Ok - just does not make sense to me - is really outside what the S100 bus was intended to do. I think you will find it very limiting when you get to Linux and try to do I/O across it - will not be pleasant to the user. I guess you are one of those types that has a hammer and everything is a nail. I struggle with the same thing at work - I am a software person (that knows enough about hardware to be dangerous) in a hardware division that builds hardware because they can not because they should. It is just a pile of sand I tell them without software :)
I think the more interesting thing is to look at putting an FPGA on the S100 bus to be a super I/O chip and build around that. You basically had that with the LAVA board but never really looked at how it did things internally - there was a lot of potential there. The beauty with a FPGA you can attach a lot of different I/O and just reprogram it to do what you want and it is much faster than any logic you have on current boards - really a hardware persons dream though you have to learn verilog which is much nicer than the low level programming you have been doing on GALs. With respect to the TS-4900 connector good luck on trying to get that aligned so you can plug the board in by hand soldering. Dave On Monday, August 25, 2014 10:09:35 AM UTC-5, monahanz wrote: > > Different things for different folks Dave. Agree we could do with more > software input. Not to make excuses, but I consider myself a “hardware > type” knowing enough software just to be dangerous! I did the basic 6502 > and Z80, 8086, 80386 and 68K monitors (all from scratch), just to get the > hardware working. While not for everybody I just like putting hardware on > the S100 bus. Which is really just building computers. What one does > with them is a personal/hobby thing. If one wants a simple working > system you can go with an number of SBC’s out there, Ardunio, Raspberry Pi > etc. but I and (apparently a few others) just like seeing new hardware > arrangements working, I have to admit powerful modern software would be > nice too. I know I could do it, a decade ago I wrote C++ code to run a > custom program (MDI/MFC) for windows from scratch with a total of over 500k > lines of code, but it soaked up 100’s of hours of “hobby time”. In the > end I realized I was not having as much fun as doing hardware. That’s > just me. I know there are others that have exactly the inverse response and > “Need” hardware just to get the fun software written. > > > > For our small but apparently growing group there appears to be a wide > spectrum of interests. Some like the truly antique S100 systems, and in a > perfect world would use only genuine original boards, and for new boards > would like to use chips just of that era. Others would like to “pep up” > there systems to add on and play with extra components in an already > established system. Others would like to put in the S100 bus as much > hardware as they can lay their hands on and get the different CPU’s to > interact with each other. As you know in theory anyway, a unique feature > of the bus is that there can be up to 16 CPU’ present. > > > > My own personal goal is to get towards the latter in terms of hardware. > The current 80386 master/slave board probably is the last board I will do > with discrete 74LSxxx chips only. Things like the 80486, our next board, > will require a few GAL’s as well just to get things to fit on the S100 > board, but the utility of the board increases. It should be able to work > with any 8,16 or 32 bit interface – even old S100 RAM 8 bit boards. > > > > I see the likes of these ARM mini boards as, in essence, as a super chip. > We would somehow place them in the middle of an S100 board and build > around them like the 80386/80486. The upside being that the software types > could really play and have real fun with the setup. Linux in the S100 bus > would in fact be practical. Is there an easier way to have Linux at home, > definitely yes – just go with a Pi, or any number of other SBC’s out > there. The fun is in having Linux and the likes running on your unique > hardware and in the end interfacing with other unique hardware boards you > build, be they other CPU’s, graphic boards, sound boards, data storage > boards, I/O and Internet boards etc. > > > > So yes, we absolutely need more software types but first we need to get > them hardware they can trust and build upon. > > > > Case in point is your observation with the 68K board. If in fact that > board does not work with the IDE board we need to solve that hardware > issue. I’m more or less in a holding pattern on the next version of that > board pending your observations. I (and others I’m sure) really > appreciate your keen eye in finding issues like that. Andrew probably > has the numbers, but I’m guessing all told, probably 500, perhaps 1000, > N8VEM boards have now gone out. We hear little about what happens to them. > I suppose no news is good news, but it would be nice to hear some > feedback. > > > > Specifically on the TS-4900 connector support board requirement, you may > be right but actually soldering two SMD connectors to an S100 board is not > that hard. I have just finished doing this for RAM chips on our 16 & 32 MB > RAM boards. Connectors are less scary since you rally cannot overheat them. > > > > > > John > > > > > > > > > > > > *From:* yoda [mailto:[email protected] <javascript:>] > *Sent:* Monday, August 25, 2014 6:42 AM > *To:* [email protected] <javascript:> > *Cc:* [email protected] <javascript:> > *Subject:* Re: [N8VEM-S100:5018] An ARM CPU on the S100 bus > > > > I seriously doubt that - it is a 100 pin connector with pins that are > 0.8mm apart - you are not going to find anything that can do that with > 0.1in spacing thru hole. Again I don't see the point of trying to put > modern 32/64 bit processors on a crippled bus. I think what we really need > is to stabilize what we have and make sure there is supporting software. I > am not seeing many people in this group writing software (sure like > soldering boards) or the 68K issue would have been surfaced a lot sooner. > I still have not gotten IDE to work with it yet. I am not sure it is > software or hardware but I am strongly leaning to the latter. I am in the > process of verifying my software on the mini-68K ECB which I know the IDE > port works on as I have used it with John C's software. Once I confirm > that my code works there and not on the S100 version then I will know for > sure. > > > > Dave > > On Sunday, August 24, 2014 11:01:07 PM UTC-5, monahanz wrote: > > I suspect for the mating board connector there is an equivalent plated > through connector. There are so many of these thing s these days there > almost has to be one. > > > > As to speed, I looked at the Pi, it runs at 700MHz, this one is at 1GHz. > At that speed assembler driven ARM code should have absolutely no trouble > interfacing the S100 bus. A high level language, possibly. It may require > data latches on the S100 bus, but I don’t see why one would not use ARM > assembler for key interfaces. > > > > John > > > > > > *From:* [email protected] [mailto:[email protected]] *On > Behalf Of *yoda > *Sent:* Sunday, August 24, 2014 2:40 PM > *To:* [email protected] > *Cc:* [email protected]; [email protected] > *Subject:* Re: [N8VEM-S100:5018] An ARM CPU on the S100 bus > > > > That mating board has SMT connector - so I don't see where you make sense > here and John that sample mating board has not through hole connectors so I > don't see how you would connect it. I have looked at several boards like > this and they base board brings out connection to headers parallel to the > board that allows small modules to plug into it. They generally don't have > pins perpendicular to the base board so it is not friendly to mounting to > another bigger board. You can get the connectors like on the base board to > mount on the S100 board but they are SMT connectors not thru hole because > the pin spacing is much smaller than thru hole. Look carefully at the > specifications of the module and the base board. > > On Sunday, August 24, 2014 3:44:47 PM UTC-5, Andrew Bingham wrote: > > The mating side of the connector may be SMT only - > http://wiki.embeddedarm.com/wiki/File:TS-Socket_connector_photo.jpg > > On Sunday, August 24, 2014 1:35:11 PM UTC-7, monahanz wrote: > > Dave we would work with the complete board. Just plug it into sockets on > the S100 board > > > > > > *From:* yoda [mailto:[email protected]] > *Sent:* Sunday, August 24, 2014 1:11 PM > *To:* [email protected] > *Cc:* [email protected] > *Subject:* Re: [N8VEM-S100:5012] An ARM CPU on the S100 bus > > > > Hi John, > > > > The board you are suggesting will probably be a challenge as those > connectors I believe only come in SMT style and the alignment of them are > very tricky so I don't think you would be able to hand solder them. > > > > Dave > > On Sunday, August 24, 2014 1:21:35 PM UTC-5, monahanz wrote: > > Thanks for the info gb. Currently I'm leaning toward s the Technologic > TS-4900. See here:- > http://wiki.embeddedarm.com/wiki/TS-4900 and > http://www.embeddedarm.com/products/board-pictures.php?product=TS-4900 > > They offer a mini-board with two 100 pin connectors on the back that > would make the placement on an S100 board very nice. The S100 board would > be modeled after one of their "TS-Sockets" and should in theory allow one > to use a number of their "Computer on Module boards". They supply a > free IDE programming interface but somebody told me programming the I/O > control lines is not easy. There are 121 of them. Clearly plenty to > control the critical S100 lines for other S100 board I/O data etc. > > An alternative I've been looking at is the European Olimex line, for > example their A13 https://www.olimex.com/Products/SOM/A13/A13-SOM-256/ > > The advantage of them is they are more "hobby" oriented and supply much > more information. The down side is their boards don’t have the connectors > on the back. The could be placed upside-down (not great) or perhaps removed > and re-soldered. > > Anyway early in the process, I have currently started on an 80486 S100 > board which I will do first. > > Everybody out there, please feel free to supply suggestions as to "modern" > CPU board you would like to see on the S100 bus. If a cell phone can > contain a powerful computer, the is no reason our S100 boards cannot be > one! > John > > > > -----Original Message----- > From: [email protected] [mailto:[email protected]] On > Behalf Of G. Beat > Sent: Sunday, August 24, 2014 7:13 AM > To: [email protected] > Subject: [N8VEM-S100:5012] An ARM CPU on the S100 bus > > These System-On-Module (SoM) packages offer some interesting capabilities. > Ethernet / wireless support and integration on the S-100 card being a BIG > Plus. > > Connectivity is replacing most storage media for data/program transfers, > such a board could serve the role as a surrogate for other S-100 boards. > > Depending on implementation, an ARM based S-100 board (Linux) could also > eliminate the need fir a separate PC to address -- uploads, interfacing, > etc. > > Intel's migration to NUC, now in their 4th generation, shows another > paradigm shift and options. It could sit inside a S-100 case -- or > attached to back of monitor (VESA), being a super-smart > terminal/workstation. > http://www.intel.com/content/www/us/en/nuc/nuc-kit-d54250wyk.html > > gb > > -- > You received this message because you are subscribed to the Google Groups > "N8VEM-S100" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > For more options, visit https://groups.google.com/d/optout. > > -- > You received this message because you are subscribed to the Google Groups > "N8VEM-S100" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > For more options, visit https://groups.google.com/d/optout. > -- You received this message because you are subscribed to the Google Groups "N8VEM-S100" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
