10 Mar
2023
10 Mar
'23
3:04 p.m.
I always thought of the distinctions this way (from my basis of exposure from late 1970s
through the 1980s) and from a higher educational setting primarily:
Mainframe = repairs required multiple technicians, some possibly there full-time; regular
operator(s) present, and a locked door located between you and the machine; entire
specialized room with raised flooring, extra-high amperage specialized power sources and
wiring, and significant air conditioning
Minicomputer = Vendor still provides a technician (just one) for repairs, who drives in
out in a station wagon; only a part-time operator only; an user can be located in the same
room; 240-volt wiring, but not particularly outlandish
Microcomputer = Computer can sit on a desk or in a "normal" room; broken
computer taken by user to someplace to be repaired or self-repaired; typically one user,
and only 120-volt household or office power needed.
Supercomputer = a really fast and specialized version (primarily focusing on high-speed
mathematical computations) of a mainframe.
Kevin Anderson
10 Mar
10 Mar
3:11 p.m.
New subject: mainframe vs mini
On Fri, 10 Mar 2023, Kevin Anderson via cctalk wrote:
I always thought of the distinctions this way (from my
basis of exposure
from late 1970s through the 1980s) and from a higher educational setting
primarily:
Mainframe = repairs required multiple technicians, some possibly there full-time; regular
operator(s) present, and a locked door located between you and the machine; entire
specialized room with raised flooring, extra-high amperage specialized power sources and
wiring, and significant air conditioning
Minicomputer = Vendor still provides a technician (just one) for repairs, who drives in
out in a station wagon; only a part-time operator only; an user can be located in the same
room; 240-volt wiring, but not particularly outlandish
Microcomputer = Computer can sit on a desk or in a "normal" room; broken
computer taken by user to someplace to be repaired or self-repaired; typically one user,
and only 120-volt household or office power needed.
Supercomputer = a really fast and specialized version (primarily focusing on high-speed
mathematical computations) of a mainframe.
extension cord / dedicated circuit / dedicated pole transformer
Which machines needed 3-phase?
3:20 p.m.
New subject: mainframe vs mini
On 3/10/2023 6:11 PM, Fred Cisin via cctalk wrote:
On Fri, 10 Mar 2023, Kevin Anderson via cctalk wrote:
Some PDP-11's although you can sometimes break them up and just use
lots of 110/220 outlets. And then you get to watch the lights in the house
dim when you fire them up.
Some Vaxen as well and some of those you cannot break up. At least I never
figured out how. But the University electricians always did a good job
of meeting
my power needs.
I would be willing to bet that the 1401, 360's and 4300 systems all
required a bit
more than you get out of a standard 100AMP 220 volt electrical feed.
And Univac 1100's didn't run on commercial power at all but required a motor
generator between them and commercial power.
bill
I always thought of the distinctions this way
(from my basis of
exposure from late 1970s through the 1980s) and from a higher
educational setting primarily:
Mainframe = repairs required multiple technicians, some possibly
there full-time; regular operator(s) present, and a locked door
located between you and the machine; entire specialized room with
raised flooring, extra-high amperage specialized power sources and
wiring, and significant air conditioning
Minicomputer = Vendor still provides a technician (just one) for
repairs, who drives in out in a station wagon; only a part-time
operator only; an user can be located in the same room; 240-volt
wiring, but not particularly outlandish
Microcomputer = Computer can sit on a desk or in a "normal" room;
broken computer taken by user to someplace to be repaired or
self-repaired; typically one user, and only 120-volt household or
office power needed.
Supercomputer = a really fast and specialized version (primarily
focusing on high-speed mathematical computations) of a mainframe.
extension cord / dedicated circuit / dedicated pole transformer
Which machines needed 3-phase? 
5:20 p.m.
New subject: mainframe vs mini
On 3/10/23 17:20, Bill Gunshannon via cctalk wrote:
On 3/10/2023 6:11 PM, Fred Cisin via cctalk wrote:
The VAX 11/780 was wired for 208 V 3-phase, but all the
internal power supplies ran line-line, so were single-phase.
Also, the cooling blowers had 208V single-phase motors.
The DECSystem 2020 (KL10B processor) ran directly off
3-phase power with a HUGE transformer in the bottom of the
cabinet.
On Fri, 10 Mar 2023, Kevin Anderson via cctalk
wrote:
Some PDP-11's although you can sometimes break them up and
just use
lots of 110/220 outlets. And then you get to watch the
lights in the house
dim when you fire them up.
Some Vaxen as well and some of those you cannot break up.
At least I never
figured out how. I always thought of the distinctions this way
(from my
basis of exposure from late 1970s through the 1980s) and
from a higher educational setting primarily:
Mainframe = repairs required multiple technicians, some
possibly there full-time; regular operator(s) present,
and a locked door located between you and the machine;
entire specialized room with raised flooring, extra-high
amperage specialized power sources and wiring, and
significant air conditioning
Minicomputer = Vendor still provides a technician (just
one) for repairs, who drives in out in a station wagon;
only a part-time operator only; an user can be located
in the same room; 240-volt wiring, but not particularly
outlandish
Microcomputer = Computer can sit on a desk or in a
"normal" room; broken computer taken by user to
someplace to be repaired or self-repaired; typically one
user, and only 120-volt household or office power needed.
Supercomputer = a really fast and specialized version
(primarily focusing on high-speed mathematical
computations) of a mainframe.
extension cord / dedicated circuit / dedicated pole
transformer
Which machines needed 3-phase? I would be willing to bet that the 1401, 360's and
4300
systems all required a bit
more than you get out of a standard 100AMP 220 volt
electrical feed.
And Univac 1100's didn't run on commercial power at all
but required a motor
generator between them and commercial power.
The IBM 360/50 and /65 used a "Converter Inverter" that
converted 208 V 3-phase utility power to single-phase 120 V
2500 Hz sine wave power.
All the logic supplies ran off that 2500-Hz power, and were
astonishingly small.
At least the mid-range 370's had built-in motor generators
that converted 208V 3-phase power to 120/200 V 415 Hz
3-phase power. The internal supplies ran off that, vastly
reducing the size of transformers and capacitors.
Jon
3:26 p.m.
New subject: mainframe vs mini
On Fri, 10 Mar 2023, Kevin Anderson via cctalk wrote:
None that I am aware of - even the very power hungry 9021 'mainframes' I mentioned
earlier could be powered from a single phase, if you wanted to. The power company might
have expressed a view, but that's a different subject. 3 phase is relevant to circular
motion, not to the amount of power you need - although lots of electrical equipment tends
to come in 3 phase versions, especially if the amount of power goes up.
And yes, my memory has been refreshed - it was 9021, not 9020 I meant in the other post.
I always thought of the distinctions this way
(from my basis of exposure from late 1970s through the 1980s) and from a higher
educational setting primarily:
Mainframe = repairs required multiple technicians, some possibly there full-time; regular
operator(s) present, and a locked door located between you and the machine; entire
specialized room with raised flooring, extra-high amperage specialized power sources and
wiring, and significant air conditioning
Minicomputer = Vendor still provides a technician (just one) for repairs, who drives in
out in a station wagon; only a part-time operator only; an user can be located in the same
room; 240-volt wiring, but not particularly outlandish
Microcomputer = Computer can sit on a desk or in a "normal" room; broken
computer taken by user to someplace to be repaired or self-repaired; typically one user,
and only 120-volt household or office power needed.
Supercomputer = a really fast and specialized version (primarily focusing on high-speed
mathematical computations) of a mainframe.
extension cord / dedicated circuit / dedicated pole transformer
Which machines needed 3-phase? 
11 Mar
11 Mar
10:40 a.m.
New subject: mainframe vs mini
On Mar 10, 2023, at 6:26 PM, Sytse van Slooten via
cctalk <cctalk(a)classiccmp.org> wrote:
True. I remember that the house where I grew up in Holland had 3 phase service because we
were one of the (then unusual) households to have an electric cookstove. It was fed off
380 volt 3 phase power.
But apart from circular motion (i.e., feeding motors) there is a second reason to use
three phase power: reduced ripple. Full wave rectification of n Hz mains power gives you
a 2n Hz ripple; full wave rectification of 3-phase n Hz mains produces a 6 Hz ripple, and
tha amplitude is a lot less as well. So the filters are much smaller, which is a
non-trivial matter if you need to filter several hundred amps. That of course assumes
your power supplies are actually fed off the three phase mains, rather than being
individual single phase supplies spread over the phases. The former was certainly the
case with the CDC 6600, you can see it clearly in the schematics.
paul
On Fri, 10 Mar 2023, Kevin Anderson via cctalk wrote:
Which machines needed 3-phase?
None that I am aware of - even the very power hungry 9021 'mainframes' I
mentioned earlier could be powered from a single phase, if you wanted to. The power
company might have expressed a view, but that's a different subject. 3 phase is
relevant to circular motion, not to the amount of power you need - although lots of
electrical equipment tends to come in 3 phase versions, especially if the amount of power
goes up. 
2:30 p.m.
New subject: mainframe vs mini
Yes I can concur with this. I operate some Navy transmitters (at home) and while on the
vessel they used 3 phase power supplies which were physically smaller and produced less
ripple. At home, I use a 1 phase power supply just fine. These were 1KW output (4KW input)
transmitters so not a lot of power but benefitted from low ripple.
73 Eugene W2HX
-----Original Message-----
From: Paul Koning via cctalk <cctalk(a)classiccmp.org>
<snip>
But apart from circular motion (i.e., feeding motors) there is a second reason to use
three phase power: reduced ripple. Full wave rectification of n Hz mains power gives you
a 2n Hz ripple; full wave rectification of 3-phase n Hz mains produces a 6 Hz ripple, and
tha amplitude is a lot less as well. So the filters are much smaller, which is a
non-trivial matter if you need to filter several hundred amps. That of course assumes
your power supplies are actually fed off the three phase mains, rather than being
individual single phase supplies spread over the phases. The former was certainly the
case with the CDC 6600, you can see it clearly in the schematics.
paul
2:45 p.m.
New subject: mainframe vs mini
On 3/11/23 14:30, W2HX via cctalk wrote:
Yes I can concur with this. I operate some Navy
transmitters (at home)
and while on the vessel they used 3 phase power supplies
which were
physically smaller and produced less ripple. At home, I use a 1 phase
power supply just fine. These were 1KW output (4KW input) transmitters
so not a lot of power but benefitted from low ripple.
A full-wave 3-phase rectifier's output is about 96% DC (4% ripple).
--Chuck
3:57 p.m.
New subject: mainframe vs mini
OB_3-phase anecdote:
Shortly before I started teaching at the college (so I got all of the
details secondhand), . . .
They had a PDP (I don't even know what model) that they used for the
prograamming classes. But, they had constant problems with the
[after-market?] disk drive, so it was down a lot. Not good when students
need it for their homework. One semester, it was down almost all
semester.
So, they decided to replace it with microcomputers. They sold it to a
local school district, and bought a room full of 5150s. Plus an
unreliable network ("can support up to 128 simultaaneous users"; "oh,
we've never connected more than 8 at a time", and a single user could
bring the network to its knees).
The school district was thrilled to get the PDP. Then they had PG&E set
up the power for it. Some PG&E technicians did not know the difference
between "Delta" and "Wye"/"Y" three phase! Seriously
dsmaged the
machine.
PG&E bought a replacement machine for the school district, on the
condition that all involved go along with a false story that it had been a
lighning strike!
I started teaching at the college. Fortran, "Micorcomputer Disk Operating
Systems", and later BASIC, PC Assembly language, C, and even COBOL.
We had a room full of 5150s, and used the IBM Microsoft Fortran and COBOL.
There are some problems with the first version of the IBM/Microsoft PC
Fortran, but it was OK for teaching beginning programming.
--
Grumpy Ol' Fred cisin(a)xenosoft.com
4:30 p.m.
New subject: mainframe vs mini
On 3/10/23 15:57, Fred Cisin via cctalk wrote:
OB_3-phase anecdote:
PG&E bought a replacement machine for the school
district, on the
condition that all involved go along with a false story that it had been
a lighning strike!
Doesn't seem to be an uncommon practice. I remember an electrician
hooking up a 120V60Hz water cooler to the 250VDC supply used for
welding. Sent the unit back for replacement because of "manufacturing
defects".
--Chuck
4:38 p.m.
New subject: mainframe vs mini
> OB_3-phase anecdote:
> PG&E bought a replacement machine for the school district, on the
> condition that all involved go along with a false story that it had been
> a lighning strike!
On Fri, 10 Mar 2023, Chuck Guzis via cctalk wrote:
Doesn't seem to be an uncommon practice. I
remember an electrician
hooking up a 120V60Hz water cooler to the 250VDC supply used for
welding. Sent the unit back for replacement because of "manufacturing
defects".
It wasn't even my first experience of it.
When I built my auto repair garage (building built for me) in 1975, got
three phase for the air compressor. Until we did our own testing, there
was over 200 volts to the wall outlets.
--
Grumpy Ol' Fred cisin(a)xenosoft.com
5:07 p.m.
New subject: mainframe vs mini
On Fri, Mar 10, 2023 at 3:58 PM Fred Cisin via cctalk <cctalk(a)classiccmp.org>
wrote:
OB_3-phase anecdote:
That internet convention (OB) is so old now it's archaic, and vintage in
and of itself. Amazing.
So, they decided to replace it with microcomputers.
They sold it to a
local school district, and bought a room full of 5150s. Plus an
unreliable network ("can support up to 128 simultaaneous users"; "oh,
we've never connected more than 8 at a time", and a single user could
bring the network to its knees).
Typical.
The school district was thrilled to get the PDP. Then
they had PG&E set
up the power for it. Some PG&E technicians did not know the difference
between "Delta" and "Wye"/"Y" three phase! Seriously
dsmaged the
machine.
Typical.
PG&E bought a replacement machine for the school
district, on the
condition that all involved go along with a false story that it had been a
lighning strike!
Ha! PG&E has always been involved in shady shit, but I never suspected
insurance fraud.
Sellam
5:20 p.m.
New subject: mainframe vs mini
> OB_3-phase anecdote:
On Fri, 10 Mar 2023, Sellam Abraham via cctalk wrote:
That internet convention (OB) is so old now it's
archaic, and vintage in
and of itself. Amazing.
You rang? (cf. Maynard G. Krebs)
> The school district was thrilled to get the PDP.
Then they had PG&E set
> up the power for it. Some PG&E technicians did not know the difference
> between "Delta" and "Wye"/"Y" three phase! Seriously
dsmaged the
> machine.
Typical.
PG&E
bought a replacement machine for the school district, on the
condition that all involved go along with a false story that it had been a
lighning strike!
Ha! PG&E has always been involved in shady shit, but I never
suspected
insurance fraud. 
7 p.m.
New subject: mainframe vs mini
that's still happening today. Up here in the north woods, trees in
lines is a weekly thing and the PGE supers will tell you directly to
"make sure you get EVERYTHING that we broke" but not to worry about
receipts.
steve
On 3/10/23 5:07 PM, Sellam Abraham via cctalk wrote:
On Fri, Mar 10, 2023 at 3:58 PM Fred Cisin via cctalk
<cctalk(a)classiccmp.org>
wrote:
OB_3-phase anecdote:
That internet convention (OB) is so old now it's archaic, and vintage in
and of itself. Amazing.
So, they decided to replace it with
microcomputers. They sold it to a
local school district, and bought a room full of 5150s. Plus an
unreliable network ("can support up to 128 simultaaneous users"; "oh,
we've never connected more than 8 at a time", and a single user could
bring the network to its knees).
Typical.
The school district was thrilled to get the PDP.
Then they had PG&E set
up the power for it. Some PG&E technicians did not know the difference
between "Delta" and "Wye"/"Y" three phase! Seriously
dsmaged the
machine.
Typical.
PG&E bought a replacement machine for the
school district, on the
condition that all involved go along with a false story that it had been a
lighning strike!
Ha! PG&E has always been involved in shady shit, but I never suspected
insurance fraud.
Sellam
9:01 p.m.
New subject: mainframe vs mini
On 3/10/23 19:07, Sellam Abraham via cctalk wrote:
> The school district was thrilled to get the PDP.
Then they had PG&E set
> up the power for it. Some PG&E technicians did not know the difference
> between "Delta" and "Wye"/"Y" three phase! Seriously
dsmaged the
> machine.
>
This goes a bit sideways here. I've dealt with a lot of power people,
shady company or not, and the equipment they have isn't for amateurs
that don't know the type of service or whatever. Just doesn't fly.
And as far as any dealings I know about, all of them put the order of
the service responsibility on the customer. The power company will put
in what you order. They could put in the wrong service, but again
referring to the previous paragraph, there are many more issues with the
equipment and people involved for someone to have not caught it. I'd
put the blame on the company at 5% as far as the story.
As to the business issues, can't say. Again, getting far into grassy
knowl territory for someone not to talk, and there are a lot of PG&E
people who would dearly love to drop that company into boiling oil as
there are customers.
No way to verify on my end, just observing the state of the story at
this point from what I read here.
(Not Sellam's story, clipped from his level of the tread).
thanks
Jim
5:09 p.m.
New subject: mainframe vs mini
On 3/10/2023 6:32 PM, Chuck Guzis via cctalk wrote:
On 3/10/23 15:11, Fred Cisin via cctalk wrote:
Univac 1100
IBM 360/40 mounted in two 65 foot trailers and known as The Army
Standard System
bill
extension cord / dedicated circuit / dedicated
pole transformer
Which machines needed 3-phase?
Extra points for requiring an MG set furnishing
400Hz 3-phase.
5:42 p.m.
New subject: mainframe vs mini
On Mar 10, 2023, at 6:32 PM, Chuck Guzis via cctalk
<cctalk(a)classiccmp.org> wrote:
On 3/10/23 15:11, Fred Cisin via cctalk wrote:
CDC 6000 series is the one I know of. But they'd also use either 1 or 3 phase regular
50/60 Hz power. For example, the console display uses 3 phase 400 Hz for its power
supplies (nice for low ripple) and 60 Hz 110 V for fans. The CPU cabinet takes 400 Hz
power to run its power supplies but 60 Hz power for the compressors that cool the Freon
that cools the logic.
As for 3 phase, PDP-11 CPUs did not, but some of the disk drives (like RP06) do since they
use 3 phase induction motors to drive the spindle.
paul
extension cord / dedicated circuit / dedicated
pole transformer
Which machines needed 3-phase?
Extra points for requiring an MG set furnishing 400Hz 3-phase.
--Chuck
7:36 p.m.
New subject: mainframe vs mini
On 3/10/23 17:42, Paul Koning via cctalk wrote:
CDC 6000 series is the one I know of. But they'd
also use either 1 or 3 phase regular 50/60 Hz power. For example, the console display
uses 3 phase 400 Hz for its power supplies (nice for low ripple) and 60 Hz 110 V for fans.
The CPU cabinet takes 400 Hz power to run its power supplies but 60 Hz power for the
compressors that cool the Freon that cools the logic.
"Power for the basic computer consists of one 250 kva, 400 Hz
motor-generator set. The motor-generator set has the capability of
providing power to the CPU, MCS, I/O and the MCU. The optional memory
requires the use of an additional 80 kva motor-generator set."
Cooling?
"Cooling for the basic computer consist of two 30-ton water-cooled
condensing units. These units cool only the CPU, MCS and I/O sections.
The MCU is air-cooled. With the optional memory, the basic computer
requires an additional 10-ton condensing unit."
--Chuck
9:40 p.m.
New subject: mainframe vs mini
On 3/10/23 17:42, Paul Koning via cctalk wrote:
CDC 6000 series is the one I know of. But they'd
also use either 1 or 3 phase regular 50/60 Hz power. For example, the console display
uses 3 phase 400 Hz for its power supplies (nice for low ripple) and 60 Hz 110 V for fans.
The CPU cabinet takes 400 Hz power to run its power supplies but 60 Hz power for the
compressors that cool the Freon that cools the logic.
I believe some of the tape drives used 208V 60Hz single-phase. I'm
pretty sure the big disk drives (e.g. 808) had 3-phase spindle motors
powered from the 240V feed. I'd have to check, however. 400Hz for the
logic. I don't recall what the Bryant 4000 drives used, but I"m sure
that bitsavers can tell you.
--Chuck
412
days inactive
413
days old
19 comments
11 participants
participants (11)
-
Bill Gunshannon -
Chuck Guzis -
Fred Cisin -
jim stephens -
Jon Elson -
Kevin Anderson -
Paul Koning -
Sellam Abraham -
steve shumaker -
Sytse van Slooten -
W2HX