These (edited) memories have been contributed by visitors to this web site.
Prior | Memories |
Next
I loved your website.
I worked as an "FE" at Burroughs from 12/75 till
6/81. There was two Group classifications of Field Engineers at Burroughs--
Group II and Group III. I belonged to Group II and we handled all Group I
(Calculators, mechanical adding machines, Sensimatics and the like) and
Group II which included the "Mini-computers" (L/TC's), B700/800/900 and
B80/90 Burroughs hardware. The Group III engineers did hardware maintenance
on the larger B1700, etc product line.
I have one correction to the L8000 section. You stated: "It was also
equipped with a faster 40 character-per-second dot matrix printer (or
"needle printer, as it was called)." All L8xxx models had the "ball
printer". The 20cps printers were controlled with the use of the "oil
filled decoder" which contained the tilt and rotate mechanism / relays to
drive the "bands" that operated the print ball. Later versions of the L8xxx
had the 30cps printer, which incorporated separate stepping motors for the
tilt and rotate functions. The 30cps L8 models failed regularly due to
driver transistors burning out and mechanical timing slipping. The dot
matrix printer was released with the L9xxx series. It was somewhat superior
to the L8, but had problems with the ribbon dot matrix cable shorting and
the needles being bent due to paper jams, etc.
The B700 was also challenging to support / maintain. The "CPU" was solid,
even the early "core memory" models. Most service calls were to fix the
B9480/81 removable disk drives. The 100 track per inch drives were
generally maintenance free, unless the heads crashed, however the 200tpi
models never really worked in my opinion. These included temperature
compensations circuitry that always needed adjustment and failed regardless
of how they were adjusted. They had a servo positioning motor to position
the "heads" that ran on rails and would become "gouged" over time from dirt
and usage. Once this happened, the only solution was to replace them and
try to align the replacement rail so it could read all the customer's
existing media. Of course, you didn't leave the customer sight until you
verified you could read all their disks after the operation, or you were
back the next day by popular demand.
The B800 line was even better than the B700. It had an "Electronic" eight
channel paper tape reader feeder as opposed to the "Mechanical" reader on
the B700. This saved a ton of time when it came to loading maintenance
paper tape programs to diagnose CPU issues. It used the same "FE Cards"
that were plugged into the CPU that allowed for manual step though the
maintenance programs. Based on a halted led register pattern, a list of
IC's was presented for replacement spread among some 25+ 8 x8 in printed
circuit cards. All "chips" were plugged into removable sockets, so FE's
usually carried a box of 50+ different 14 pin chips that could solve all
their problems :-)
The B900 was the last model in this product line. The biggest difference in
this model from the B800 was the ability to assign different functions to
different CPU's. For example, your could assign Data Comm or Printing to
one CPU set. Additionally, MPLII or COBOL applications could be dedicated
to one or more CPU set(s).
My six year stint at Burroughs was a learning experience to say the least.
It was a time when support involved technical skills. Mechanical skills
were necessary to repair old F and E Series Sensimatics. There were
transistors to be physically replaced in the L/TC line; there was no surface
mount technology. There were voltages to be measured and set accordingly to
operational tolerances. There were disk heads to be aligned using "tri-bit
/ cat-eye" patterns as viewed from your Model 422Tektronix scope. Ah, the
good ole days!
Thanks and keep the Burroughs page up,
Steve Svoboda
Salt Lake City, UT
Prior | Memories |
Next