On Mon, 15 Oct 2007, Tony Duell wrote: I like the 750. It can be plugged into a 30A 150V outlet without any problems. It isn't as fast as the 780, but you can rebuild 4.3BSD from sources in about 10 hours. I have a goal to decap the gate arrays and take die photographs, but haven't done so yet. The 780 wants 3 phase power, but you can jury rig the supply box to run on 2 phase with an external supply for the air flow sensors. It generates a LOT of heat, so running without airconditioning is a sure path to hardware failure. 730's are dog slow. I'd take a uVax or VS3100 over a 730 except for curiousity sake. Unfortunately, many were cannabalized for the 2901's to repair video games, and the PALs are hard to find anymore. The printsets also appear to have errors (delibrate?) with regards to the PAL fusemaps. I got a second one to reconcile the errors (Thanks Emanuel!), but haven't had time to do so. Clint

Ethan Dicks wrote: Hmm. I didn't find it all that hard to find some. As a matter of fact they seem to find me. After "disposing" of 2 of them (well 3 actually...the third was everything but the cabinetry), I still have 4 (one's in my shop, the others are in storage). If anyone really wants one, I'll make a deal! :-) If any one's serious about it and are out for VCF, we can do some digging in storage (they're packed in tight). The power controller is 3-phase but that's more to get the amperage reasonable rather than anything really needing it. It's just the power controller and it just splits the phases. All of the supplies in the cpu are 110v with standard 15A 3-prong plugs (there's just a lot of them...I think the one in my shop has 5 or 6 supplies). Of course I have plenty of 3-phase in my shop so I don't need to come up with a "creative" solution. -- TTFN - Guy

On Tuesday 16 October 2007 08:37, Mr Ian Primus wrote: It should only be $1k or so to get one shipped cross-country. They're not too heavy. After moving an 11/780 (big footprint) and IBM S/390 (small footprint) into my place (the 780 by myself, and the 390 with the help of a friend), well I can say I feel perfectly fine after the 780 (which I had no help moving), unlike the sore knees I had for a few months after moving the s/390 (which I had help with) (pushing both of them up a 10% or so grade from the alley behind my place to the door). IOW, 11/780 *light*, S/390 (or any similarly sized IBM gear) HEAVY. :) Pat -- Purdue University ITAP/RCAC --- http://www.rcac.purdue.edu/ The Computer Refuge --- http://computer-refuge.org

On 16 Oct 2007 at 10:43, Sridhar Ayengar wrote: Same goes for the comments on 3-phase power supplies. A lot of the old iron really used 3-phase AC, usually employing a star rectifier configuration. Ripple frequency for 60 Hz mains is 360Hz, with a high constant DC component to the unfiltered output. Some older equipment required a 3-phase 400Hz supply (i.e. get your MG set running). Cheers, Chuck

Jochen Kunz wrote: That's the sort of thing worthy of video. Big power control stuff is fun :) Was that rating for the entire system, or just the CPU & I/O (no drives etc.)?

What would you see? All the rotating stuff in any MG post-1920 is all enclosed. The only thing that might be visible is the flexible coupling between the motor and generator (great to get your tie stuck in). -- Will

William Donzelli wrote: Hmm, I think the latest one I had dealings with was a Dorman out in NZ - that was 1950's (ex-telecoms) and the generator side was mostly exposed (and no covers at all on the motor side of things). The British Polars I did some messing around with were mid-30's; the generators were something like 1200kVA each (and stood about 6ft tall) - could power a fair few mainframes off those! It's mainly the sound of these things running that's so captivating (and the smell) I suppose, rather than the visuals.

Dave McGuire wrote: Did not IBM tech's have to wear one?

On Oct 18, 2007, at 1:17 AM, woodelf wrote: Pretty much all of them did at one time or another. I just have a large philosophical objection to the notion that one's clothing is more important than one's technical abilities. I've turned down many a job due to "dress code" garbage. I work best when I'm comfortable. -Dave -- Dave McGuire Port Charlotte, FL Farewell Ophelia, 9/22/1991 - 7/25/2007

woodelf wrote: That's not geek. That's nerd. 8-) Peace... Sridhar

woodelf wrote: Those are kooks. 8-) Peace... Sridhar

check out e-pay auction 290171794664 s shumaker Bob Rosenbloom wrote:

On Tue, 16 Oct 2007 21:49:44 +0100 Jules Richardson <julesrichardsonuk at yahoo.co.uk> wrote: You can visit it: http://www.cray-cyber.org/ It was just the MG starting. The machine was switched off. Current droped to a few A once the MG got to speed. The MG has a big flywheel. Therefore it needs all that energy to get up to speed. The idea is that the flywheel stores enough energy to compensate short power outages. If power stays off longer the disel kicks in. (But at Cray-Cyber they skiped the disel. ENOSPACE.) AFAIK a minimal configuration of the machine (CPU + I/O) needs about 20 kW. -- tsch??, Jochen Homepage: http://www.unixag-kl.fh-kl.de/~jkunz/

On Wednesday 17 October 2007 12:22, Chuck Guzis wrote: So what kind of problems were caused by the power outages? I assume that these are hardwar faults, not software-related things (like filesystems that were screwed up after the outage, etc). Pat -- Purdue University ITAP/RCAC --- http://www.rcac.purdue.edu/ The Computer Refuge --- http://computer-refuge.org

Jochen Kunz wrote: Rats, seems to be broken at the moment (I get a connection, but no data ever comes back) :-( Ahh, I think we're talking about different things. To me, a motor-generator is a diesel (or other non-electric power source) driving a generator to produce electricity; I've always known something which just converts electric between forms as a rotary converter. (but I'm not sure quite what I'd call a flywheel-storage device) Maybe the terms are ambiguous... or I'm just plain wrong ;-) (I've had a few experiences with generating plant, but not in a computer context...)

That tends to be UK usage too. The thing under the bonnet (OK, hood) of a car is called an 'engine' over here. 'Motor' for that would be very uncommon. To me, a 'motor generator' is just that. An electric motor driving an electric generator, to be used to convert voltage/frequency, etc. A 'rotary converter' (or 'Dynamotor; if oyu go back far enough) is a simialr thing with common field windings/frame/etc rather than 2 separate electrical machines with the shafts coupled. A 'generator' over here is either just the mechanical-to-electircal energy converter or , as you said, a heat engine coupled to such a machine. Certainly if you hired a 'generator' (as somebody suggested we should do for a VCF-type event), you'd expect to get something with the engine included. I don't know how common these other term are across the Pond, but an 'alternator' is an AC-output geneterator, a 'Dynamo' (short for 'Dynamo-electric machine' is a generator, normally DC output, with a wound field, and a 'Magneto' (short for 'Magneto-electric machine' is a genatore with a permanent magnet field. Which means the common bicycle dynamo (as it's normally called over here' is in fact a magneto.... -tony

Back on the subject--a friend runs his 3-phase shop equipment off of a rotary "phase converter" (Rotocon). At first blush, this looks like nothing more than a 3-phase motor with single-phase feeding one (delta-connected) winding with the others being fed through a large capacitor form one side of the mains or the other. The equipment to be powered is connected in parallel with the windings of the Rotocon motor. I take it that this might work for motors, but probably would wreak havoc with a less-reactive load such as a VAX. Cheers, Chuck

--- William Donzelli <wdonzelli at gmail.com> wrote: I found my 11/750 in a dusty corner of a warehouse... They do seem to be pretty reliable, after a lot of cleaning (it was filthy!) it powered up and sorta worked. The RDM board is bad, but that's not required to get the machine to run. It appears to be failure of the RAM on the RDM board, actually. Which of course is soldered to the board, and there is no diagnostic routine to find out which chip is bad... So, yeah, it seems like it's a fairly robust critter. Years of service, then years of storage in an unheated warehouse, and it still works. But then again, this is only personal experience... -Ian

William Donzelli wrote: Sounds like "Multiwire" (as others on the list reminded me a year ago when I was trying to remember the same thing).

On 10/16/07, William Donzelli <wdonzelli at gmail.com> wrote: Multiwire, as was pointed out elsewhere (though there's another technology that we used for our VAXBI boards, called "Unilayer" which is much like Multiwire, except, I think, the difference is that the Unilayer wire nest is laid out on a bit of pressure/heat-sensitive adhesive sheet, then bonded to the substrate as a unit, not one wire at a time. I suppose that's possible, but unless there has been mechanical damage or massive cyclical temperature shifts, I think the wires are somewhat sturdy. I certainly can believe there's a bad SRAM on that board, especially if they are 2114s or similar. But as the OP said, it's not an essential board. -ethan

On 10/16/07, William Donzelli <wdonzelli at gmail.com> wrote: They are remarkably reliable. I think I've seen one or two gate array failures in the 10-15 years that I was around 11/750s on a regular basis (and none lately, but that's a sample size of effectively zero). Even so, that's two out of dozens of problems that were in less esoteric arenas (like bad power supplies and such). I am entirely willing to accept that ratio (and since I have two 11/750s... that means that somewhere in my pile should be 1/5 of a 11/780, right ;-) -ethan

On 10/16/07, Tony Duell <ard at p850ug1.demon.co.uk> wrote: Right... it says "R80" on the front, but in a variety of places (Ultrix device drivers, perhaps?) it's also called the RB80, presumably to distinguish it from the RM80 and RA80 (same HDA, for folks unfamiliar with it, but 3 different host interfaces). AFAIK, yes... the RM80 is an "R80" HDA on an RM02/RM03-like pedestal, with a Massbus cage below. I wouldn't be surprised to learn there are numerous board and packaging similarities )in the Massbus portion) between an RM02/RM03 and an RM80. From experience, I know that the RM02/RM03 HDA is a repackaged CDC 9762 (same guts, different shell), *but* a real 9762 is SMD (and works with Emulex controllers, SI 9900s, etc) and the RM02/RM03 does not, despite identical cabling. I was assigned the task, once, to attempt to convert an RM02 back into a real SMD drive - in effect, making it a 9762 with a DEC faceplate, but it was beyond my skill level at the time. 80% or so of the boards in the HDA card cage were absolutely identical between the two drives, and the obivous differences didn't seem insurmountable, but it just didn't work. Thinking back, one of the things I _didn't_ check was the card cage backplane wiring - perhaps there was a difference there. I think I saw that once or twice in the real world - tape rather than RL02 on top. Yes... I didn't explicitly state it (but implied it via the device names DQA0: and DQA1:) that the RL02 did not require a RL11. As far as we were concerned, it was a feature, since RL02 was our "sneaker-net" medium of choice, and at the time, an RL11 was on the order of $600. That seems newer. Our 11/730 was ordered the day they were announced at a DECUS, if I remember the story correctly. Our guys had a devil of a time with the cable management, and had to escalate their calls a couple of levels inside DEC for our "self-install" of the machine. There was this especially tricky "Z fold" to measure. Get it right, and the cables for the IDC disk (1 x 40 + 1 x 60 + 1 x 26), the DMF-32 (1 x 25 + 3 x 40), and the Unibus (wide, white, mylar flat cable) all fit nicely in the tray and were properly protected when extending the CPU. Get that Z-fold wrong, and it was a thick mess that was nearly twice as thick as would fit. :-P Once the cables were installed, though, it was a nice package. The fact that it was an inexpensive (for the time) Unibus VAX, made it a critical part of our operation for over 8 years. I learned to install Unix (Ultrix v1.1) on that machine in 1985, and later, it was our VMS 5.0 host for relinking our product images against the VMS 5 run-time libraries so we could support 4.x and 5.x simultaneously. I wrote a script to stuff our object files over DECnet (via the DMF-32 DDCMP-compatible sync serial port) and pull the linked files back, so that we could run one script on our build machine and have it crank out both versions of the binaries into our tape-cutting directories. That one little machine, tied to our 11/750, is how we kept our company going for years and years. When we finally did get that 8300, it was solely for VAXBI product development, never for day-to-day use. To that, we trusted our VAX-11 hardware right to the very end day in 1993. -ethan

On 10/18/07, Tony Duell <ard at p850ug1.demon.co.uk> wrote: I remember fiddling with those once or twice, long ago. Interesting. Didn't know they were that flexible. ISTR that was the standard arrangement, just that the integral RL02 routed the cable such that it perfectly reached the I/O panel. One could still route the signal to an external drive if desired (which we did once or twice for reasons I now cannot recall). Good question. Back in the day, I owned RL01 drives (since they were under $200 and RL02s at the time were over $1000), but we never had RL01s where we used VAXen. ISTR there was something about being able to use RL02s but not RL01s with the RL278 interface on a DECmate I, but I also have not had the chance to explore that, either. If you had one handy, it might make an interesting experiment. There's no reason why it should break anything, but the driver/controller might attempt to seek to a non-existent cylinder on an RL01. I was pretty impressed when I saw the guts of an 11/750 for the first time - it was so much smaller than an 11/780. I, too, was impressed when I saw all of that essential functionality in an 11/730. My first MicroVAX was even more amazing, but that's another product line and another topic. -ethan

Tony Duell wrote: Talking about the 11/730's and 11/725's. Anybody on this list has one, he doesn't use ? And probably would like to get rid of ? Cheers, emanuel

How?