> -----Original Message-----
> From: cctalk-bounces at classiccmp.org
> [mailto:cctalk-bounces at classiccmp.org] On Behalf Of Holger Veit
> Sent: donderdag 3 november 2005 14:08
> To: On-Topic Posts Only
> Subject: Re: FPGA VAX update
>
> woodelf wrote:
>
> > Bj?rn Vermo wrote:
> >
> >>
> >> Apple comes to mind. You can pick up a G4 really cheap these days.
> >> Whether you want to run BSD (OS X) or Linux is a matter of taste.
> >> IBM have some PPc development systems (Walnut if I recall right)
> >> which can run both Linux and QNX, but I have no idea how
> to get one
> >> or what they sell for.
> >
> >
> > I guess nobody even considers building a computer from scratch. :(
> >
> About thirty years ago, when I was pretty much younger :-), I
> built a classical TTL computer (with 74181 ALUs etc.) almost
> from scratch, basically from schematic fragments from the TTL
> databook and TTL cookbook and some electronics magazines;
> well - if it were really from scratch, then even so "highly
> integrated chips" like the 74181 were prohibited as well.
> Looking back, this had almost all characteristics of a real computer.
>
> There are two aspects IMHO why this does no longer happen today:
> - what was a known magazine in the past, would today perhaps
> realistically called "Un-Popular Electronics", the knowledge
> of electronics from the ground up is dying out; and with the
> continuing "digitalization" of technology, it is an ever
> increasing hurdle to get started - the classical AM detector
> radio I built as a newbie will nowadays no longer attract
> anyone - you can get a gadget which is better by several
> magnitudes for a fraction of the expenses you'd have for
> soldering your thing.
> - The tools you have are too user friendly (!); i.e. you
> could rather easily click something together, be it software
> or VHDL code for an FPGA without ever needing to understand
> what is really going on. The "soul of a new machine" guys
> that traced glitches with a logic analyzer in a large
> wire-wrapped TTL graveyard are gone - such a machine had the
> necessary wow factor. No surprise when a VAX in an FPGA - see
> subject - is not really interesting. It is possible - that's
> it. But what does one gain? It is like solving a 10000 pieces
> puzzle; spend time and don't learn really much.
> When Hillary climbed the Mount Everest 50 years ago, it was
> something new, extraordinary. When tourist nowadays use sort
> of stairways to reach the top (okay, it's not *that* extreme
> now), it is just uninteresting.
> But standing on the top, is still a challenge and an
> experience for the individual. However, what we lost now is a
> sense of wonder - that small ALU based TTL circuit was built
> by myself, and I did it, and it is irrelevant that some idiot
> could download a digital simulator and click the same circuit
> together on a PC screen. I guess this is what makes the
> difference between a real PDP-11 with some ridiculous 5MB
> storage disks compared to a SIMH emulator running on a 3GHz Pentium.
> Watch the blinkenlights.
>
> Holger
Actually, I don't need to comment on this. Being a HAM radio amateur,
I did built a diode detector with a high-impedance headphone and a
varco plus coil. The experience *I made that*, and the first time
you hear a radio station are unforgettable impressions.
To add to your last sentence: I agree, but I have built something
that sits "in between". I use the SIMH software, but *added* a real
blinkenlight console! See my website. I do have several real PDP-11
models, but the blinkenlight SIMH is fun too (well, sort of).
Perhaps most of the fun was building it ...
- Henk, PA8PDP.
>
>
This message and attachment(s) are intended solely for the use of the addressee and may contain information that is privileged, confidential or otherwise exempt from disclosure under applicable law.
If you are not the intended recipient or agent thereof responsible for delivering this message to the intended recipient, you are hereby notified that any dissemination, distribution, or copying of this communication is strictly prohibited.
If you have received this communication in error, please notify the sender immediately by telephone and with a "reply" message.
Thank you for your cooperation.
>
>Subject: Re: FPGA VAX update, now DIY TTL computers
> From: "Chuck Guzis" <cclist at sydex.com>
> Date: Wed, 02 Nov 2005 22:41:42 -0800
> To: cctalk at classiccmp.org
>
>On 11/3/2005 at 7:00 AM Jos Dreesen wrote:
>
>>Much too big to actually build in TTL.
>
>Consider the very old Packard Bell PB250--22 bit words, fewer than 400
>transistors and 2500 diodes, 63 instructions. Power consumption about 40
>watts, exclusive of I/O:
>
>http://ed-thelen.org/comp-hist/BRL61-p.html
>
>The trick, of course, is to use bit-serial methods. It seems to me that
>one could greatly simplify construction of a homebrew machine that way.
>We're not doing this for speed, right?
>
>Cheers,
>Chuck
As they say in some parts "it depends". ;)
I can think of three reasons to build a machine.
Recreate a system that is scarce or unobtainium such as the Apollo Block
One. Speed here is possibly important to understanding what it's limitiations
or capabilities were.
Experiment with an archetecture that may exist but is not a chip. Some speed
would be nice here as you may end up running real code.
Or just to build a computer that one can completely understand why and how
it works. Here speed is not essential, slow may aid is seeing all the bits
work in real time.
It would be hard to beat commercial hardware. However there is little
reason for not pushing the logic some if you plan to run an OS on it.
At the other extreme the ability to go glacially slow is a good debug
tool (all them lights don't hurt either).
Allison
>
>Subject: Re: FPGA VAX update, now DIY TTL computers
> From: Jos Dreesen <jos.mar at bluewin.ch>
> Date: Thu, 03 Nov 2005 07:00:20 +0100
> To: "General Discussion: On-Topic and Off-Topic Posts" <cctalk at classiccmp.org>
>
>
>>
>> What surprises me about the home-built computers from scratch is the
>> lack
>> of imagination when it comes to architecture. Most are basically
>> one-address-cum-accumulator designs. It would seem that larger
>> register
>> files are much easier to build nowadays and would open up the door to
>> some
>> 2 and 3-address designs.
>
>The reason is the effort it takes......
>Ask anybody who actually made a CPU ( I myself did a 12 bit
> TTL based, single address single accumulator machine).
>The real effort is actually building and debugging the unit.
> There is only so much time you have....
>>
>> I've always thought that 24 bits is a nice word size for a small
>> computer.
>
>Much too big to actually build in TTL.
>
>
>In 1974 Elektor magazine started a series of DIY articles that described
>an extensible ( 12 bit or 16 bit width ) three address, clockless
>computer.
>TTL based, with a shiftregister based memory. (6 or 8 512x2
>shiftregisters .)
>An expensive extra wasa 2102 based RAM.
>The machine did hardware multiply and divide +
>Instructions were build into the address map .i,e, address XX is the
>adder, adress yy is a shift and so on..
>
>The start of this series actually predates the Mark8 computer !
>
>
> Jos Dreesen
With current generation MSI TTL functions the highest parts consuming
section the ALU is reduced to a few peices. Registers with tristate
outputs are easy now. Those things were not easy to find back in 1973-4.
The whole point and problem solution of serial machines was flipflops
were 6 trasistors each and a logic gate could be one transistor and
many diodes. Now a 4 bit multifuction arithmetic slice is 74181
(transistor and diode equivelent may have numberd in hundreds).
Another area that was tedious to the max was memory. In 1974 the 2102
was a over 10 dollar part and you needed 8 for a byte parallel ram.
By 1980 1kx8 parts could be had for 3 dollars. Now 32kx8 parts are
cheap with far larger available. So building a memory system is
simplified as well.
So building a larger machine is not as daunting a task as in 1974 or even
1980. Even without FPGAs.
Allison
This may be appropriate for another list, but it seems to me that there's
plenty of applicable knowledge here.
Right now, we're using an older Compaq Deskpro P3-600 box as our incoming
Linux server. Basically, it connects with the DSL modem and contains IP
masquerading, DNS caching, firewall and fetchmail/procmail/qmail tasks
(SpamAssassin included). It runs 24x7 with no problem, as it has for the
last 5 or 6 years. Built like a tank.
I'm wondering if something a little less power-hungry might be preferable,
however. Are there any low-power alternatives for the job that anyone can
recommend?
Cheers,
Chuck
Tony wrote:
> Do you have a logic analyser? If so, then I would try looking
> at the 8080 address bus. See what bit of the program it's
> executing, see if it does the right things when you press the
> keys, etc.
Yes, I have a Biomation D250 I got from Edward recently.
However, I think I am not yet out of all options to pull out the
"big gun". First I must hjave a look at the printset to see how
the keyboard scanning works, but I suspect that the signals are
generated by the 8008. Makes sense to me ...
If I am lost, I will pull the D250 out (and the manual, because
I will need to study/learn how to use the beast).
> If you don't have a logic analyser, a trick that's helped
> many times before is to connect an n-bit comparator to the
> address bus, with the otuer input from switches. You can use
> that to see if the processor ever accesses a particular
> address (e.g. to see if a particular routine is being executed).
I used that method to detect why my homebrew 68010 did not work.
I was used to 680x devices, so I programmed the 68010 software to
set up the stackpointer at the highest RAM location. Big mistake!
It should have been highest minus 1 (even address aligmnent!),
but that's another story. Btw, the 68010 works fine :-)
> I am pretty sure it doesn't do UNIBUS accesses when you type
> in a number.
> Firstly, it doesn't know when you've finished the number
> (until you press onoe of the other keys), and secondly, the
> number you type in may well not be data for the current
> addresss -- it might be a new address, or a value to put into
> the Switch Register.
Yes, that is what I was thinking too.
> I wonder, though, if the thiog needs signals from the arbiter
> to work properly. I can't see why it would, but maybe if one
> of the bus lines is in the 'wrong' state it gets very confused.
Ha, this sounds interesting. Although I removed most of the cards
out of the backplane, there are still a few that could hurt if
the arbiter fails, because of wrong bus lines IF they play in
this scheme. Again, I must dive deeper into the M7859 diagram.
thanks Tony!
- Henk.
This message and attachment(s) are intended solely for the use of the addressee and may contain information that is privileged, confidential or otherwise exempt from disclosure under applicable law.
If you are not the intended recipient or agent thereof responsible for delivering this message to the intended recipient, you are hereby notified that any dissemination, distribution, or copying of this communication is strictly prohibited.
If you have received this communication in error, please notify the sender immediately by telephone and with a "reply" message.
Thank you for your cooperation.
I was getting a Soroc IQ-135 ready for VCF and noticed the "mold" on the
outside edges of the screen between the CRT and the outside glass
shield. Someone on the list wanted a sample (can't remember who) to do
an analysis to find out if it was indeed mold. I now have samples if
there is anyone who can do an analysis of the stuff.
In looking at it, my guess is that it is not a function of something
living (plant or otherwise) but rather a crystallization taking place in
what appears to be a silicon based substance (RTV?) It looks like it
would be relatively easy albeit time consuming to separate the CRT from
the cover glass. Does anyone know what material was originally used?
I was reminiscing to someone about an old GE mainframe with an analog meter
that registered Kops/sec. I'm sure it was little more than an RC
integrator hooked to the RNI signal reading out on a voltmeter, but it
seemed pretty spiffy at the time.
Has anyone done a similar thing with their vintage systems? On the 8080,
the M1 signal could be used; I'm not sure offhand about other processors.
I know--it's more simple amusement for a simple mind...but then, blinking
lights always fascinated me.
Cheers,
Chuck
> A better idea is to place a current-limiting resistor in
> series with the alligator clip to ground
Not really. You'd need to use an HV resistor, it's unlikely
that anything else will survive discharging a tube without
being damaged. Not much good when you go to use it a second
time.
> Not using the resistor will still discharge the tube, but
> it runs the risk of damaging it.
Not really, any tube can stand very many such discharges
without harm.
Lee.
..
___________________________________________________________
How much free photo storage do you get? Store your holiday
snaps for FREE with Yahoo! Photos http://uk.photos.yahoo.com
> Can anyone reassure me that this is a semi-safe procedure?
It's safe, much safer than not doing it.
> Has anyone done this sort of thing before?
Lots of times. I connect the croc clip directly to the aquadag
connection, usually an uninsulated wire or spring stretched
across the back of the tube, this is the other end of the
capacitor formed by the tube.
You may not hear any discharge, there may be none, but poke the
end of the screwdriver into the hole that the HT cap clips into
and keep it there until you lift the cap and visually confirm
that it has connected.
I did a load of scrap monitors like this recently and the only
ones that had any residual charge were the ones that I'd powered
that day to see if they were worth saving.
Lee.
..
___________________________________________________________
Yahoo! Messenger - NEW crystal clear PC to PC calling worldwide with voicemail http://uk.messenger.yahoo.com
Okay, I'm going to try and take another crack at replacing the flyback
on my VT100. A few months ago I almost did, and chickened out. I'm
very very afraid to do so, as the idea of trying to discharge the CRT
scares me very much. The terminal has been off and disconnected from
power for at least 2-3 months now.
This is what I've learned so far from reading:
Put one hand in your back pocket. Use an insulated screwdriver with
alligator clips on either end, one to the screwdriver blade, one to the
frame (VT100 service manual says the frame right above the CRT is where
you should ground to). Place the screwdriver under the anode cap and
touch the metal connection underneath. Wait till the crackling stops.
Remove the anode cap.
Can anyone reassure me that this is a semi-safe procedure? Has anyone
done this sort of thing before?
Thanks
Julian
