19 Jan
2009
19 Jan
'09
1:20 p.m.
Patrick Finnegan <pat at computer-refuge.org> wrote:
What you need to understand here, is that there is nothing in a 11/780
that is actually driven by 3-phase power. And the power controller is
nothing like what you are trying to describe.
The 11/780, along with very much of all DEC equipment that used 3-phase
power, did it just to lessen the load per fuse, along with getting
enough power to the system.
If you look inside, you'll see that the 3-phase goes into a power
distribution box, and from there you have a number of single-phase
sockets. And they are grouped in three separate groups. Each group is
driven by one phase of the 3-phase power, and the other pin is ground on
all sockets.
And with this kind of construction, it really isn't any problem at all
in using the same phase for all three phases. There are very few things
DEC did which really required 3-phase power.
Off my head, I know that the 86x0 machines really require 3-phase (the
fans are AC motors), the RP06 drives use AC motors (probably RP04 and
RP05 as well). I'm not sure, but it might be that KL-10 machines
actually used 3-phase. It's been a while since I looked inside one.
But as always, if you try do run a machine with 3-phase power off a
single phase, you need to know how that machine is built, and if it
actually is possible do "rewire" it. Don't play around with this if you
don't really know, because it can be lethal both to the machine and you.
Johnny
On Sunday 18 January 2009, Bob Armstrong wrote:
I'm not going to argue with you on this, however, see below...
People need to understand the danger in this -
in any multiphase
power system, the neutral current is the DIFFERENCE of the current in
the individual phases. If the load is perfectly balanced, then the
neutral current will be zero.
But, if you run the same load from three separate single phase
circuits and (worst case) the three circuits are all the same phase,
then the neutral current will be the SUM of all three phase currents.
That's 3 times what it would be otherwise and runs a serious danger
of melting the neutral and starting a fire.
This is exactly the point I was trying to get across. Actually I
thought the DEC power controllers had a circuit breaker
in the neutral leg too just to prevent somebody from doing something
like this. The split phase models do - are the three phase power
controllers different ?
I *think* that it does, but without opening one up or digging through
bitsavers, I can't remember for sure. I do seem to remember though,
that the breaker has the 3 phases, neutral, and a shunt-trip run
through it for thermal overload/lack of airflow.
Anyways, in my experience on my 11/780 (which isn't a full config), the
CPU, memory box, unibus controller, fans, 11/03, and RX01 together draw
no more than a total of 24A at 120V, so it may be possible to run the
system off of a single 120V supply, but I wouldn't recommend it.
Adding anything like a second memory box, DU780, an RH780 or something
else to the CPU (does the LSTTL in the 785 draw more or less power than
the 780's cpu at their respective clock speeds?)