Maker Pro
Maker Pro

OT? The latest bare-bones from Fry's, Windows-free

  • Thread starter Rich Grise, but drunk
  • Start date
K

Keith Williams

There were PC's with 186's? The only 186 I've worked with was the embedded
version of the 8086 - it included real-time clocks, address decoders, I
think an interrupt controller - a bunch of peripherals on-chip - I'm
pretty sure that it was intended to be the embedded version of the 8086.

The 80186/ was *intended* as an integrated 8086/8. It ended up as
an embedded processor because IBM stomped all over Intel's
"reserved" interrupts and addresses that were intended for the
8018x. The result was that vanilla DOS could not run on the
80186/8, so were depreciated into to the embedded market.

BTW, at least at one point the 80188 and the 80186 were the same
die with a bond-out option at final packaging. Kinda neat from the
supply chain standpoint.
 
R

Rich Grise

IIRC there were a few others, though nothing with a '186 was 100% IBMPC
compatible (BIOS IRQs were all messed up).

FWIW, the '186 was designed for embedded apps; it had address decoders,
counter/timers, an interrupt controller, and some other stuff that I don't
recall at the moment. I had a job maintaining the firmware for one once.
It was actually kind of fun - I have a "PROM simulator" story, too. ;-)
And an 8048-based FIFO story. ;-)

Cheers!
Rich
 
K

Keith Williams

FWIW, the '186 was designed for embedded apps;

Not according to Intel, at the time. It was IBM that screwed up
the PC so the 186/8 couldn't be used for anything other than
embedded applications.
it had address decoders,
counter/timers, an interrupt controller, and some other stuff that I don't
recall at the moment.

Sure, but that's not the point. PCs had all that stuff too. The
integrated controllers would have brought the cost of the PC bag-o-
parts down significantly. If only...
I had a job maintaining the firmware for one once.

I was in a procurement engineering group at the time. We had
Intel's ear rather close.
It was actually kind of fun - I have a "PROM simulator" story, too. ;-)
And an 8048-based FIFO story. ;-)

Go fer it!
 
J

Joseph2k

Pooh said:
My first HD had only 32 of them ! And that was simply an RLL version of
20MB drives.

Graham
That's right. In those really old drives the signal sent back to the
controller card was just the read head selected amplified and clipped.
They did not always have "servo" written on the disk either. But that was
primarily on the earliest and smallest disks 5 MB, 10 MB, and 20 MB.
Stepper motor head positioning, 85 mS average seek time, 210 mS Max. How
many remember drum memory?
 
S

Sjouke Burry

Rich said:
FWIW, the '186 was designed for embedded apps; it had address decoders,
counter/timers, an interrupt controller, and some other stuff that I don't
recall at the moment. I had a job maintaining the firmware for one once.
It was actually kind of fun - I have a "PROM simulator" story, too. ;-)
And an 8048-based FIFO story. ;-)

Cheers!
Rich
Those 186 computers were sold in the Netherlands
for less than a year, number six we bought a short
while later was a 286.
We had a big row with the supplier,because with
an expensive card you had 1M memory, however
we found out that you could only use 640k,
and got our money back for that extension card,
500+ dollar each card,for 360k EMS memory only,
which we didnt know how to use -)-)-)
 
M

Michael A. Terrell

Joseph2k said:
That's right. In those really old drives the signal sent back to the
controller card was just the read head selected amplified and clipped.
They did not always have "servo" written on the disk either. But that was
primarily on the earliest and smallest disks 5 MB, 10 MB, and 20 MB.
Stepper motor head positioning, 85 mS average seek time, 210 mS Max. How
many remember drum memory?


Drum? How about mag core?

--
Service to my country? Been there, Done that, and I've got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida
 
R

Rich Grise

That's right. In those really old drives the signal sent back to the
controller card was just the read head selected amplified and clipped.
They did not always have "servo" written on the disk either. But that was
primarily on the earliest and smallest disks 5 MB, 10 MB, and 20 MB.
Stepper motor head positioning, 85 mS average seek time, 210 mS Max. How
many remember drum memory?

Me! Me! Me! ;-)

The first computer I ever played with was a Bendix/Control Data G-15, in
high school, in 1965/66. Its main memory was a rotating drum, and its
accumulator(s) and index registers were on the same drum.
http://members.iinet.net.au/~dgreen/
There's a diagram of the drum memory on page 6 of this:
http://acms.synonet.com/bendix/tech/TechMan/tekrpt1.pdf
And page 8 shows the "short lines" where the "A register" and scratchpad
and index registers were. :) The machine instructions had a field for the
address of the next instruction - when you wrote code, you had to know how
long it would take for your instruction to execute so that you could
put the next one in the spot that'd be swinging around on the drum just as
the previous instruction finished executing. It was all serial, 29 bits.
It was way kewl! ;-P

(If you look at the schematics, those wierd things that look like a gamma
and an upside-down tee separated by a dashed line are "toobz". ;-) )

Cheers!
Rich
 
R

RST Engineering \(jw\)

How many remember GLASS memory, where the serial signal bounced around
inside of a square flat box off of mirrors in the corners?

Jim



" How
 
K

Keith Williams

Drum? How about mag core?

Sure. In the '70s I bought (boss paid for ;-) a PDP11/35 as part
of a Tektronix Signal Processing System. It had two floppies and
56K of core. Fortunately, it had a bootstrap ROM. ;-)

There was a "drum" (really a fixed-head disk) on the mainframe
controlling system test (a 370/155, IIRC), with four heads per
track, used as paging store.
 
K

Keith Williams

How many remember GLASS memory, where the serial signal bounced around
inside of a square flat box off of mirrors in the corners?

Haven't seen anything like that, but I have seen one with coax
memory.
 
R

Richard Henry

Drum? How about mag core?

In my Navy days, I worked on a Collins UHF radio control box that had a
small mag-core memory to store channel vs. frequency settings. The next
older control box had a mechanical memory: a small metal drum with rows of
pegs that could be moved to represent various frequency codes, picked up by
microswitches when the drum was rotated into the corresponding position. In
fact, I saved one of those little drums from the trash, discarded becasue
the painted channel numbers had worn off.
 
R

Rich Grise

Haven't seen anything like that, but I have seen one with coax memory.

I've read about something that was like an acoustical memory - there was a
pool of mercury, in a 7-sided container, with ultrasonic transducers, that
sent a wave that bounced off 5 sides and was read out by the other
transducer, but apparently that wasn't a big seller. Or the storage CRT. :)

Cheers!
Rich
 
M

Michael A. Terrell

Richard said:
In my Navy days, I worked on a Collins UHF radio control box that had a
small mag-core memory to store channel vs. frequency settings. The next
older control box had a mechanical memory: a small metal drum with rows of
pegs that could be moved to represent various frequency codes, picked up by
microswitches when the drum was rotated into the corresponding position. In
fact, I saved one of those little drums from the trash, discarded becasue
the painted channel numbers had worn off.

The old "Music box" memory?


--
Service to my country? Been there, Done that, and I've got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida
 
R

Rich Grise

Go fer it!

Ok, I'm a little bored, and al ittle high, so here's the PROM simulator
story. I was working on a device that had a(n?) 80186, which controlled
the video on a 9" monitor, and the company's claim to fame was the "bend
board" - we'd write a pixel pattern of a waveform, but then there was a
special little circuit that would introduce a little magnetic deflection
between pixels, to get rid of the pixellations and produce a trace that
looked like a long-persistence analog EKG, like you see on TV. This was
a 4-channel "remote heart-rate monitor", used in hospitals, where the
patients were ambulatory, but still needed to be monitored. They'd carry
around a little transmitter, about the size of two packs of cigarettes,
with 5 or so leads plugged into it with EKG stickers stuck all on their
chest, in the prescribed spots. It was a little transmitter, of the EKG
signal, using some really cheap technology like five V/F bands in the
audio range. Anyway, the unit I worked on had four receivers, for
four channels, so they could monitor four patients. The software
(firmware?) took the four analog signals which had been ADC'd, and did
some digital filtering on them, and showed the patient's EKG on this
monitor.

So, anyway, I was hired as "programmer", I think, although I think the
fact that I'd been a tech might have played in there. What my coworker
and I did was maintain and improve the firmware for this 80186 that ran
this display device - four parallel digital filters, display drivers, bend
board drivers, a strip chart recorder - but that's the "8048 FIFO" story -
and we'd make a mod to the program, burn a pair of EPROMs, slap them in
the test system and test it. Everybody's heard that one, haven't they!
Edit, compile, burn, plug, crash. Edit, compile, burn, plug, crash. Edit,
compile, burn, plug, run! Hooray! ;-)

So, we ascertained that it would be cheaper in the long run to go ahead
and make a ROM simulator, where we could bypass the erase/burn cycle, and
it's just be edit, compile, load, run/crash. ;-)

So, I slapped together a little design with a couple of 6116s and LS245s
and stuff, that looked like a PROM to the DUT and like a memory block to
the host PC.

Well, I was all set to go ahead and prototype it up myself, but one of
the guys who outranked me said, "Just give it to the production people -
they'll have one for you by this afternoon."

Well, does anyone remember those schematics in the early magazines and
stuff, where they show all of the bypass capacitors on a bus, and just
"Vcc" and "Gnd" labels at the chips? This board was wire-wrapped
impeccably - yea, verily, perfectly - to my schematic. The assembler had
put all of my bypass capacitors right next to each other on busses right
at the power entry point to the board. Each chip had Vcc and Gnd, of
course, but EEk! I really had to restrain myself and be all patient and
stuff as I 'splained to the assembler, "Oh, gosh, I'm sorry, I only drew
them that way to indicate that they're _electrically_ in parallel, there
was supposed to be one per chip..."

And, not necessarily surprisingly, she immediately understood the point I
was trying to make, and while she was redoing the wiring I redrew the
schematic with all of the damn lines drawn to all of the damn chips and
all of the damn bypass capacitors right next to their damn chips; but it
worked like a champ! :) :) :)

Cheers!
Rich
 
S

Sjouke Burry

Rich said:
I've read about something that was like an acoustical memory - there was a
pool of mercury, in a 7-sided container, with ultrasonic transducers, that
sent a wave that bounced off 5 sides and was read out by the other
transducer, but apparently that wasn't a big seller. Or the storage CRT. :)

Cheers!
Rich
In 1967-8 my company considered bying a french
computer with mercury delayline memory.
Luck would have it that they bought a PDP7
(8K 18bit mem) with papertape and dectape i/o,
utilised ~20 hours per day.
Later I heard it went to the DEC museum.
 
A

Anthony Fremont

Keith Williams said:
IIRC there were a few others, though nothing with a '186 was 100%
IBMPC compatible (BIOS IRQs were all messed up).

That may be, but I don't recall ever seeing or hearing about any. But I
didn't pay that much attention to them when they came out, as I was
really into the mainframe that I was using. I do know that the RS one
was seriously in-compatible. ;-)
 
K

Keith Williams

I've read about something that was like an acoustical memory - there was a
pool of mercury, in a 7-sided container, with ultrasonic transducers, that
sent a wave that bounced off 5 sides and was read out by the other
transducer, but apparently that wasn't a big seller. Or the storage CRT. :)

Storage CRT? You mean the Williams Tube Memory (I wonder how I
managed to remember that one ;)?

http://www.cedmagic.com/history/williams-tube.html
 
Top