st251s are
>faster!) and their reliability is at best ok. FYI: st225s live far
I would if I had a ST-251, but I got 4 ST-225's, and 2 of them are in use,
and the other 2 are spares.
they are slow, but they do work nicely. I also like the sound of old
hardware, the new drives you can hardly hear them run, with these, you can
hear them a mile away! :)
The cooling fan on this machine is as loud as the hard disks, and moves lots
of warm air..
>longer if they are cooled. If the fan isn't noisy it isn't moving enough
>air past them, seriously! Also if they are in the smae box as all the
>other hardware they raise the other components temperature lowering their
it is in an XT case, which is huge, and has one slot free for air to move..
and my room is not air condidtioned, but the house is, but the cold never
travels in here! <G>
>life. Heat is not your friend, this is why computer rooms are air
>conditioned an usuialy under 70 degrees F.
>
>Allison
>
>
Denizens of the Bay Area, check this out!
HMR Recycling (www.hmr-usa.com), when I had the opportunity to visit,
reminded me greatly of one of the Surplus Warehouses of Olde. Reasonable
prices, huge assortment to choose from, and 100,000 sq. feet of warehouse
to peruse!
Example: Nearly new-looking Tektronix 7844 dual-beam O-scope, with a pair
each of 7B92A timebase and 7A26 vertical plugins, with cart: $400.
Other examples: Emulex P3000 print server, $20. DEC RRD42 SCSI CD-ROM
drive, $20. Exabyte EXB-8200 2.3 gig 8mm tape drive, $125.
One of my finds there was a pair of DEC storage expansion cabinets at $45
each (bargained down from $50 since I bought both of them). They each
contained a pair of DEC RZ57 drives (made by Micropolis, 1.0 gigs each, all
SCSI). Heck of a deal!
They get new stuff in literally every day. I went there twice on different
days, and found some neat deals both times.
08:00-16:30, Monday-Friday only, minimum purchase $20.00, and you need to
sign the visitor log and wear a visitor badge while looking around. Other
than that, enjoy!
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Bruce Lane, Sysop, The Dragon's Cave BBS (Fidonet 1:343/272)
(Hamateur: WD6EOS) (E-mail: kyrrin(a)wizards.net)
http://www.wizards.net/technoid
"Our science can only describe an object, event, or living thing in our own
human terms. It cannot, in any way, define any of them..."
>>>jpero...
Sorry if this is long but it is important to users of classic machines.
<I gently disagree with this assertion that ST251 series is good than
<ST225 series. I got too many dead ST251 series compared to ST225's
<due to weaker structral design, pushed design by driving steppers too
<hard byond 40ms average and not as rugged as ST225's.
Can't say I disagree. You however confused my assertion of better
performace in terms of access times, with reliability. Its faster and more
storage for less power and heat than two st225s. I never said it was more
reliable.
My expereince is that the st251 is less tolerrent of heat but if cooled it
seems to do as well as most st225s (n=10k+). An experience with a customer
in Austrailia using both 225s and 251s proved it. FYI neither like heat!
<goal set by designers. If one wants faster seeking time, go voice
<coil! I am dreaming wishing that ST251 drive was fitted with voice
<coil instead of stepper and better motor design...sigh.
true, but there are voice coil dries out there that do not hold up well too.
It's a design trade made by some companies thinking the mechanics will never
outlive the electronics. Much of it is process quality and handling. Most
bad drives I've seen were from warm enviornments or dropped in their life.
<reliably. Whole problem was that design of this ST251 spindle motor
<is ill-designed to begin with in my opinion: Trying to fit all the
<coils and one bearing INSIDE that spindle results in weak axle that
<gets bent easily due to this design and the bearings failure rate on
<this one is high too.
It has a problem with gyroscopc precession, this makes it very susceptable
to small shocks when running. The bearing does not like sideloads either.
I agree it could have been better thought out.
generally going to the smaller 3.5" was actaully an improvement in
reliability for disks as it took less power to spin it and also move the
heads. With less mass in the media the motors, bearings and all could be
more rigid for it's size. The lower heat made bearing life better and also
helped dimesional stability.
< Recently I pulled this ST225 out of dumpter and fixed up the XT with
< this hd and LLF'ed it. Gave whole thing to poor guy to use vax at
< college. Still works.
Most 225s were not fails, just retired due to small space available. FYI
you get better results if the disk is LLF'd in situ and well warmed up.
I must have 6 st225s of them all good and solid. I use them but, when I
need speed I try to use others (quantum q540 30mb is nice!).
<I tore down all kinds of drives from old and 3 years old types for
<post-failure analysis which gave pretty good info for me.
Same here, mostly limited to:
St506, st412, st225, st251, micropolus1325 and quantum Q540, Q2190s and a
few other dec disks.
I have an st238 that refuses to die!
<Snip! Tales of positive things about cooling...
<Actually, all drives old and new benefits from cooling and boards
<likes the flowing cooling air.
Very true, vacuuming the vents from time to time also is a good thing.
The big advantage of the newer drives is they consume less power therfore
produce less damaging heat enhancing reliability.
<We have no choice as many do not have AC so they ran in warm air
<but if you are careful, enough fans to make hurriance out of it and
<put a fan or two on monitor as well. My 17" needs two otherwise the
<HOT transistor will pop again. ARRGgghhh.
Tell me! I run microvaxen, Q-bus PDP-11s and s100 systems and 20 years of
experince (and reliability studies) has taught me 90F is a hard ceiling and
to shutdown with fans running if it gets hotter. I know this from my days
at DEC in the mill when my office system (11/23) was fried by an AC failure,
the area hit 98f and the 11 went down for the count. I've been pulling the
plug regularly this year as we've had a lot of warm here in MA. ;-)
Even without AC, every effort to remove heat is good! Even if you can't
lower the ambient, keeping the guts closer to it will help greatly.
My s100 crate has two 5" 120cfm fans pushing air up through the cage and it
runs stone cold. The original layout had an 80cfm fan in the PS section
pulling the air around and the boards ran HOT. It makes a racket but
compared the the 8" drive and considering it's 20 years old... it still
runs.
Even my ba123 microvax, I replaced the fans to blow up as someone put them
in backwards for less noise. You fight convection and lose cooling
efficiencty. Since I have few spares I prefer to keep it going and tolerate
a slight increase in noise. This is something to watch for on older systems
where things may have been replaced but with the wrong (underrated) items.
One last item. MAKE BACKUPS OF EVERYTHING. make duplicate backups and test
the backups for integrity. Disks fail, and the cheapest insurance is a boot
and backup package that works when you replace/repair the offending drive.
Allison
<I'm trying to track-down a mod that Tandy on the Expansion Interface. On o
<of my early EIs (SN# 000352), there was a 6-pin DIN modification, which
<someone at one time told me was to correct some erratic signal problems.
the mod was to bring out ras* cas* and MUX* with terminating resistors at
one or both ends. These signals went through the EI cable but timing was
critical at best. the later EIs derived cas* and mux* off of ras* inside
rather than pipe them over. Generally speaking the mod was one where if it
was mostly working it MIGHT help, sometimes it made it worse.
There was another mod after that called the buffered EI cable...not a good
idea either. Fixing the EI was the solution.
Allison
At 02:45 PM 7/15/97 -0400, you wrote:
>
>One last item. MAKE BACKUPS OF EVERYTHING. make duplicate backups and test
>the backups for integrity. Disks fail, and the cheapest insurance is a boot
>and backup package that works when you replace/repair the offending drive.
>
>
>Allison
>
This is why I like the 20 and 40 meg drives, I can back up key elements of
the OS and configs, and restore them easily. as for 3.5 being more
reliable, they are not, as I bought a Conner CFS-850A 850 meg, and it worked
nicely for a year, then one day I am greeted with this error:
BOOT DISK FAILURE, SYSTEM HALTED.
it suffered a head crash, it would spin up normally, and when the controller
tested the head actuator upon startup, there was a time when the head would
seek to a location on the disk, and GRIND to a halt. the drive never worked
again... I lost 700 megs of stuff, half unreplaceable (my fault, but I have
no tape drive). fortunately the drive did have a 3 year warrenty, and a
replacement was shipped, a 1.2 gig seagate ST-31276A. and here is another
question:
Is this going to last me more than a year? If Seagate made good drives in
the past, will this new one live up to this, or is this drive chock full of
corner cutting?
> Most 225s were not fails, just retired due to small space available. FYI
> you get better results if the disk is LLF'd in situ and well warmed up.
That's doubly true for RLL encoded ST-506/412 drives.
These drives should be LLF'd from time to time anyway. Most HD controllers
have a built-in LLF routine which you can access via DEBUG.
> What should be on display considering:
> 1) space limitation
> 2) Mainframes, minis, micros
> 3) What would most attract visitors
> 4) What would most attract funding :)
I'm afraid number 3 and number 4 will turn you into The Boston Computer
Museum.
Roger Ivie
ivie(a)cc.usu.edu
