master/slave using pc hardware

[Crumpite]

Hell no, AdeV, don't go to General,  this is Daryl's and my little corner, and we encourage rambling about old favorite computers.   Besides, Bob starting it by reusing old PC hardware in an ingenious way.  

Yup, it's Bob's fault !
No one that creative should be allowed to show the rest of us up ! (IMHO  )

My first computer was a SOL-20, S-100 buss, I soldered it together myself - by the time I got rid of it I had 64kb memory and all sorts of assemblers, compilers, chess programs, etc.
Then an OSI with built in (defective) basic. I helped a local farmer build a financial program for it.
Then the Commodore 64. My wife was doing desktop publishing with it while IBM folks were dreaming about it.
Next came the Amiga and it's video and awesome graphics, we were multitasking while IBM folks dreamed about it.
And of course, all of the IBM, Univac and DEC systems - yikes ! I suddenly realize that I'm getting old, Sob....

Did I really use punchcards and card sorters and such through collage
Hmmm, perhaps the good old days weren't *quite* that good at times.
I do believe that the muscular strength of my left side is from the days of carrying two boxes of punch cards around all the time between classes.

My gout is acting up so I'm up in the middle of the night, annoying folks on this fine forum.
Daryl

[dubbleUJay]

Hi Daryl I don't think there's many folks around as you guys are all sleepin', but I'm up, it's 10h30 in the mornin'
Wilhelm

[mobile_bob]

my first was a timex/sinclair, the cute little bugger that you connected to a tv for a monitor
and a cassette tape deck for storage, it came with a whopping kilobyte of memory, and i bought the massive
i am talking huge, monsterous 4kb expansion module for another hundred bucks.

it worked, although finicky as hell with that joke of a touch sensitive keyboard made for the tiniest of fingers.

moved up from there a few years later to an ibm pc, the first one, with a 10 meg hard drive, wow
still it had 640 kb of memory, and a better keyboard what at least worked reliably even if the rest of the machine
was a joke.

wasn't till i got involved with ibm's ps2 line of microchannel machines that i found a home, with os2 warp, i was thrilled
right up till the day ibm quit the machine and later the operating system.

i haven't been right since,, course i probably was a half bubble off to start with.



bob g

[AdeV]

I was lucky(?) enough to miss the punch-card generation, by a good 10 years I think. The college I went to had a DEC VAX 8800, of which they were extremely proud. One was given an account on that machine, and a cubby hole into which your printouts were placed; you could tell if your batch job had gone wrong, as the pile of paper in your cubby hole was massive...

I have fond memories of writing a football (soccer) "simulator" in DCL - which half of my year played as a "fantasy football" thing, trading players off each other. Another, far cleverer than I, chap wrote an interactive "magazine" entirely in DCL. None of us considered that it looked remarkably like a primitive form of WWW - none of us having seen any WWW content at that point... no graphics, though, as it had to work on standard WYSE terminals.

AH, happy days. There's probably more power in my desktop computer than there was in that VAX, but I really got to like that machine - insofar as any machine is likable...


PS: It's 8:30am here, and I'm only up because I'm expecting the plumber anytime soon to come fix the central heating...

[quinnf]

VAX?  Yeah my employer had one of them.  We'd collect data from our instruments on an IBM mainframe.  The raw numbers would accumulate in user accounts and dump to a lineprinter.  We'd take the printout and go over to the VAX and manually enter the numbers to number crunch and print out nice pretty graphs on another printer.  Then we'd take the graph and processed data and go over to the Wang word processor to write the text of our report.  The data and graphs couldn't be embedded in the report.  They were attached as an appendix.  And data was archived on (yet) another IBM mainframe.  And none of these machines talked to each other!

One day, out of frustration a guy brought his Mac SE and Imagewriter color printer from home and started doing his work on it.  He had a critical deadline to meet, and expected to take the computer home after he was finished.  The guy suddenly became very popular, and the computer stayed.  Soon, there was a clipboard with a signup sheet screwed to his office door.  People were willing to work on the Mac during lunch and after work in preference to the VAX/Wang because it was so much faster to use and the output was so much nicer.  

Within a year, the entire IS group had been laid off and Macs were in every office.  Ten years later, the Macs were replaced by networked PCs.  And a few geeks were hired to ride herd on the network.  A few years later a new IS group was formed.  Then server-based applications and multiple databases began to appear on the network.  None of the databases share data or are globally searchable.  IS doesn't see that as a problem.  

Is history repeating itself?

Merry Christmas!

Quinn

[BruceM]

Ah, yes, the Timex Sinclair, I remember it well.  TV video, scanned out by the Z80 addressing, only when the program (basic interpreter) wasn't running; blank screen until program finished.  Clever minimalist hardware and cheap, I had a lot of fun with it.

I did mess with a Comodore 64, but didn't own one.

Next (for my home hardware) was the TRS-80 from Radio Shack. I disassembled (Z80) their "Electric Pencil" word processor, which only did cassette.  I wrote a diskette file system for it,  modded the video for upper and lower case characters.  Made a decent word processor for cheap.  Alas, before I could sell more than a few,  RS came out with a new diskette based version.

After the TRS-80 I upgraded to the North Star Horizon S-100 bus machine, a clone of the "Terminal Interface Multiplexors" (TIMs) which my then boss and I had spread throughout our lab-  the AF had slapped a moratorium on small computers because of fear of proliferation and idiotic "big iron" thinking.  So we built our own non-computer TIMs from S100 components, and officially used them for "off line" data and program entry, which programs could then be transferred to the main computer for compilation.  Of course, to further save time, you could do a local Fortran compile "to check for syntax errors", and the mathematicians needed regression and statistical analysis...

It was a highly successful program (around 1978) , every engineer and mathematician got a full house computer, crudely networked to the main flight simulator computer and it ushered in a new era of projects and productivity for our lab.  

I started working at that AF research lab about 1977, where some old tube and transistor analog fight simulators where still being used in some lower priority research projects. I when I first started as a college student summer hire, I operated one system that had to be booted from paper tape, then mag tape, and finally a vacuum drum (hard disk the size of a beer keg).  It was a great experience to see and use some of the (recent) history of real time computing. We still did punch cards, but only for a couple years more.

Quinn just posted about small computer networks vs Big Iron.  I saw the same stupidity at Motorola and McDonnel Douglas Helicopter.  My "renegade" projects (the ones where money was no object but schedule was not negotiable) always bypassed the big iron- because with a large software group, the big VAX or whatever would ALWAYS be on it's knees come major milestones and deadlines. (The big hard drive bottleneck.)  With dedicated computers, your own compile and load build time (or your small groups) build of new software was always predictable and fast.  

Yet upper management and IT are often more concerned about "control" than productivity, and the urge to centralize and "control" seems eternal.  As Quinn suggests, I think stupid ideas (big iron, central computer)  just keep coming back around again.

We never had one iota of problems with software version and baseline control in any of the distributed development systems I used.  

A failure of a single system meant nothing to the project, and often  dedicated systems were loaded with actual interface hardware, allowing a developer his own hardware, right at his desk or in his own spot in the lab, for early testing and integration, making subsystem hardware and software development much faster, and virtually eliminating the typical "systems integration" schedule debacle.

Big Iron is dead (yea!)-  ling live Big Iron (boo).

[loonogs]

Mobile Bob,
any update on the project, i just came across it and its cool

Philip

[mobile_bob]

haven't started serious assembly... yet
however i have done some subassembly work for testing, and am very happy with the results

my mode of operation is quite different and maddening to some folks, actually most folks i guess

i don't start a build on anything until it is perfected in my mind, all drawn out, and where i feel i have
a near 100% chance of success, so

this leaves me at times waiting a very long time by most folks standards before i start a project.

the upside to my approach is, generally what starts out as a complex problem turns out to have a rather
simple solution.

also i had been waiting for a source of a specific little part and found some on ebay recently, got them in
so now i can start assembly.

i think i have all the hardware sorted out now, at least i hope so  

bob g

[loonogs]

well good luck with it,
you have a lot of interested people here waiting for the next installment.

Philip

[Westcliffe01]

I just bought one of these:

http://www.freescale.com/webapp/sps/site/prod_summary.jsp?code=DEMO9S12HY64&fsrch=1

Cost $99
Software: Free  
Code: Assembly language (unlimited)
         C (32k limit)
Many additional development / Demo boards available

[Westcliffe01]

This page gives an idea of the extent that large scale integration has gone in simplifying the hardware side of things for small engine control.

http://www.embeddedstar.com/weblog/2009/08/28/mcz33812-s12-mcu/

[mobile_bob]

thanks for this link, it might well be the dual fuel solution i have been looking for
it also has the O2 sensor which is really cool.

bob g

[Westcliffe01]

On this page http://www.freescale.com/webapp/sps/site/prod_summary.jsp?code=KIT33812ECUEVME&parentCode=MC33812&fpsp=1&nodeId=01435979968460
Freescale offers a complete development board with all of the pertinent devices installed and a basic set of code to allow you to run an engine.  Cost $235.31



Their highly detailed handbook for this kit and software is too large to attach but it can be found here:
http://cache.freescale.com/files/analog/doc/user_guide/KT33812ECUUG.pdf?fpsp=1