Newsgroups: comp.robotics
Path: brunix!cat.cis.Brown.EDU!agate!tcsi.tcs.com!uunet!hobbes!earth.armory.com!rstevew
From: rstevew@armory.com (Richard Steven Walz)
Subject: Re: STEPPER DRIVER INFO NEEDE
Organization: The Armory
Date: Thu, 11 Aug 1994 08:16:06 GMT
Message-ID: <CuD2yv.E39@armory.com>
References: <9408070631.A6509wk@circellar.com> <328nsb$h6c@handler.Eng.Sun.COM>
Sender: news@armory.com (Usenet News)
Nntp-Posting-Host: deeptht.armory.com
Lines: 92

In article <328nsb$h6c@handler.Eng.Sun.COM>,
Chuck McManis <cmcmanis@Sun.COM> wrote:
>stephen.griswold@circellar.com wrote:
>:  My main problem with the base, was being unable to calibrate the two drive
>: motors, to the same speed.  One, would always seem to spin faster than the
>: other, causing the base to turn on a wide arc, when I wanted it to go
>: straight.
>
>Strangely enough, this simple sounding problem drives most hobby roboticists
>to distraction! I know that is was one of my personal demons for a long time
>and still pops up now and then. I nailed it once by designing a modern
>version of a device invented by the chinese circa 200 BC. It was called a
>"chinese differential" and in its original incarnation was a series of gears
>that attached to a vertical pole with a man pointing attached to the pole.
>As you walked (pulling this cart) if your direction changed it the difference
>in wheel travel would cause the pointing guy to turn (relative to the cart)
>in the opposite direction. By paying attention to the gearing the pointing
>guy always pointed the same direction regardless of how you turned. Thus
>you always knew your original vector, even if you happened to be going
>up a bunch of switchbacks or walking across a featureless plain. My 
>use was to use my virtual "pointing guy" as an error correction signal that
>was added to the motor driver circuit. Anyway I submitted the design to
>the CCI design contest but no joy, after all it was just a bunch of gates,
>no CPU ;-). 
>
>:  Having seen this, when I got back from the meeting, I immediately dug into
>: my 'Junk Box' (Ok? Who also has one of these, or for that matter knows what
>: one of these is?) and found two identical stepper motors I had pulled from
>: a 'Fried' line printer, and a set of 4 Fairchild FT5753M Driver ICs.
>
>Are you kidding? My junk box fills half my lab :-) I'd suggest you bag the
>driver ICs unless by some random chance you get docs on them. You can build
>stepper driver functionality into your software if you've got spare cycles
>or build a simple digital circuit (or sophisticated uP based circuit). Also
>there are current generation drivers out, like the sprague UCN6003 (check
>that number, its from memory so could be bogus), that can be had inexpensively
>but not free. (I know, you wanted to use something in your junk box, how
>about some 7474 flipflops and a 7400 rigged as an oscillator? Or maybe 
>7490 connected to a small rom? How about using a 2716 EPROM burned with 
>the step sequences so that you could drive both by clocking the address
>lines independently?) 
>
>:  Does anyone have the Pin-out for the FT5753M? the device is size equiv.
>:  to 4 TO-220 power transistors side-by-side, 12-pin SIP.
>:  Also, can you provide the specs on the chip? (maximum load limits, heat sink
>:  requirements, power, etc.)
>
>Sorry, I don't have a Fairchild data book that lists this part. If you know
>approximately when you aquired it you may check the local library for
>a copy IC Master at that time and find some cross references.
>
>: Also, side note, a few weeks back, someone began serious "Flaming" about
>: CP/M machines..  
>
>CP/M machines are fine, I've got 3 of them. My most reliable is the Ampro
>LittleBoard system. The fastest CP/M machine I have is my 486 laptop
>emulating a Z80. Its a very functional system and TurboPASCAL is my
>favorite systems language (I know, weird huh?) The only downside of these
>things are there MIPS/pound ratio is really bad. The Cromemco System III
>consumes the better part of a desk if its on top.
>
>: ... the 'Now!' generation has been severely MSDOS-Poisoned!
>
>They of course see it as you being stuck in your ways :-)
>
>: Besides, My robot uses the 8085 as its heart. I need only write the program
>: in assembler as I had before, feed it through ASM now, and feed the hex data
>: to the robot through the RS-232. (ASM would run a lot faster, than the
>: original BASIC Assembler I had been using. and I've already ported over the
>: BASIC programmer, converting it from the old read/write of the RS-232, to
>: I/O port INP/OUT for the Kaypro.
>
>And it will run faster still on your PC laptop running ZSIM or an equivalent
>emulator :-)
>
>--
>--Chuck McManis	     All opinions in this message/article are
>FirstPerson Inc.            those of the author, who may or may not
>Internet: cmcmanis@firstperson.COM   be who you think it is.
--------------------------
A Cromemco System III!!! Gawd, that S-100 piece of junk! First CP/M like
computer I ever programmed on professionally!!! Big black 19" box with a
power supply and S-100 MB on its side. Then you pick your terminal! Ran
their CDOS (CP/M clone OS). I about died a year later when I saw the first
Ampro LittleBoard that would do the same as the monstrous CROMEMCO desk
version! I finally have three of the little Ampro's and a bunch of Morrow's
as well. I collect those old gems! I just hated the S-100 bus. It never
worked right for a month at a time. It was either dirt or re-seat all the
cards or something, and it would crash exactly when you really didn't need
it to, i.e., before saving!!! Damn thing!:)
-Steve  rstevew@armory.com

