Guys I think if we have multiple copies of EDA files around/distributed we are asking for trouble. There really has to be one location where they reside. From there, if a “run” is to be done Andrew can forward the most current file just before the trigger is pulled.
On updates, there is nobody more frustrated than me in having to do an update/correction or V2 of a board. This is because I want to do new more exciting boards. 80486, ARM CPU’s, Ardunio support boards etc. I hoping to clean up the current set of boards by spring time with V2 versions and then move on. The main differences with the V2 boards are: first a clean-up of minor bugs/errors, but for the CPU boards installing the sophisticated master/slave handshaking circuit I just described for the V2- 80286 board. This will tremendously increase the capability of these boards in complex multi CPU/DMA systems in the future. So for example we need to slightly modify the 68K board. I’m not sure if we should go back and do a V2 of the 8088 and 8086 boards since the 80286 is better anyway, comments please. The good news is the upcoming 80386 and later in the year 80486 will have it installed.
Boards that I think are ‘done’:- Serial-IO, 4M-RAM, Parallel –Ports, Console-IO, Dual-IDE, ZFDC, Keyboard Converter, both prototype boards, the extender board and the RTC-PIC.
Very minor changes are suggested for Z80 board, 68K and 1M ROM+RAM boards
The 6502 CPU board is fine as is, but would need to be tested/modified as mentioned above for the master/slave to sub-slave circuit. Works fine currently as a slave.
The LAVA video board update is on hold because MYLUM is doing an updated video chip – probably mid-year.
V2 updates are in the works for the SMB and MSDOS-Support Boards. These are substantial and worthwhile I think.
Andrew and I are currently doing an extremely fast 16M S100 Static RAM board. Yes, one board for the complete RAM space of the S100 bus! The catch is it will require soldering of eight surface mount 2M RAM chips. I will illustrate how to do this (it’s not difficult with these bare boards, there is sufficient solder on the board that you just touch each RAM pin with a soldering iron). Nevertheless it’s not for everybody and you may wish to stick with the 4M Board.
Please note the above boards go through numerous prototypes and when significant, are described on S100Computers.com. Andrew and I generally announce their arrival, but as you read about progress etc. please feel free to offer questions/suggestions before the final “production” version is spun. Once a production run is done it’s a real pain (for all of us) to go back and change things. We had this happen recently with the parallel ports board. There was a bug with the addressing of the PC-printer port that required cutting traces on the original board and a second “production” board. The more eyes that review the schematics the lesser the chances of things like this happening.
Well the status of the N8VEM S-100 board suite as a whole is rather imprecise. This is the main reason I removed the KiCAD EDA files last year. They go out of date as soon as they are posted. Builders make boards only to find out the EDA files are lagging the posted schematic and PCB layout files. Then they get upset for wasted effort.
Really though it is prudent to verify the files are current *before* placing a PCB order. This is way too complicated to assume everything is exactly current all the time. At any given time there are probably 8-10 boards in work between John and myself not counting side projects like S-100 Utility board (Thanks Neil!), S-100 VDP (Thanks Leon!), etc. It is safe to assume everything is changing with each release. There have been only a very few true unmodified PCB reorders that did not involve some kind of respin for whatever reason.
My suggestion is rather than try to take on the whole mass at once, pick one or a small number of boards we can definitely baseline – for example they are not presently in work or being respun. I think also if we are going to distribute these boards across the builders there should be some sort of configuration control. Right now there is none other than the directories on my hard drive. Fortunately I do have off-line back up so it won’t disappear.
I think the S-100 regular prototyping board, the S-100 68K CPU V3, the S-100 Serial IO, the S-100 6502 CPU would all be good candidates. The S-100 SMB V2, S-100 Z80 CPU V2, S-100 MS-DOS Support boards would not because they are being changed at the moment.
Also some boards are in quasi-retired status due to low demand. For example the S-100 parallel ASCII keyboard or S-100 Console IO. Other boards are completely retired due to being replaced (S-100 4MB SRAM -> S-100 16MB SRAM) but could be brought back if builders wanted another run.
Another topic we have not discussed is the libraries associated with the board EDA files. If there is one thing that goes obsolete faster than the EDA files it is the libraries that support them. They change daily as the boards are updated and new ones are made and/or respun. Fortunately the N8VEM KiCAD libraries are segregated from the rest of the distribution but there is a lot of duplication since the KiCAD libraries themselves are absolutely full of bugs that have been fixed by N8VEM builders over the years. There are many dozens custom parts and footprints of all kinds.
Thanks and have a nice day!