FPGA Altair Advice

I was working on a Macintosh clone project (a clone of the original
128k), but I'm not willing to invest any more time in something that
Apple would probably kill with a C&D order.

I'm now working on a hardware replica of the Altair.  I'm scanning the
PCB layout and creating historically accurate replicas.  I've come to
the conclusion that the average person is not going to be able to
afford a kit "just for the fun of it".  There are literally square feet
of PCBs required and a lot of the components will be expensive.

This leads me to the idea of putting the Altair into an FPGA for people
who want the blinking light effect but could care less about the guts.
This would be similar to the PDP-8 clone that is an inch thick and can
be hung on a wall like an interactive picture.  I don't have any
experience designing logic for FPGAs, but like learning
C/Assembler/Visual basic...I can probably use code examples to teach
myself.

Basically what I'm looking for is an FPGA development board that would
be suited to hold the 8080.  I would also like to integrate an
"optional" boot ROM, RAM, serial card, cassette card, etc.  These
devices wouldn't take up too many resources I'd think.

The only possible problem I can think of for an Altair FPGA is that I
would want all the bus signals brought out for the front panel and for
optional interface cards.  Is the T80 core accurate enough to produce
the two clock phases and all of the bus signals?

What development board would you suggest I buy for this purpose?  I've
heard about schematic entry for the logic, and I'd like to have that
tool as an option. 

Thanks for your time,
Grant

Grant,

A bit off your topic, but building any computing machine for your own 
use is perfectly legal.  Oh so it was the last time I checked.

Xerox PARC tried to buy a DEC 10, but DEC wouldn't sell them one.  So, 
Xerox got the schematics, put everyone to work, and built one.  Just 
one.  Used it for research.  Even made their own operating system for it.

Nothing DEC could do:  Xerox PARC did not derive any profit from it, did 
not use it to do any product (and they had the losses to prove it!).

I wonder if today Apple would even care if you went into business 
offering (old) Mac Clones?

As long as you are not using their IP, their patents (and not impacting 
their present business), they really have no reason to care.

An Altair clone?  Why?  So that you could use those old S100 interface 
pcbs?  Run CPM?

Austin

Well, all of the old mac websites were sent C&D letters threating them
if they didn't remove system software required to run a mac from 1984.

If you aren't a mac guy already, it would be pretty hard to get one of
those old machines running because of Apple's actions.

Altair clone, because, I don't know...its the Altair!  :)  I want to
make a computer "kit" and have it be useful at the same time.

logjam wrote:

> Well, all of the old mac websites were sent C&D letters threating them
> if they didn't remove system software required to run a mac from 1984.
> 
> If you aren't a mac guy already, it would be pretty hard to get one of
> those old machines running because of Apple's actions.

  You could write to Steve, and tell him what you are trying to do ?

  Ponders: With the new Intel MACs, and Linux FPGA tools, how close are
we to being able to do a FPGA build of a MAC, on an Apple PC ?
  [ Apple's marketing guys could love that.... ]


  Mention that more enlightened companies, like Borland
- and even (gasp) Microsoft have Museum policies, where you can
either download, or freely use, some of their really old software.
  Some companies are proud of their heritage..... and the WEB is
ideal for this type of repositry.

  In some cases is it just a lawyer reflex, or they may be worried about
possible support questions, but the commercial impact is going to be Nil.

  -jg

"logjam" <grant@cmosxray.com> writes:
> Is the T80 core accurate enough to produce
> the two clock phases and all of the bus signals?

No, no two-phase clock.  You'd have to add something to generate it.
The T80 bus is much closer to the Z80 bus than the 8080 bus.

> What development board would you suggest I buy for this purpose?

I like the Spartan-3 Starter Kit, which you can buy from Digilent with
an XC3S1000 for $149.

The I/O expansion connectors suck, though.

> I've heard about schematic entry for the logic, and I'd like to have that
> tool as an option. 

Personally, I find it much easier to deal with equations in an HDL.
I can understand logic symbols on a schematic just fine, but if someone
asks me to draw a schematic, it takes me more time to convert my mental
concept into a bunch of logic symbols than to just type an equation.
However, other people find the opposite to be true.

Anyhow, the Altair logic uses lots of asynchronous crap like one-shots.
I suppose you could simulate the oneshots by counters with a really high
clock rate, but the design would still be fundametally asynchronous.
Async logic in FPGAs is not a good idea; other people here can tell you
horror stories about it.

Austin Lesea wrote:
> Xerox PARC tried to buy a DEC 10, but DEC wouldn't sell them one.

No.  Xerox wouldn't let them, since they'd just purchased Scientific
Data Systems (SDS), which they renamed to XDS.  The corporate types
thought (perhaps correctly) that it would negatively impact XDS sales
if customers got wind that XDS machines weren't good enough for Xerox'
own internal use.

Of course, the fact that PARC built their own PDP-10 rather than using
XDS machines would send the same message to customers.

> So, Xerox got the schematics, put everyone to work, and built one.

Xerox may or may not have had the DEC schematics, but they designed their
own PDP-10 clone ("MAXC") from the ground up.  There was no real similarity
to the DEC hardware; MAXC was simply designed to execute the same
instruction set.

> Just one.

Two, actually.  And the second was a redesign, not a copy of the first.

> Used it for research.  Even made their own operating system for
> it.

They wrote their own operating systems for various small computers they
developed, but on MAXC they used TENEX, which came from BBN.  A few
years later DEC used TENEX as the basis for TOPS-20, which ran on the
PDP-10 processor of the DECSYSTEM-20.

> Nothing DEC could do:  Xerox PARC did not derive any profit from it,
> did not use it to do any product (and they had the losses to prove
> it!).

That wasn't the issue.  There weren't any patents on the PDP-10 instruction
set.  If there were, DEC might have had grounds to sue, even though PARC
didn't ship it as a product.

Profit is not the same as economic benefit.  Even if Xerox posted a loss
every year that they were using MAXC, it's quite possible that they would
have had a larger loss without it.  MAXC contributed to the development
of many successful Xerox products.

If company A sells some patented product P, and company B decides that
buying a P would save them lots of money, but builds their own P to
save even more, you can bet that company A will sue if they get wind
of it.

Unless, of course, company B develops P2, which does not use the patent
claims held by A.  In which case they still might get sued, but will
have a better defense.

I've personally been involved in a case where a company build a clone
of something, carefully avoiding the patents, but was threatened with
litigation and negotiated a license rather than spend a bunch of money
to try to defend it.

> I wonder if today Apple would even care if you went into business
> offering (old) Mac Clones?

They're not going to care unless you include a copy of the Mac ROM,
which is the main component of the "secret sauce" of the Macintosh.

> As long as you are not using their IP, their patents (and not
> impacting their present business), they really have no reason to care.

But of what use is a Mac clone without a ROM?  It's arguably of even
less use than a newborn baby.

logjam wrote:
> 
> Basically what I'm looking for is an FPGA development board that would
> be suited to hold the 8080.  I would also like to integrate an
> "optional" boot ROM, RAM, serial card, cassette card, etc.  These
> devices wouldn't take up too many resources I'd think.

There are various fpga/retro boards out there.  In particular, there is
one from a woman in germany, who wanted to replicate the C64/C128 in an
fpga board.  I forget the exact name.  Anyways, they've got things going
to the point of having a number of different cores availabe, I believe
even some atari/etc ones.  I'm sure they'd welcome someone doing an
altair "port"...

-- 
 [100~Plax]sb16i0A2172656B63616820636420726568746F6E61207473754A[dZ1!=b]salax

"Tobias Weingartner" <weingart@cs.ualberta.ca> schrieb im Newsbeitrag 
news:slrndsgrat.i5.weingart@irricana.cs.ualberta.ca...
> logjam wrote:
>>
>> Basically what I'm looking for is an FPGA development board that would
>> be suited to hold the 8080.  I would also like to integrate an
>> "optional" boot ROM, RAM, serial card, cassette card, etc.  These
>> devices wouldn't take up too many resources I'd think.
>
> There are various fpga/retro boards out there.  In particular, there is
> one from a woman in germany, who wanted to replicate the C64/C128 in an
> fpga board.  I forget the exact name.  Anyways, they've got things going
> to the point of having a number of different cores availabe, I believe
> even some atari/etc ones.  I'm sure they'd welcome someone doing an
> altair "port"...
>

that woman is Jeri and he is not from Germany but from the US, he designed
and VERY UNSUCCESFUL C1- computer, sold some distribution rights
to an german guy who paid her in advance, manufactred the boards and
got financial loss afterwards as Jeri did not keep her part of the deal
(did not deliver anything working ip cores).

there was later a guy Tobias from Germany who did that what Jeri was 
commited
todo, eg write functional IP cores for C1, including several retro engines.

but Tobias also got really pissed on Jeri and the C1, so he is currenty 
supporting
more the TREX C1 board

I do have the old C1 board, its just collecting dust somewhere. do not buy 
the C1
take eiter TREX C1 or Trenz retro or somethin else

antti

Grant

If you are looking for a development board with lots of uncommitted I/O to 
add modules then have a look at our Raggedstone1 and Broaddown2 products. 
Our MINI-CAN product also has a lot I/O but is slightly harder to use. There 
should be details of a pile of add-on modules appearing on our website 
sometime this week for Raggedstone1 and Broaddown2 which may also be of use.

If you don't need the PCI on these boards we have a connection module that 
allows the use of the interface to create 50 I/O which is 5V tolerant. We 
also have a I/O connection SODIMM that fits the socket on Broaddown2 giving 
a large pile of extra I/O on that product.

If those are not enough Broaddown4 (Virtex-4 - various up to LX160) will on 
sale shortly soon to be followed by Broaddown3 (XC3S4000/5000). Both these 
product will offer very large I/O counts.

John Adair
Enterpoint Ltd. - Home of Broaddown2. The Ultimate Spartan-3 Development 
Board.
http://www.enterpoint.co.uk


"logjam" <grant@cmosxray.com> wrote in message 
news:1137186601.951204.325810@g43g2000cwa.googlegroups.com...
>I was working on a Macintosh clone project (a clone of the original
> 128k), but I'm not willing to invest any more time in something that
> Apple would probably kill with a C&D order.
>
> I'm now working on a hardware replica of the Altair.  I'm scanning the
> PCB layout and creating historically accurate replicas.  I've come to
> the conclusion that the average person is not going to be able to
> afford a kit "just for the fun of it".  There are literally square feet
> of PCBs required and a lot of the components will be expensive.
>
> This leads me to the idea of putting the Altair into an FPGA for people
> who want the blinking light effect but could care less about the guts.
> This would be similar to the PDP-8 clone that is an inch thick and can
> be hung on a wall like an interactive picture.  I don't have any
> experience designing logic for FPGAs, but like learning
> C/Assembler/Visual basic...I can probably use code examples to teach
> myself.
>
> Basically what I'm looking for is an FPGA development board that would
> be suited to hold the 8080.  I would also like to integrate an
> "optional" boot ROM, RAM, serial card, cassette card, etc.  These
> devices wouldn't take up too many resources I'd think.
>
> The only possible problem I can think of for an Altair FPGA is that I
> would want all the bus signals brought out for the front panel and for
> optional interface cards.  Is the T80 core accurate enough to produce
> the two clock phases and all of the bus signals?
>
> What development board would you suggest I buy for this purpose?  I've
> heard about schematic entry for the logic, and I'd like to have that
> tool as an option.
>
> Thanks for your time,
> Grant
> 

I'm interested in these kits Eric.  What else other than the board
would you recomend I get?  Does this board come with things like a VGA
core?  I read on some dev board descriptions that they came with a few
modules ready to use.  Are the WebPack tools good enough?

I love the 96 uncomitted I/O on the dev-board.  This will allow the
Altair's switches and LEDs to be connected without a bunch of logic
other than buffers.  :)  I've got some crazy ideas for a "serial port"
inside the FPGA that emulates a terminal with the VGA and PS2 port, but
I'm not sure how practical that is yet.  ;)

Also on a completely different project, I need some advice.  I'm
building a 176wX110h pixel display using LEDs.  I'm almost half way
through soldering!  ;)  I plan on it being a memory mapped display for
the Altair.  I bought a bunch of LEDs surplus a while back and have to
use them.  Plus, the Altair is all about blinking lights...but I don't
know of an Altair with 17,000+ blinking LEDs.  :)

The project requires 110 shift registers to be loaded every 14ms.  I'm
going to use some dual port SRAM from cypress, which will basically
take care of ALL the Altair's RAM needs!  ;)  Would a good beginning
project be an FPGA which reads from the dual port SRAM and sends the
correct byte to the correct shift register?  Basically a compicated
parallel to serial converter?  An evolution of the project will be
character generation ability (I organized the display as 22x10
characters of 8x11 pixels)

This looks like a good board for the project:
http://www.digilentinc.com/Products/Detail.cfm?Prod=D2FT&Nav1=Products&Nav2=Programmable

Any suggestions on either crazy project would be great to have.  :)