rhyde@ucrmath.ucr.edu (randy hyde) (11/29/90)
My board arrived. Currently I'm pouring over the schematics before turning on the soldering iron. I've got a couple of questions someone might be able to help me with. 1) I note that the DUARTs live at ports 2800020, ...40, ...60, etc. Yet the address lines going into the 74138 imply that they should be at base addresses ...10, ...20, ...30, etc. Is the documentation correct? If so, is this an artifact of the decoding for the ICU? 2) I'm interested in a little off-board expansion without going through the SCSI bus (I'm not ready to tackle the design of a smart SCSI controller). I was thinking of popping off one of the DUARTs, running a couple of addit- ional address lines (and unused decoding lines from the '138) over to the socket and building a daughterboard to plug into the DUART socket. I'd like to put a couple of parallel ports (perhaps a 6522 or 8255), an 8530 (I want to build an Appletalk interface), and some other devices. Any problems anyone can warn me about? 3) Along the lines of question 2, I'd like some additional interrupts. can I cascade another ICU using the daughterboard technique des- cribed above? 4) Where do I get documentation for the monitor program? 5) Are the source listings available for the monitor? I'd like to see some code for initializing all the chips. Thanks, Randy Hyde .
rjk@sequent.com (Robert Kelley) (11/30/90)
> > 2) I'm interested in a little off-board expansion without going through the > SCSI bus (I'm not ready to tackle the design of a smart SCSI controller). > I was thinking of popping off one of the DUARTs, running a couple of addit- > ional address lines (and unused decoding lines from the '138) > over to the socket and building a daughterboard to plug into > the DUART socket. I'd like to put a couple of parallel ports > (perhaps a 6522 or 8255), an 8530 (I want to build an Appletalk > interface), and some other devices. Any problems anyone can warn > me about? > > 3) Along the lines of question 2, I'd like some additional interrupts. > can I cascade another ICU using the daughterboard technique des- > cribed above? > Me too, I want to hang strange devices off my PC532 without the hassle of SCSI. I haven't thought about exactly how to do it yet, but let's keep in touch about it. Bob Kelley rjk@sequent.com
culberts@hplwbc.hpl.hp.com (Bruce Culbertson) (11/30/90)
rhyde@ucrmath.ucr.edu (randy hyde) writes: > 1) I note that the DUARTs live at ports 2800020, ...40, ...60, etc. Yet the > address lines going into the 74138 imply that they should be at base > addresses ...10, ...20, ...30, etc. Is the documentation correct? If so, > is this an artifact of the decoding for the ICU? You are right. That was to see if you were paying attention. > 4) Where do I get documentation for the monitor program? This is my fault. I wrote it and should document it. Until I do, send e-mail if you have questions. Most people seem to have figured out how to use it by reading its ROM-squandering on-line help. > 5) Are the source listings available for the monitor? I'd like > to see some code for initializing all the chips. Yes, I can mail them to you. If you order pc532-Minix, they will be included. Have fun, Bruce culberts@hplabs.hp.com
news@daver.bungi.com (11/30/90)
> My board arrived. Currently I'm pouring over the schematics before turning > on the soldering iron. I've got a couple of questions someone might be able > to help me with. > > 1) I note that the DUARTs live at ports 2800020, ...40, ...60, etc. Yet the > address lines going into the 74138 imply that they should be at base > addresses ...10, ...20, ...30, etc. Is the documentation correct? If so, > is this an artifact of the decoding for the ICU? > > 2) I'm interested in a little off-board expansion without going through the > SCSI bus (I'm not ready to tackle the design of a smart SCSI controller). > I was thinking of popping off one of the DUARTs, running a couple of addit- > ional address lines (and unused decoding lines from the '138) > over to the socket and building a daughterboard to plug into > the DUART socket. I'd like to put a couple of parallel ports > (perhaps a 6522 or 8255), an 8530 (I want to build an Appletalk > interface), and some other devices. Any problems anyone can warn > me about? A couple of suggestions. For the DUART positions, fabricate (with breakable strip sockets) PGA style sockets. Then mount the DUARTS in the regular PLCC socket and plug this assembly into the PGA style socket. The PGA style socket will provide the interconnect flexibility you will require for expansion. Also, note that PLCC sockets generally have slots on two diagonal corners for chip extraction. The slot width just so happens to be the same width as the thickness of a typical PC board -- thus making it possible to attach a prototype board to the PLCC socket. That is, the board is vertial to the socket and orientated along the diagonal. Wires then can be run from the socket to the proto- type board, etc.. The end result is presto -- a pluggable assembly, with lines [BD0..BD8], [A0..A3], /RD, /WR, /CS, /INT, etc. Also removing the associated RS232 drivers provides 2 * 6 (ie 12) additional uncommitted lines to the assembly. I trust the above is clear :-).. -- John Connin: manatee Orlando, Florida UUCP: {uunet,ge-dab,ucf-cs}!tarpit!tous!manatee!johnc
gs@vw25.chips.com (George Scolaro) (11/30/90)
[In the message entitled "Re: PC532 Expansion & other questions" on Nov 28, 18:03, randy hyde writes:] > My board arrived. Currently I'm pouring over the schematics before turning > on the soldering iron. I've got a couple of questions someone might be able > to help me with. > > 1) I note that the DUARTs live at ports 2800020, ...40, ...60, etc. Yet the > address lines going into the 74138 imply that they should be at base > addresses ...10, ...20, ...30, etc. Is the documentation correct? If so, > is this an artifact of the decoding for the ICU? The documentation is wrong. Hmm, why hasn't anyone else noticed this, every just looking at the schematics? > 2) I'm interested in a little off-board expansion without going through the > SCSI bus (I'm not ready to tackle the design of a smart SCSI controller). > I was thinking of popping off one of the DUARTs, running a couple of addit- > ional address lines (and unused decoding lines from the '138) > over to the socket and building a daughterboard to plug into > the DUART socket. I'd like to put a couple of parallel ports > (perhaps a 6522 or 8255), an 8530 (I want to build an Appletalk > interface), and some other devices. Any problems anyone can warn > me about? > > 3) Along the lines of question 2, I'd like some additional interrupts. > can I cascade another ICU using the daughterboard technique des- > cribed above? I suppose it should work. Have fun. Just make sure you don't run the interface too far from the pc532. Even though the peripheral data bus is buffered, it is still pretty fast. On second thoughts, you will probably have trouble with timing. The 8530 is real slow on the bus (last time I looked), and things like ancient 8255's are probably even slower so you might have to jiggle with the wait pal to get an extra wait state or so into the peripheral accesses - have more fun! best regards, -- George Scolaro (gs@vw25.chips.com) Chips & Technologies (408) 434-0600 3050 Zanker Road San Jose, CA 95134
news@daver.bungi.com (11/30/90)
Randy: > My board arrived. Currently I'm pouring over the schematics before turning Pouring over them? By the way, would flux remover help? I tried using alcohol with little success. But the board seems to run fine anyway. Any recommendations? > on the soldering iron. I've got a couple of questions someone might be able > to help me with. > > 1) I note that the DUARTs live at ports 2800020, ...40, ...60, etc. Yet the > address lines going into the 74138 imply that they should be at base > addresses ...10, ...20, ...30, etc. Is the documentation correct? If so, > is this an artifact of the decoding for the ICU? > > 2) I'm interested in a little off-board expansion without going through the > SCSI bus (I'm not ready to tackle the design of a smart SCSI controller). But how about a dumb scsi controller? From looking at the ANSI spec it looks like an 8751 should do just fine. A Z-80 with a PIO should be better. > I was thinking of popping off one of the DUARTs, running a couple of addit- > ional address lines (and unused decoding lines from the '138) > over to the socket and building a daughterboard to plug into > the DUART socket. I'd like to put a couple of parallel ports > (perhaps a 6522 or 8255), an 8530 (I want to build an Appletalk > interface), and some other devices. Any problems anyone can warn > me about? Any concern about speed at all? > 3) Along the lines of question 2, I'd like some additional interrupts. > can I cascade another ICU using the daughterboard technique des- > cribed above? > > 4) Where do I get documentation for the monitor program? > > 5) Are the source listings available for the monitor? I'd like > to see some code for initializing all the chips. >
news@daver.bungi.com (11/30/90)
> 2) I'm interested in a little off-board expansion without going through the > SCSI bus (I'm not ready to tackle the design of a smart SCSI controller). > I was thinking of popping off one of the DUARTs, running a couple of addit- > ional address lines (and unused decoding lines from the '138) > over to the socket and building a daughterboard to plug into > the DUART socket. I'd like to put a couple of parallel ports > (perhaps a 6522 or 8255), an 8530 (I want to build an Appletalk > interface), and some other devices. Any problems anyone can warn > me about? May I suggest you also consider using a PC as an IO server. If performance is not the primary consideration, all things considered, this seems to me to be a clean approach. -- John Connin: manatee Orlando, Florida UUCP: {uunet,ge-dab,ucf-cs}!tarpit!tous!manatee!johnc
george@wombat.bungi.COM (George Scolaro) (11/30/90)
[In the message entitled "Re: PC532 Expansion & other questions" on Nov 29, 21:12, Rick Rodman writes:] > > Randy: > > > My board arrived. Currently I'm pouring over the schematics before turning > > Pouring over them? By the way, would flux remover help? I tried using > alcohol with little success. But the board seems to run fine anyway. > Any recommendations? The best product I have found (there are probably more) is made by RAWN and the product is called FLUXSOLV, part number 11010. It works very well. Once you have soldered everything in, keep the board at a near vertical position, start at one corner and slowly work your way across and down, washing the gunk towards the bottom. If you use too much pressure (adjustable nozzle) you might spray some gunk through the vias, so use it gently. Basically, try and keep the liquid gunk on the solder side as you wash it off, you especially don't want it to get into the sockets. Also, it is best to use it just after soldering, the flux is still soft, it hardens with age and will require more flux off to remove it. Note that modern day solders have non-corrosive fluxes so it is ok to leave the flux on. The main reason for removing it is that you can then inspect your solder joints much easier - flux tends to hide nasty dry joints etc. Oh also, the FLUXSOLV likes to eat some plastics so don't use it near things like oscilloscope front panels etc :-( also use it in a well ventilated area - some of us have few enough brain cells left as it is... best regards, -- George Scolaro george@wombat.bungi.com [37 20 51 N / 122 03 07 W]
news@daver.bungi.com (12/01/90)
George Scolaro writes: >On second thoughts, you will probably have trouble with timing. The 8530 is >real slow on the bus (last time I looked), and things like ancient 8255's >are probably even slower so you might have to jiggle with the wait pal to >get an extra wait state or so into the peripheral accesses - have more fun! On third thought, you can probably get away with this on the hardware side, but Intel's docs say something like the software must emulate the action of the (non-existant) wait-state generator. The usual hold-off for Intel MOS parts is 450 ns. I usually just used a tuned loop or an interrupt from a timer device. If you did not do this you can get the microcode in these parts into some pretty exotic states. -- J.R. (Use the Source, Luke) Stoner "Dying is easy, | asgard@montana.portal.com comedy is hard." | asgard@zorch.sf-bay.org
news@daver.bungi.com (12/01/90)
> > My board arrived. Currently I'm pouring over the schematics before turning > > Pouring over them? By the way, would flux remover help? I tried using > alcohol with little success. But the board seems to run fine anyway. > Any recommendations? If you promise not to harass me, giggle, or whatever I will tell you what I used.. Well anyway, I found some automotive "Wax and Grease Remover" in the garage -- you know, the stuff you use to prep a car before you repaint it.. Just scrubbed it down using an old tooth brush and then washed it off with ivory liquid soap. To dry it I used a hair dryer. Works great provided you are careful to not let the petroleum distillate run to the top side of the board and get under the components. -- John Connin: manatee Orlando, Florida UUCP: {uunet,ge-dab,ucf-cs}!tarpit!tous!manatee!johnc
news@daver.bungi.com (12/01/90)
I believe I screwed up the address the first time I posted this, so I will give it another whirl.. > My board arrived. Currently I'm pouring over the schematics before turning > on the soldering iron. I've got a couple of questions someone might be able > to help me with. > > 1) I note that the DUARTs live at ports 2800020, ...40, ...60, etc. Yet the > address lines going into the 74138 imply that they should be at base > addresses ...10, ...20, ...30, etc. Is the documentation correct? If so, > is this an artifact of the decoding for the ICU? > > 2) I'm interested in a little off-board expansion without going through the > SCSI bus (I'm not ready to tackle the design of a smart SCSI controller). > I was thinking of popping off one of the DUARTs, running a couple of addit- > ional address lines (and unused decoding lines from the '138) > over to the socket and building a daughterboard to plug into > the DUART socket. I'd like to put a couple of parallel ports > (perhaps a 6522 or 8255), an 8530 (I want to build an Appletalk > interface), and some other devices. Any problems anyone can warn > me about? A couple of suggestions. For the DUART positions, fabricate (with breakable strip sockets) PGA style sockets. Then mount the DUARTS in the regular PLCC socket and plug this assembly into the PGA style socket. The PGA style socket will provide the interconnect flexibility you will require for expansion. Also, note that PLCC sockets generally have slots on two diagonal corners for chip extraction. The slot width just so happens to be the same width as the thickness of a typical PC board -- thus making it possible to attach a prototype board to the PLCC socket. That is, the board is vertial to the socket and orientated along the diagonal. Wires then can be run from the socket to the proto- type board, etc.. The end result is presto -- a pluggable assembly, with lines [BD0..BD8], [A0..A3], /RD, /WR, /CS, /INT, etc. Also removing the associated RS232 drivers provides 2 * 6 (ie 12) additional uncommitted lines to the assembly. I trust the above is clear :-) -- just don't ask me to draw a picture of it. -- John Connin: manatee Orlando, Florida UUCP: {uunet,ge-dab,ucf-cs}!tarpit!tous!manatee!johnc
rhyde@ucrmath.ucr.edu (randy hyde) (12/01/90)
>> Have you considered a dumb SCSI...
Yeah... that's for my second project. *I'm* too dumb to deal with
SCSI just yet. Once I get the board up and running and I hack it
to pieces hanging things here, there, and about, I'll probably
learn how to build a Smart Peripheral.
Indeed my second project will probably be to move a file system
onto a second computer (I dread the thought of using a PC, not
because I've got anything against the PC, but because MS-DOS
is my only option and it is a *LOUSY* file system). Perhaps
if MINIX 2.0 is running (I dunno about it today, I'm still
catching up on the articles from May) I'll make Minix's
file system remote.
As for the reason for sticking the peripherals on the DUART
socket, this seemed like the most reasonable expansion point
for (hardware) hackers like myself. It requires the fewest
cuts and jumps on the PC532 motherboard. I don't need 8
serial channels, so it made sense. As for why I'm doing
this, as I said, I'm a hacker when it comes to hardware
(hacker=someone without considerable education in a
given area, one who "hacks" things out [like a hack writer]).
I enjoy doing stuff like this for fun. I may never actually
use the additions, the fun is in making them work.
BTW: I just purchased a "Serial Box" from Genovation which
lets you connect a PC keyboard up to a serial port. Lets me
put a full-featured keyboard (i.e., not limited to ASCII) on
the PC532. As soon as my board is up and running I'll let
you know about it.
*** Randy Hyde O-)rhyde@ucrmath.ucr.edu (randy hyde) (12/01/90)
I'm certainly considering using a PC as an I/O processor. As posted earlier, however, my motiviation for adding I/O devices is just for the fun of it, not for any serious process. I'm trying to catch up on the old articles. As of about May no one had found a decent SCSI card for the PC which talks nicely to the PC532. Has this situation changed recently? *** Randy Hyde