On Thu, 9 Dec 2004, Fred Cisin wrote: Shag, berber, or ...?

At 04:31 PM 12/9/2004, you wrote: I'm sorry, all answers from Exposition Services require a $100 non-refundable down-payment in advance. It's $200 if you want the answer carried from the loading dock to your booth. - John

Fred Cisin wrote: Blown, or did the fuse brake from being slammed against the floor? ;=)

John wrote: AFAIK, there's nothing particularly magic about 3.3V. CMOS can be designed or optimized for a wide range of voltages, and 3.3V just happened to "win" as one of the most common. But there are also systems that run on 3.0 and other voltages. And most modern VLSI chips actually use lower voltages for the core logic than the I/O, so it's common to see chips with a 1.2, 1.5, 1.8, 2.0, or 2.5V core supply, while 3.3V is still common for the I/O supply. Some chips with 3.3V I/O are "5V tolerant", but many are not. Anyhow, the migration from 5V to 3.3V was mainly motivated by two things: 1) reducing power consumption 2) reducing field strength in high-density IC core circuitry to avoid damage Eric

Not likely in the 6000, but large machines sometimes have a supply of lesser voltage for use in the bus terminations (like the -2.0 volts found in both 10K and 100K ECL machines). William Donzelli aw288(a)osfn.org

Did anyone save any VAX 9000s? I saw one at a junkyard several years back. William Donzelli aw288(a)osfn.org

On Thu, 2004-12-09 at 22:55 -0800, Tom Jennings wrote: Out of interest, why do we use 50Hz power in the UK and 60Hz in the US? (I don't mean why are they different, I mean why those numbers) I saw yesterday a claim that the 60Hz in the US was a Tesla invention, but the info didn't elaborate on why he chose that figure if so... Wouldn't surprise me if there was a connection to the telegraph system, of course. cheers Jules

At 09:24 AM 12/10/04 -0700, you wrote: I think you have it backward. The motor speeds are the results of the line frequency not vice versa. Telsa did a LOT more than invent a better electric motor. You should go read up on him and his work sometime. It's fasinating. FWIW the "better" electric motor was a LOT more than just better, it was revolutionary. Joe

Hi, ben franchuk said: It also depends on the number of poles your motor/generator has. Other frequencies were tried in the early days - 100Hz for example. (The Anderson radio transmitter at Grimeton, Sweden generates 17.2KHz. It's a pruely mechanical transmitter using a big generator.) -- Cheers, Stan Barr stanb(a)dial.pipex.com The future was never like this!

Hi, John Lawson said: Of course...brain not working well lately. See http://www.wireless.org.uk/mechrx.htm (if it's still there) for a matchimg *mechanical* receiver. I was going to have a go at making one myself... -- Cheers, Stan Barr stanb(a)dial.pipex.com The future was never like this!

Tom Jennings wrote: No a good hack is *mechanical* Computer. Now somebody in England had develeoped a nice mechanical display device using rotating mirrors before WWII, and a method of displaying one TV scan-line at time that was projected for display. That was a neat design instead of CRT. Ben.

It was chosen to be very close to the NTSC vertical sync rate of 59.94 Hz. :-)

On Fri, 10 Dec 2004, Jules Richardson wrote: 35 years ago, I was working for a British physicist, and I asked him: "Why are US devices 60Hz, and UK devices 50Hz?" "So that they will work with the local mains power." "Well, why are the US outlets 60Hz, and the UK outlets are 50 Hz?" "So that the devices plugged into them will work."

On Fri, 10 Dec 2004, Jules Richardson wrote: I don't know, but 60Hz is awfully convenient for calculating time (with synch motors). I guess you 50Hz people have to use square gears or something to do 50/60ths :-)

On Monday 13 December 2004 09:29, Paul Koning wrote: Probably 115/120V was chosen, because it was the correct voltage to make a pickle glow. ;) Pat -- Purdue University ITAP/RCS --- http://www.itap.purdue.edu/rcs/ The Computer Refuge --- http://computer-refuge.org

This is fascinating. I had no idea so many different plugs were in such wide use, especially all the ways grounding has been added to the basic European two-prong plug. Unfortunately it's got a couple of bugs. It seems a bit confused about transformers versus converters; and the "Trick to know the local voltage and frequency" doesn't work in, at least, North America (NA light bulbs are marked with voltage but not frequency). The descriptions of converters and transformers make the transformer sound like the more complex device, while in fact almost nothing could be simpler than a transformer, and the description of a converter sounds as though the generated wave is a sinewave with peaks clipped - like what you get from an amp driven into clipping - which works only to lower voltage, delivers too much power into RMS devices like heaters (or else delivers too low a peak voltage to anything else), and would be difficult to build. I've sent off a note to the address given on the page. We'll see what happens. /~\ The ASCII der Mouse \ / Ribbon Campaign X Against HTML mouse(a)rodents.montreal.qc.ca / \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B

Paul Koning wrote: Almost random, I concur, but there were definite technical factors behind. It took a while for people to settle on 60Hz or 50Hz. Some interesting tidbits mostly taken from "Engineers and Electrons" by Ryder & Fink, IEEE Press: First there was the AC vs. DC debate. Edison firmly believed that DC was better, while Tesla-Westinghouse pushed things the other way. Edison even tried (though friends) to push legislation that would outlaw voltages greater than 300V, thus nullifying AC's advantage for transformation to higher voltages (and therefore transmission over longer distances). In 1888, a Harold P. Brown, now widely believed to have been funded and helped by Edison, published letters stating that "AC can be described by no adjective less forcible than damnable". In Ohio, Brown even electrocuted several animals in front of witnesses using AC in order to prove its perilousness. Edison's interest was not safety: at the time, he held the patent to the only available power usage meter, but it only worked with DC. Also, some of the competing light companies used arc lamps, which require high voltages. Finally, Westinghouse held Tesla's patent for the induction motor. Thus, when bids were called from Westinghouse and General Electric to construct the generators for the Niagara Falls project, Westinghouse won thanks to its AC experience. What really tilted the balance in favor of AC was the 1891 demonstration of AC 3-phase transmission at 30,000 volts from Lauffen to Frankfurt, separated by 100 miles (Brown of "Brown-Boveri fame-another Brown- was behind this). DC continued to be used only in transportation systems thanks to the higher starting torque of the DC motors. As for frequency, the factors that usually affected the choice were suitability for use in different devices and the preference of mechanical engineers to design steam engines operating at integral RPM specs. At roughly 25 Hz, flicker became "tolerable" for Edison's cotton filament bulbs, but not so for arc lamps, which required higher frequencies. However, higher frequencies made mechanical design more difficult for AC motors. For example: for the Niagara Falls project, both 16 2/3 and 33 1/3 Hz were proposed. 25Hz was ultimately decided on. This was because the turbine speeds were set at 250 RPM and you need an integral number of poles. A 1888 Westinghouse alternator ran at 2000 RPM and had 8 poles for 133 1/3 Hz. Transformers, filament and arc lighting worked fine at that frequency, but it was too high for AC motors. Around 1890 a lower RPM figure was used when steam engines first shared a single shaft with alternators, and L.B. Stillwell proposed 3600 cycles per minute or 60Hz, which was on the high end for AC motors but good for all lighting and also for efficient and not-too-bulky transformers. That way you could have 1800 rpm steam turbines or 225 rpm steam engines. Carlos. -- Carlos E. Murillo-Sanchez Dean of Engineering Universidad Autonoma de Manizales, Manizales, Colombia ----

On Mon, 13 Dec 2004, Paul Koning wrote: First, I don't mean to imply 'therefore 60Hz is superior', though we can have fun making jokes about it :-) 60 is convenient (not necessarily superior) in calcing time re: navigationally. It divides nicely into 360 (degrees in a circle) and all the related measures thereof. Of course with actual machinery such relations are merely pretty, especially on hallucinogens (really). I'm on another mailing list where one person's .sig contains URLs to rants about the evils of the metric system, and 'the govamint' taking away mercurochrome, plus a few other tinfoil-hat subjects. [I found the sig: -- Non-political causes I support- We want mercurochrome - http://www.mercurochrome.org Self explanatory - http://www.metricsucks.org The former being *particulary* good. ] (The English system is handy in that, it seems to be based upon halving and doubling, and therefore is good with profoundly low-tech implementations (eg. stick and string or chain). If this were the 17th century you could make some argument about it, but certainly not in the last 100. Not that I would, anyways, unless it led to hilarity or heightened beer-consumption. Though the French I believe played with decimal time -- we never heard of THAT again (though I did once have a 24hr clock with 100 minutes, presumably for timecard-type purposes (this is far too many parentheticals))). I think that is the correct answer, if such a thing is even possible. At least here in the sensible north hemisphere (where "up" really means "up" and the water goes down the toilet the proper direction. Maybe you saw the episode of The Simpsons where the U.S. Embassy has a rather complex toilet with a mechanism that FORCES the water to go in the proper direction. American technology at work -- screw you guys!)

Tim wrote: And civilian aircraft as well. When I flew from San Francisco to London in 2002, I had to make arrangements to use my CPAP machine. They were concerned because the machine is rated for 50 to 60 Hz. But it uses a switching power supply which is actually perfectly happy to run on 400 Hz, so I was able to use it. It's irritating that the manufacturer defines the rating based on what power they expect the user to have, rather than practical operational limits. Unfortuately, the mechanic assigned to wire the outlet for me (by a long extension cord and an adapter from the aviation power plug to a standard NEMA 5-15R receptacle) apparently plugged it into the "ground service bus" which is intended for non-flight use for vacuum cleaners and such, and is automatically turned off on takeoff. So I had to stay awake for the whole flight. On the return I asked them to make sure it was NOT wired to the ground service bus, and it worked fine. Speaking of annoying specifications, I was just reading the manual for an HP logic analyzer circa the mid-1990s. They make a distinction between "specifications" and "characteristics", defining the former to be something they actually test the instrument against, and the latter to be non-tested and non-guaranteed. They list the maximum clock fequency and minimum clock pulse width as specifications. But they list the number of channels and the memory depth as "characteristics", which means that if the analyzer only had 67 channels instead of the "characteristic" 68, they could still claim that it met the specifications. In practice I'm sure they wouldn't try to screw customers on that point, and couldn't get away with it anyhow because the channel count and memory depth are prominently featured in the sales literature. But after they go to all the trouble of explaining a distinction between specifications and characteristics, it seems really bizarre that they would put the channel count in the latter category since it is clear that they do actually have to test them all. Eric

On Mon, 2004-12-13 at 18:03 -0500, shoppa_classiccmp(a)trailing-edge.com wrote: Twisting this back on topic, the IBM 650 had a power distribution that made something like 133Hz AC or even higher. -- Tore S Bekkedal <toresbe(a)ifi.uio.no>

On Tue, 14 Dec 2004 19:42:16 +0100 Tore S Bekkedal <toresbe(a)ifi.uio.no> wrote: Reminds me. The CDC Cyber 960 from http://www.cray-cyber.org/ runs at 400 Hz. The higher frequency allows smaller transformers. As this is three phase you can get DC quite easy and efficient with a six pulse rectifier. I was toled that the PSUs have no regulation. The machine is build from ECL. ECL has a constant current consumtion, in oposite to TTL or CMOS. So all you need for a PSU is a six pulse rectifier and a smoothing capacitor. The voltage drop caused by the PSU load is compensated by a slightly higher PSU output voltage. Simple, efficient and elegant. The 400 Hz three phase current is generated by a motor-generator. The rotating mass of this machine is enough to keep the Cyber running for some seconds. (Power consumption of the Cyber at minimal configuration is around 20 kW.) Enough to start a big diesel generator in case of a power outage. (They don't have a diesel at the cray-cyber basement, but it was usual at "regular" installations.) -- tsch??, Jochen Homepage: http://www.unixag-kl.fh-kl.de/~jkunz/

Tom Jennings wrote: Actually they do. APC makes UPSes up to the multiple megawatt range. There's an IBM 1401 restoration project at the Computer History Museum. They have a German 1401, so they need to convert the line frequency from 60 Hz to 50 Hz to run it. I was surprised to learn that rotary converters (i.e., motor-generators) have much higher MTBF that static inverters (all-electronic). Eric

On Tue, 14 Dec 2004, Eric Smith wrote: That's a lot of beige plastic.

On Tue, 14 Dec 2004 18:41:46 -0500 (EST) William Donzelli <aw288(a)osfn.org> wrote: With smaller transformsers you can squeeze more power in the same volume. The "energy density" of the PSU is higher / the PSU gets smaller and the whole machine gets smaller. The size is important as the runtime of the signals in the machine limmits its max. operating frequency. Ripple, of course, is also important. Reminds me of an other Cray thing: Cray said that computer design is like plumbing: Plumbing of data paths, plumbing of power supply and plumbing of cooling. :-) I saw disels of that size starting in less then 10 seconds. Note that we are talking about 100 kW or the like. 100 kW is not a big diesel. Many cars have biger diesel engines. A ship diesel engine with multipe MW is big. An other important fact. -- tsch??, Jochen Homepage: http://www.unixag-kl.fh-kl.de/~jkunz/

On Tue, 2004-12-14 at 18:41 -0500, William Donzelli wrote: We had an electric start Dorman generator at the fortress in NZ (37KW IIRC) and it'd turn over and start happily in a second or so. I've got some nice footage somewhere of a power fail test - kill the supply to the fortress and the lights dim for a couple of seconds whilst the Dorman kicks in automatically and then everything picks up again. The pair of British Polar straight-8 diesels up at Palmerston Nrth were another matter. Each engine's 725 litres in capacity (yep, you read that right) and have to be barred over prior to starting by hand (it's back- breaking work). Each engine has a couple of enormous air recievers, which also need to be charged first and then fire the engine on two cylinders for starting. The engines run at 300rpm, which gives a wonderful ground-shaking rumble and thumping noise from the exhaust stacks. Output's 3300V three-phase - that'd supply any computer collection nicely ;-) Not bad for something of 1930's vintage. Designed for continuous running obviously though, not exactly the sort of thing you'd want to be stopping and starting all the time :) (takes a couple of hours start to finish to run through the startup procedure) cheers Jules

On Wed, 15 Dec 2004 13:56:43 -0500, Paul Koning <pkoning(a)equallogic.com> wrote: Good idea, but isn't that a bit of overkill for your home collection. Dan