On 2012-02-18, Ignoramus23204 wrote:
On 2012-02-18, Michael A. Terrell wrote:

Ignoramus23626 wrote:

If I bought it, I would junk the old control even if it was not yet
broken.

First, electronics that is that old would not be reliable. Second,
this is a very old and obsolete control, compared to the PC based
controls (Mach3 and especially EMC2).


That is your choice, but you can't repair the electronics without
step by step instructions. There are 80 year old pieces of electronics
still in use, and a good tech can keep them working.

Note that a certain age of electronics has problems with
electrolytic capacitors. The formula for the electrolyte was pirated,
but what they stole was a wrong formula, and many capacitors made with
that formula continue to fail. Most seem to be used in computers and
similar devices -- which would certainly impact CNC controllers made
during that period.

It may make sense for other types of devices, but old CNC controls, I
think, are not worth the effort. I am beyond happy with my EMC2 setup.

Considering how much progress there has been in CNC controls,
and "linuxCNC" (the new name for "EMC2", which turns to to be too
similar to some computer security company) has many advantages over the
old ones.

For example, the old controller for the Bridgeport mills (BOSS-3
through BOSS-6 at least) used a DEC LSI-11 CPU card, which is limited to
64K of address space. This includes the firmware which implements the
CNC itself, and the RAM into which G-code programs are loaded.

And memory in older controllers used to be specified in "feet"
(feet of punched paper tape, with each byte taking 1/10", so a foot of
tape is equivalent to only 120 bytes of program. If the whole address
space of the LSI-11 were available for the program, that would only be
546.13 feet of memory. And at least half of that was really taken up by
the firmware. Lots of PROMs.

The memory in the BOSS-3 through BOSS-6 machines is quite
limited, so they had to be set up (for larger programs) to read in as
much as the memory could hold, run that program, then read the next
chuck in and continue. This is a significant limitation if you were
trying to call functions (which had to be in memory the full time) or
were trying to run a loop (which might be longer than the memory present
in the machine). Disk drives were not even a thought. The choices for
input we

1) Punched paper tape (produced on another computer in the main
office of the plant).

2) Hand entered on a teletype or computer terminal (or now,
downloaded from the serial port of a PC, which might be
the very machine which generated the G-code.

2a) The teletype would have a punched tape reader, and punch
available, so that was sort of a third way to read things in.
and to save a program once you had it doing what you want.

Newer machines had miniature cassette tape drives, including the
Anilam Crusader II, and my little Emco-Maier Compact-5/CNC. (The latter
has the 6502 CPU which is at least as limited as the LSI-11, and
probably more so given the less efficient instruction set, so more
firmware was needed to do the same amount of control.) (I do plan to
build a linuxCNC controller for the spare Compact-CNC, and replace the
steppers with servos eventually.)

Enjoy,
DoN.


--
Remove oil spill source from e-mail
Email: | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---