> It should work, provided you have the license code,
> IPL app, and mainframe
> image installed. I have a couple of these machines
> myself.
>
> Peace... Sridhar
>
> On Tue, 15 Jan 2002, Colin Eby wrote:
>
> > Date: Tue, 15 Jan 2002 16:10:00 -0800 (PST)
> > From: Colin Eby <out2sea00(a)yahoo.com>
> > Reply-To: classiccmp(a)classiccmp.org
> > To: classiccmp(a)classiccmp.org
> > Subject: Re: IBM R390 IPL woes...
> >
> > Folks --
> >
> > Tomorrow I pick up the IBM 5362. But I have
> another
> > obscure piece of hardware I'm playing with. Anyone
> > ever hear of an IBM R390. This is an RS/6000 with
> an
> > SBC mainframe in the MCA. The share IO, but
> nothing
> > else. I've been rebuilding this thing as my UNIX
> > workstation at the office. The problem I'm having
> is
> > getting the mainframe to IPL (of course). This is
> > probably, because -- fool that I am -- I FDISK'd
> the
> > box and upgraded to AIX5L. Does anyone know if
> there's
> > a binary compatibility issue with the mainframe
> > board's system files, and the newer AIX binary
> set?
Sridhar --
Thanks, for the note. When I wrote this earlier in the
evening, I only had the original disks. After some
considerable web searching I found a reference to the
P390 ftp site. The reference was not I hasten to add,
to be found anywhere on IBM's support pages. The docs
in the ftp site don't specifically endorse AIX5L, but
it did say the system dumps I've been having were
common when upgrading to 4.3.3. And I should therefore
upgrade the r390 binaries I had. Low and behold there
they were on an anonymous FTP server. And shockingly
enough I am now sitting in front of an MVS
installation that no longer system dumps all over me.
Technically it still doesn't IPL, but I'm positively
gleeful anyway.
Thanks,
Colin Eby
Senior Consultant
CSC Consulting
__________________________________________________
Do You Yahoo!?
Send FREE video emails in Yahoo! Mail!
http://promo.yahoo.com/videomail/
Tothwolf wrote:
>Did you try flexing this cable while checking its resistance? I've
>replaced a few similar cables in other printers that had intermittent
>breaks in the conductors when they were flexed a certain way.
will try it. The cable was near-zero resistance,in a write-head position
about halfway across the page, so I don't think that's it. Also the missing
line of pixels is all the way across the page, so it's not an intermittent
open over the range of motion across the page.
>Good luck with the glue, super glue and glue designed for polystyrene
>plastics does not bond ABS plastics very well.
Yeah, I know. I was being a bit facetious about the "good as new".
>A cheap transistor checker? A multimeter can sometimes be used, but it
>helps to have the pinout and specs for the transistor on hand.
Hm. Replacing all of them only runs about $4.50. But Tony suggested a
1st-order test with my VOM (see below).
>A dull knife hit with the same
>rubber hammer on the glued seams can also work,
This brick seems to have no glue. On the "wall" side (with the plug blades
sticking out), it has three *deep* holes with what look like black painted
metal or hard plastic beads at the bottom. Those act like bolts that are
holding it together. I felt the label, no screw heads obvious under it, but
I'll peel it up and check again. It's not right anyway, output is nowhere
near the 9V it claims :-( .
> 4) What am I likely to find toasted in there?
>take an ohmmeter and see what the primary and secondary winding
>resistances are.
OK, will try it if/when I get the thing open.
>....simply replace the whole brick.
What's a good source for ancient Mac parts at bargain prices?
>Of course, if you own a coil winding
>machine and have the proper wire on hand,
Um..... What's a good source for ancient Mac parts at bargain prices?
--------
Matt London suggested:
>OTOH, Araldite (an epoxy) works a treat in my experience
If everything else starts working but the cover pieces won't stay together,
I'll try it.
----------
Tony Duell said:
>How many dots is this printhead? 25 connections could be 24 dots + common
>or sowething like that. I would have suspected 1 driver transistor per
>dot, though, and 9 dots seems very small for a modern-ish inkjet printer.
Either 8 or 9. I think 8, but don't have any good evidence for that other
than the row of 9 (what look like) driver transistors, of which one is
slightly offset. 8 pixels would just about square with the width of the
missing row vs. the distance between rows when I try to print a solid black
block, but 9 might work too.
The printer is 1991-93 era (about?) and was 1st generation ink-jet for
Apple. I guessed 8 pixels based on the number of big wide traces coming
across the ribbon cable and on the theory that it takes one big wide trace
per pixel. Do they do some wizardry in the print head where one driver
drives 3 pixels, but serially, or some such?
>There's no easy way of telling. At least not without seeing the machine,
>which is kind of difficult :-)
I have a digital camera which produces .jpeg file pictures, about
450kbytes/picture. I could let you see images of many parts of it, if you
can decode jpegs.
Alternately, I have the shipping carton for it, so it becomes a matter of
whether I'll pay more for 2-way shipping to London or for the parts I fry
trying to debug the thing....
>Is this AC or DC? In other words is the wall-wart just a transformer, or
>does it contain a rectifier and smoothing cap as well?
I made both measurements DC. I never thought to check AC before breaking
it, as the label said DC output. I also never thought to check AC after
breaking it. I did try the DC measurement backward, after it broke, and it
definitely had (about) a 0.2 V DC level on it, same sense as the correct
output had been before.
>Once you get inside, it's easy to check/replace <the fuse>.
I'll look forward to it. Is it easy to identify?
>Could well be. Those sound like printhead drivers. I'd be inclined to
>trace all the connections from the printhead ribbon cable connector to
>see where they end up, though.
:-) Uh oh. Tony is trying to educate me, I'll bet. What was it Huck Finn
said about Aunt Polly? <sigh> OK, I'll give it a shot.
>Read that as 2SB1243 -- which is not in 'Towers International Transistor
>Selector'. Can you check that number, please. In fact, please post all
>markings on the original transistors.
I thought I did, unless there are more markings on the side against the PC
board? I'll look again.
>A good first check is to desolder them and then use the ohmmeter function
>(on an analogue VOM) or diode test function (on a DMM) to check that the
>base-emitter junction and base-collector junction test as diodes. And
>that the 'resistance' between the collector and emitter is very high both
>ways round.
I knowed it, he's a-tryin' to eddicate me! :-)
OK, I'll dig out the soldering iron.
>To find out which pin in base, which is collector, etc, you either have
>to look the transistor up in a databook ('Towers International Transistor
>Selector' is a well-known one in the UK that most hardware hackers have
>on their bench) or trace out enough of the surrouding circuitry to work
>out at least which pin is the base.
I think I get to cheat. I remember seeing "E", "C", and "B" markings on the
PCB.
----------
Pete Turnbull said:
>Yes, it's a PNP power trannie, in an ATV package, which is a bit like a
>TO220 but without the metal tab. It has a beveled edge on the front top,
>instead of a metal tab on the back.
This sounds familiar.
>Pin order is ECB (opposite of the common TO220).
That is what I remember from the board, though I'll check.
>Near equivalent 2SD1864.
Thanks! I'll comparison-shop for them both.
--------
All within 24 hours. Is this a *great* list to be on or what?
Will work on collecting info tonight, responses tomorrow. Thanks very much!
- Mark
On Jan 15, 19:20, Tony Duell wrote:
> >
> > On Jan 14, 20:19, Tony Duell wrote:
> > > >
> > > > Hi, I've just been having a go at fixing an old Acorn AKF40 monitor
(a
> > >
> > > OK.... Is this better known as something else? I don't think Acorn
ever
> > > designed their own monitors, did they?
> >
> > They used Philips or Microvitecs. I think that model is a rebadged
Philips
> > VGA. There's a stock LOPT fault on one of those, but I'm not sure if
it's
>
> >From the component references in the fault description you gave, I'd bet
> it's a Philips...
I think so too, but I can't be sure.
> I think the AKF60 was a Microvitec.
It is.
--
Pete Peter Turnbull
Network Manager
University of York
On Jan 15, 14:17, Eric J. Korpela wrote:
> Apparently there are are lots of Araldites. Araldite 2020 and 2026 are
> clear. There are a few dozen other Araldites as well. I'm sure the stuff
> you generically call Araldite has a type number as well. It's possible
> the type varies depending upon how the epoxy is branded.
It's sold here as "Araldite". Full stop. No number. It's a retail
product that's been around for decades (Araldite was invented by Aero
Research in the UK and came to fame during WW2; it was sold to Ciba after
the war). We can get the numbered types as well, but they're less common;
I'd need to go to an engineering (or electronics or microscopy) supplier
for most of them. Unfortunately that means I can't tell you what the
equivalent number version would be in the States :-(
The standard stuff is a 50:50 mix, comes in blue and yellow tubes, and like
all epoxies is thermoplastic. It softens at around 150 deg.C. You can
vary the mix slightly to make it a little more or less rigid/hard. The
normal mix sands/files well. Does that help?
--
Pete Peter Turnbull
Network Manager
University of York
Hi there. I am looking for three identical IBM PC Server 310's. I'd
prefer the PCI/MCA one, but I'll take the PCI/ISA one. I am also looking
for one or two PCI/MCA PC Server 320's. (Part 8640-MD?).
Peace... Sridhar
Folks --
Tomorrow I pick up the IBM 5362. But I have another
obscure piece of hardware I'm playing with. Anyone
ever hear of an IBM R390. This is an RS/6000 with an
SBC mainframe in the MCA. The share IO, but nothing
else. I've been rebuilding this thing as my UNIX
workstation at the office. The problem I'm having is
getting the mainframe to IPL (of course). This is
probably, because -- fool that I am -- I FDISK'd the
box and upgraded to AIX5L. Does anyone know if there's
a binary compatibility issue with the mainframe
board's system files, and the newer AIX binary set?
Thanks,
Colin Eby
Senior Consultant
CSC Consulting
__________________________________________________
Do You Yahoo!?
Send FREE video emails in Yahoo! Mail!
http://promo.yahoo.com/videomail/
Hello Clint,
I think you might find what you are looking for at
www.UnitechElectronics.com This is virtually the only place on earth that makes cables specific to A1200's.
The prices are in Australian Dollars so if you are in the
US you can expect to pay approx. half the listed price.
Jeff Rose (Site owner & Amiga Distributor) is a good friend
of mine.
cheers,
Keith
cavador(a)iprimus.com.au
From: Chad Fernandez <fernande(a)internet1.net>
>I wish I had some 16meg simms..... I would have put 256megs in my 386
>:-) Yes my 386 had 16 simm slots, and would take more than 32 megs!
Yes, but...
I've seen a lot of i386 boards that could take a lot of memory, but many of
those boards couldn't cache more than 64MB. Resulted in awfully slow
performance.
Ken
Tony:
You first question is answered by your second question.
The reason that I propose to "insulate the computer from the world" is to solve the first problem that you pointed out:
"applying the reforming voltage ... to random points in the logic"
If the computer itself was "insulated from the world", the chasis would be able to float so that no actual
voltage was applied to any circuit other than the capacitor being reformed.
You would need to be sure that you did not even touch the computer during the process.
If my theory actually works -- it could save a lot of trouble for computer restorers.
My assumption is that the two sides of the capacitor are isolated from each other.
This seems like a reasonable assumption to me.
But, just in case, I thought I should check with those of you that might have more experience rehabilitating old computers.
Perhaps I just need to find something less precious than my Altair to try it out on ....
Why don't I want to unsolder the capacitors ?
My experience is that I can do a lot of damage unsoldering things.
My soldering skills are not as good as they were when I put the Altair together 25 years ago.
My fingers and eyes do not seem to work as well as they did back then either.
Perhaps I just need to get myself one of those new high-tech temperature controlled soldering irons.
Tony:
Reading your later messages, I notice that you recommended using a VOM meter in the Ohms mode to test the caps.
I have always avoided the Ohms mode of a VOM when working with TTL circuits.
The VOM applies a voltage to its test leads when it is in Ohms mode.
As I understand it, delicate circuits an actually be damaged by these meters.
I have not worked with hardware in years though. Maybe the modern VOM meters are safe.
-Rob
-----Original Message-----
From: Tony Duell [SMTP:ard@p850ug1.demon.co.uk]
Sent: Monday, January 14, 2002 6:09 PM
To: classiccmp(a)classiccmp.org
Subject: Re: Prophylactic replacement of electrolytic capacitors?
>
> I have not yet tried this yet --
> In theory, electrolytics can be re-formed.
> There is a very thin aluminum oxide layer that disappears after a long time.
> Running current through the cap will rebuild the insulating layer
> until current will no longer go through it.
OK, yes, reforming can work...
>
> I want to reform the original capacitors in my Altair "in place" --
> i.e., without unsoldering them.
Why do you not want to desolder them? It'll do a lot less damage than
applying the reforming voltage (a little above the working voltage of the
cap) to random points in the logic.
For example, there are likely to be some 16V electrolytics between the
+5V line and ground. Do you seriously want to try running TTL at 18V?
Desolder the darn thigns!
>
> I have the following setup in mind:
>
> 1. Insulate the computer from the world -- unplug it and put it on a
> rubber table.
What's the point in doing this?
> 2. Use ultra mini test clips to connect to both leads of a capacitor.
> 3. Connect the test leads to one of those lab power supplies that have adjustable DC voltage and a milli-ammeter.
> 4. Connect + on the lab power supply to + on the capacitor (very important).
> 5. Start at zero voltage and increase slowly while looking at the ammeter.
> 6. Stop increasing the voltage when the ammeter reads anything more than a few milliamps.
> (If it draws too much current, the capacitor will heat up and may blow up)
I would include a current limiting resistor (a few k) in series with the
cap. And please desolder it!
-tony
