Having read below I remember some of this now. All the later machines so the 43xx and 93xx series have some kind of control CPU and perform some kind of Initial Microcode Load or IML at power up. On a fully cabled 43xx machine with full "Emergency Power Off" (I think that?s right we always said EPO) cable just hitting the power button on the console would load a menu on the console. One option allowed the machine to be powered up which was done in sequence so that the start surges from each motor starting were spread out over a period of time. I think on the 4381M3 at NERC Wallingford (google for it for lots of info) with two strings of 3380 disks, round tapes and square tapes took it around 30 Minutes for all the peripherals to power up and for the microcode to load into all the controllers, before you could IPL the system. I seem to think many controllers had a floppy disk with the microcode on. Certain the process was a joy to observe as the room changed from being very quiet as the aircon ticked over with virtually no load as everything was powered down, to being a place of , at times almost painfull noise as everything came to life. Dave G4UGM > -----Original Message----- > From: cctalk-bounces at classiccmp.org [mailto:cctalk- > bounces at classiccmp.org] On Behalf Of Arno Kletzander > Sent: 24 September 2014 20:26 > To: cctalk at classiccmp.org > Subject: Re: Inventory of remaining IBM S/360 CPU's. > > Christian Corti <cc at informatik.uni-stuttgart.de> wrote: > > > > We have a 4331, too. Perhaps you already know our site/machine. > > Nope, I just found out about your machine when I saw it on Henk's list. > > > Hmm...I don't remember the specifics after all those years, but basically: > when you switched on the machine, the fans and some minor power > transformer/PSU would power up, the Service Processor would boot from an > old squeaky 8" floppy drive and issue some progress indications on the > terminal. At one point, you could hit a key (combination?) to get into some > kind of setup menu, where you could e.g. enter what sort of device was > connected to each terminal port. You could stay there and browse that menu > to your heart's content, but if you did not enter or did quit the menu, the SP > would try to power on the main part of the machine - you heard a clunk from > some large contactor, the buzz of the main transformer(s) would kick in for > like 1-2 seconds, then everything (including the SP supply) would be shut > down (several more clunks, fans spinning down), leaving the machine > powered off again - without an error indication on the terminal. > > > > One related machine I know of (a 4361) is kept at IBM's Museum in > > > Sindelfingen; we were there to get the pinout of the System Panel > > > > They also have a 4331 AFAIK. At least they took the 3370 drives that > > were part of the machine we got some years ago. We kept the 3340 and > > 3344 drives. > > Perhaps they didn't have it back then, or it was in storage and we were only > shown the 4361 because that was on exhibition. I don't however remember > anybody telling us about a 4331 there. > > > > machine (manuals, schematics and faultfinding charts) but don't > > > really know where to start with our Problem. > > > > MAP - Maintenance Analysis Procedures. > > Yep, that's what they were called :) > > jwsmobile <jws at jwsss.com> wrote: > > I would rely on the MLM's and other service documentation. They have > > pretty good procedures of testing to follow and solution matrices to > > follow to remedy the problems. If you are following those and have > > encountered a problem, then you should have something that can be > > helped with. I've brought up a few systems and you have to go thru > > the tedious process of running all the tests sometimes to find > > something, but things I have had wrong fell out of tests where I would not > have thought to look. > > Of course we tried following those procedures, but there were several entry > points and maybe we just always missed the right one. Somehow the path > always went off into the woodwork, with things like preconditions that were > not met in our case. > > Christian Corti <cc at informatik.uni-stuttgart.de> wrote: > > You haven't asked so far, did you ? (I don't remember) > > Probably not since you joined the list. Keep in mind that of this was more > than ten years ago and I had not pursued those issues much, as the system is > not in my custody, I only get to chat with its current owner a couple times a > year at most, and we both have had other fish to fry through all those years. > But after all, I still yearn to bring that project to some rewarding conclusion :) > > > What drives and OS tapes do you have? > > No (intact or restorable) drives except for the two internal 8" floppy drives, > no software except for two IML (bad original that had been stored inside the > drive, intact backup) and I think one diagnostic floppy that came with the > system. What's there is a 3262 high speed printer and some sort of remote > terminal server that came from a large mail-order outfit (Quelle Versandhaus > F?rth). I'm all for sourcing period peripherals if possible at all, but if we can't, > our alternate idea is to furnish some sort of channel drive emulator with a > modern hard drive, like Mr. Zabolitzky and his guys at Cray-Cyber.org did for > their CDC machines. > > > The 4361 is great because it has the storage director integrated in, > > and you can run a complete system with just a terminal, a tape drive > > such as the 4311 which has the tape controller and drive integrated > > together, and disk drives. > > > > It is a great VM personal system if you have one. A friend had one to > > himself for 2 years at UMSL at one time. Sad these all got broken up > > for the price they brought. > > I shall see whether I can get together with the owner at some time again and > get some diagnosis and work done on the machine. However I don't have > any idea whether that is even possible in the place where the machine is > currently stored. Of course I'm all ears for more suggestions and anecdotes > about these machines and leads towards associated stuff. > > So long, > Arno