Can anyone please tell me if the Olivetti M20 used a Z8010 memory
management unit?
When I worked for Marconi Elliott Avionics at Rochester we designed
the anti submarine
sub system for the WG34 (Later EH101) helicopter using a dozen Z8001s
and Z8010s
plus lots of specialised processors. After I left the company, the
project dragged on for
many years before being cancelled and a rival, more mainstream system
fitted.
The Z8001 had 16 low order address lines and 7 segment selector
lines, so even
without an MMU, it should have an 8MB address range. IIRC, the MMU
took the
top 9 address lines and the 7 segment selector lines and could map
them onto
a 16MB address range. For speed and simplicity, I came up with the
idea of never
setting the bottom bit in the segment base registers so we could send
the bottom
8 bits of the CPU address directly as the row address to the RAM
chips and
by the time they were ready for the column address, the MMU look-up
was done.
Hence we allocated memory in 256 byte chunks instead of 128 bytes as
envisaged
by Zilog.
In the M20 uses a Z8010 I might get one if one comes up for sale in
Europe.
Roger Holmes
On 10 Jan, 2007, at 16:36, cctalk-request at classiccmp.org wrote:
> Message: 8
> Date: Tue, 09 Jan 2007 22:49:27 -0800
> From: "dwight elvey" <dkelvey at hotmail.com>
> Subject: Re: Olivetti M20
> To: cctalk at classiccmp.org
> Message-ID: <BAY130-F3347994A26286E3E09D04DA3B20 at phx.gbl>
> Content-Type: text/plain; format=flowed
>
>> From: Gene Buckle <geneb at simpits.com>
>>
>> dwight elvey wrote:
>>> Hi
>>> For those that might be interested in an Olivetti M20, there is one
>>> on eBay. This was one of the few machines made with a Z8000
>>> processor.
>>> I think the only personal computer.
>>> If someone on this list gets one, contact me and I'll help you get
>>> it up and running.
>>> Dwight
>>
>> That's cool. I just posted PDFs of the CP/M for Z8000 manuals on the
>> Retroarchive site. :)
>>
>> g.
>>
>>
>
> Hi Gene
> I would suspect that the one on eBay isn't ready for CP/M8000 right
> out of the box. I'm sure it'll handle PCOS, at least version 2.0f.
> The problem is that most machines didn't come with enough memory
> for CP/M8000. Although, the manual for CP/M8000 states that one
> can run in a machine with 128K of memory, because of a number
> of allocation issues in the M20, you really need 384K.
> The M20's comes with 128K on the mother board and usually
> have two 32K expansion memory cards. Even with three of these
> 32K cards, your only up to 224K. Still not enough.
> When I first got CP/M8000 running on my machine, I modified
> two of my memory cards to take 64K DRAM chips. This brings it
> to 384K. I was always going to increase it to 512K, the most that
> the M20 decodes, but I don't see the need any more.
> A fellow in Italy had thought that it might be easier to take two
> 1 meg simm's and use those ( you need at least 2 because it
> does access 8 bit or 16 bit and needs the separate selects ).
> We talked it over and came up with a way to wire them up
> such that they can provide 384K with a minimum of external
> hardware( total of 512K with the mother boards memory).
> The only painful issue is that the M20 normally has a select for
> either 32K or 128K at each board position. The first solution
> used three board edges to connect the two other connectors
> selects. Another fellow built one that he made a header to
> go under the NAND gate chip that normally sent out the
> selects. This did require lifting one part and putting it on a socket.
> It also seems that some of the memory boards came with sockets
> and could be configured for either 32K or 128K by jumpers
> but none of my board were that easy.
The link below shows pictures and a brief description of the MODCOMP II that
we rescued on Friday 1/12/2007. Restoration work has started on the CPU/MEM
chassis - the I/O chassis is as we got it. [Note that the pictures are kinda
large - so if you have a slow speed connection, they'll take awhile]
http://bickleywest.com/modcomp.htm
In case you're curious, here's a summary inventory of the systems we have
inhouse [unfortunately, it's incomplete]:
http://bickleywest.com/system1.htmlhttp://bickleywest.com/system2.html
Cheers,
Lyle
--
Lyle Bickley
Bickley Consulting West Inc.
Mountain View, CA
http://bickleywest.com
"Black holes are where God is dividing by zero"
I'm looking for a source of transformers like those used in the Altair
680. In addition, does anyone have any idea what the rated current output
is with the Altair 680? I do not have an assemnly manual. If anyone has
an idea of the watt rating I'd appreciate it. : )
It has 102620 and 9967736 written on it.
It only has one primary coil.
There are two secondary coils. One with a center tap and one without.
Here are some stats I took:
Primary: 121.4vAC
Secondary 1: 11.13vAC (used for +9v/+5v)
Secondary 2: 14.57vAC parallel, 29.2vAC series (used for +/-16vDC supply)
Thanks,
Grant
>
>Are you perhaps thinking of the Bede 5? A friend decided to assemble
>one in the living room of his second-story apartment. I think he
>ruined the place utterly, what with the swarf from the thousands of
>holes drilled or the broken-off stubs of pop rivets that were left in
>the green shag carpeting, not to mention the oil from the snowmobile
>engine, etc. I seem to recall that there some problem getting the
>completed craft out of the building...
>
>At the same time, I was building a harpsichord in my third-story
>apartment and while the carpet was pretty dirty from all of the
>sawdust, I think it was salvageable.
>
>Cheers,
>Chuck
>
>
I think I must be thinking of the Zenith:
http://www.zenithair.com/stolch801/construction.html
It uses what seems to amount to "Cherry" rivets. I've forgotten the details of the difference between Cherry and CherryMax, but the "Max" are significantly stronger (and worse to drill out)...
I could have sworn that I had seen a flying example of a 4-seater a/c that looked like the Zenith, with a big a/c engine providing the power. That was probably at least 10 years ago...
Good to hear that you got the harpsichord completed, and that it is still playing!
Oh, BTW, one of the engines that lots of people finally settled on for the BD-5 (prop version) was the 1200 Honda Civic engine (the "49 states" model, not the CVCC). You've got to have balls to fly the BD-5, especially with a Honda 1200 screaming at rpm's that it was never intended to run at, especially with high manifold pressure...
Regards,
Robert Greenstreet
Found on rec.aviation.homebuilt by veeduber via google groups...
http://tinyurl.com/y8qfao or http://preview.tinyurl.com/y8qfao Riveting
101
http://tinyurl.com/se4pd or http://preview.tinyurl.com/se4pd Riveting
101 - continued (Part 2 of 3)
http://tinyurl.com/y7ql2r or http://preview.tinyurl.com/y7ql2r Riveting
101 - Part Three of three.
QUOTE/
Riveting 101
...or Meet Mr. Smiley :-)
Wanna learn how to rivet?
Go down to the hardware store -- or whatever passes for a hardware store in
your community -- and buy yourself a packet of COPPER rivets. They come in
different lengths and copper washers are usually included in the packet.
Now find yourself a small hammer (...no, smaller than that... about 8oz).
Have you got an old shoe? An old LEATHER shoe? Okay, then howzabout an old
handbag. Or even some heavy canvas. Okay, good... use that. (I never
thought of riveting cardboard... I usually just weld it... )
What are you going to use for your anvil? (And yes, an anvil makes a good
anvil. But real anvils are kinda thin on the ground this season.) No, wood
won't work for an anvil, not even lignum vitae. (Yep, it makes good
prop-shaft bearings for aircraft carriers. But it don't work worth spit as
an anvil.) Well-made bench-vise... that oughta work. Face of a
sledge-hammer will do too. You work it out.
Last thing you need is a HOLE to put the rivet in. Lotsa folks forget that
the hole is part of the process. When you're riveting leather, canvas or
cardboard the hole isn't a very important part of the process -- not like it
is when you're riveting metal -- but you can't rivet without a hole so go
make one. Make it as big around as your rivet. You can make it with a punch
or an awl or a drill. Got it?
Okay, put the rivet through the hole, put a washer on the rivet and squeeze
the rivet down tight against whatever it is you're riveting. How much of the
rivet's shank is sticking up? That's too much. You only want two diameter's
worth of the rivet's shank sticking up past the washer, so cut it off.
Now hit it with the hammer. (Okay, then hit it again!) Keep hitting it
until the shank is hammered into a button about twice as wide as the shank.
Do a couple more.
/QUOTE
On 13 Jan 2007 at 23:45, William Donzelli wrote:
> It should be noted that painting the yellowed case with a mild paint
> is actually the most proper thing to do, in the mind of a professional
> conservator. Even more proper, and as done with some aircraft at the
> Gerber facility, is to apply a wax coating to the original finish, and
> then paint over the wax. This allows the paint to be removed
> completely and safely when it comes time to refinish the outer layer.
Yup. If you're just doing it for show, a water-based paint would do
the job, if you didn't handle it too much. Easy to remove--and
actually serves to preserve things by sealing out atmosphere and UV
rays.
Cheers,
Chuck
Is there a cross-reference available on the web to show what machines used what
floppy disks? I'm looking specifically for what machines use what 8" floppies,
i.e. SS, DS, hard or soft sectored, TPI, etc. It seems like it would be a useful
thing to have in a FAQ.
ard at p850ug1.demon.co.uk (Tony Duell) wrote:
> Oh yes, IIRC, if yuo put a TI58C on the cradle, it'll work, but you lose
> the contents of the continuous memory if you turn off or unplug the
> cradle from the mains.
Regarding turning off the cradle, it depends on the model number. The PC100A and 100B
models would cut off poewr to calculator when you turned off the cradle (using the
big brown slide switch on its side). The 100C cradle will continue supplying power
to the calculator even when you turn the cradle off. This is documented in the manual
and was done specifically for the 58C.
**vp
"The flash card adapter board contains the basic
circuitry needed to add an SD/MMC, miniSD, and
microSD/TransFlash socket to virtually any
microcontroller application using as few as four I/O
pins."
introductory price - $39 US
http://www.ecs87.com/adaptor.html
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
> 9000 VAX wrote:
>
>> The SCSI connector on 3100 76 is wired differently from the standard.
>> I was told that it was because there was no standard when the first
>> vaxstation was out. You need a DEC specific SCSI cable to use the
>> port.
>
>
> The wiring inside the VAXstation 3100-76 may well be "non-standard",
> but the external connector works perfectly well with any SCSI
> cable which physically connects. Internally there is a single
> cable harness even though the internal devices and external
> devices are separate SCSI channels. So you cannot replace
> the harness (with anything other than the exact same
> harness) without knowing exactly what you are doing.
>
> If the external connector has been replaced with an
> Amphenol one, then perhaps that was not done properly.
>
> There's a terminator on the harness too - and my recollection
> is that without it things do not work well (if at all).
>
> Antonio
>
Internal works just fine, external has the Amphenol-50, but since there
are 2 colors of loose wires and a proprietary internal connector it's
unlikely I can figure the "proper" wiring out. I suppose I could trace
it from the NCR chips, but for now I just pull the cover and plug stuff
in the 1st bus. The ground wires on the external bus are all one color,
and they seem to go to the right places - so it's partially right. If
only DEC had better documentation.
I hand installed NetBSD from the shell option, and it seems to be
working better (no panics yet). sushi complains about libcurses and
won't run, but vi runs ok. It isn't my main VAX, anyway, so I'll
probably shove it back under the bed for now. I was going to junk it if
it didn't work, but it pulled itself together. I'd throw VMS on it, but
(a) everyone needs at least one BSD VAX and (b) the console is shared
with the 4200 so it would be a bother to run them together as a
VAXcluster. I'm considering clustering the 4000-200 and my 3000-300X,
though. Too bad DEC doesn't have the ultimate useless feature: cross-OS
clustering. Imagine a VMS VAX, Tru64 Alpha, and VMS Integrity
clustered. Completely pointless, but cool sounding.