[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [N8VEM-S100:5045] Re: ARM CPU on the S100Bus-II



Well there is already a open core for ARM, also there has been mention of PDP 8, 11, MIPs etc which people have been interested in.  I really don't think it is that hard to use FPGAs though I am still learning.  Sat down and read a couple of books and there are a lot of tutorials on the web   Any ways you used one and did not really exploit it - LAVA is a FPGA and there are a lot more capabilities than you use in that project.  Any way I am going to move forward with my ideas - more for a  super I/O chip not a processor.  I think once you get beyond a 16 bit processor on the S100 bus, you are trying to put an Indy car engine in a Volkswagen which is not very practical.

On Tuesday, August 26, 2014 5:58:54 PM UTC-5, monahanz wrote:

Hi Dave, A while back I took a look at programming FPGA’s, got the G.R Smith FPGA’s 101 book.  Scared the daylights out of me!   I conclude the only way I could ever get up to speed programming those things would be to attend some serious programming course(s).   That in itself is not a show stopper, but even if I thoroughly mastered the art, there is no way I could hammer one into shape to emulate an ARM or Atom CPU.     To my mind why try and do this when a piece of dedicated silicon is already available so that one can get Linux etc. up and going quickly on the S100 bus.  Going back to the hammer and nail analogy, I see FPGA’s as the software flip side of the hardware types with silicon.

 Thinking more about what would be really nice for us perhaps would be something equivalent to the old Propeller CPU (but with 32 bit registers, GHz speeds and 2-4GB RAM).    It has  most I/O pins available, a simple (downloadable)  built in video & keyboard capability.  As I understand it, their next 32 bit CPU design failed last year in fabrication so there is little chance of anything close to the above soon.  Bare chip ARM, atoms CPU’s are not available to us and are BGA and so unusable anyway.  This leaves us with SOC’s & COM’s (system on a chip, computers on a module… etc.).    To my mind the closer we could get to an straight ARM/Atom etc. with just DRAM, video I/O and as many GPIO pins the better.

Andrew, the Beagle Bone has some possibilities, I don’t like the only 512MG RAM limitation. Also by today’s standards speed is slow (~600MHz). It’s going to be kind of kludge to fit it onto an S100 board.  The connector pins are on the wrong side for that.  Perhaps one could do two real short ribbon cable connectors to 0.1’ pins on the board. Alternatively de-soldering them and put them on the bottom (probably risky) could work.  Upside-down will not work. 

John

 

 

From: n8ve...@googlegroups.com [mailto:n8vem...@googlegroups.com] On Behalf Of yoda
Sent: Tuesday, August 26, 2014 11:30 AM
To: n8ve...@googlegroups.com
Subject: [N8VEM-S100:5045] Re: ARM CPU on the S100Bus-II

 

John

 

 

It opens up a lot of possibilities - you can make you custom CPU to do what you want, it has memory and you can attach about anything you want with "some programming"   I have a couple of these and you can get them without the pins soldered so you can reverse the connectors.  I am looking at build a graphics display with it for the S100 and maybe some other devices on the same board (a super I/O board).  It might be even possible to split the 32MB of memory into 16 MB of S100 bus and 16MB for graphics so you could get all I/O and memory on a single board.  The other way you could use is there are all kinds of cores available on opencores.org so you could have a generic CPU board.  Download a z80 core and it is a z80 board, download a 6809 core and it is a 6809 processor - many different cores to choose from.  And forget all that S100 glue and multiprocessor stuff - develop a s100 interface in the chip and be done with it.  You would still need transceivers and open collector drivers to drive the bus but everything else is designed in the FPGA and you just reprogram it if you make a mistake - no ripping up boards - just re-route in the chip.

 

Any ways this is the direction I am looking at - just have to find some time to start working on it.

 

Dave



On Monday, August 25, 2014 8:11:46 PM UTC-5, monahanz wrote:

Hope it's OK with everybody but I started a new tread on this topic of getting an ARM CPU on the S100 bus because the earlier one was getting long and deep.

My suggestion of using a EmbeddedARM.com TS-4900 raised serious questions about the practicality of fabricating an S100 support board with two SMD 100 pin connectors and getting the aligned right with hand soldering to the overhead CPU mini-board. 

 

I want back to the drawing boards and discovered outfits that supply the ARM CPU's using SODIMM  connectors. Common on laptops etc. This outfit "Toradex" seems to have a few that look suitable. See for example

 

They supply a base board to get one started.  I would use that to build up an S100 board. The  "Colibri T30" is a Cortex-A9 based CPU and should provide decent Linus and graphics.  If I understand the terminology correctly the board has 110 GPIO lines some of which one would use to drive the S100 bus signals to talk to S100 I/O boards etc.

 

 

Could those of you familiar with such things take a glance at the above URL to see if I missing something major before I dig further.

Thanks

John

 

--
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 n8vem-s100+...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.