21 Aug
2008
21 Aug
'08
2:23 p.m.
Hi,
to make use of my spice tool set was indeed the first idea, but I
dont have enough time yet so I didnt answer.
Its no problem, there are still models for tubes available and its
not to hard to make new models for the originally used tubes.
The schematic is quite simple, maybe you can remember the first
logic chips available where made in so called RTL (restistor
Transistor logic), which is exactly the same as the ASM do, only the
levels are shifted.
Because of the problems with the resistor networks, the so called
TTL chips are developed, which replaces some resistors by transistors.
I am shure there is enough information out. Once you have seen how
RTL works, you will easily find out how ASM works.
Sorry, that I am short?with time, but I am very interested on this
project and as I can spare some hours, i will join, if possible.
With best regards
Gerhard
21 Aug
21 Aug
6:25 p.m.
DI Gerhard Kreuzer wrote:
The schematic is quite simple, maybe you can
remember the first
logic chips available where made in so called RTL (restistor
Transistor logic), which is exactly the same as the ASM do, only the
levels are shifted.
Because of the problems with the resistor networks, the so called
TTL chips are developed, which replaces some resistors by transistors.
I am shure there is enough information out. Once you have seen how
RTL works, you will easily find out how ASM works.
(As much as I like to call the ASM logic "RVL" (Resistor-Vacuum-Logic)),
the (gate input) design of the ASM logic is not the same as that of RTL.
It is not simply a swap of transistors for vacuum tubes, they are
fundamentally different approaches to gate design, at least if we're
talking about the common families of RTL ICs (that I've seen internal
schematics for).
The ASM gates have a purely resistive summing point of the input resistors
and a bias resistor before feeding into a single non-linear device (tube grid).
Further it achieves different gate types by simple varying the (bias) resistor
value.
In RTL the summing point is multiple transistor collectors with a common load
resistor: each input has it's own saturating non-linear device to
"discretize" levels before feeding into the summing point.
(And only one gate type is targetted)
Consequently, the ASM circuit is far more finicky and tight in it's requirements.
22 Aug
22 Aug
1:43 a.m.
It is not simply a swap of transistors for vacuum
tubes, they are
fundamentally different approaches to gate design, at least if we're
talking about the common families of RTL ICs (that I've seen internal
schematics for).
Well, for one thing, isn't RTL generally bipolar junction transistors?
Vacuum tubes are operationally more like FETs than they are like BJTs.
(A BJT is current control, current output, whereas a FET or vacuum tube
is voltage control, current output.)
/~\ The ASCII der Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
4:28 a.m.
der Mouse wrote:
Yes, there is that distinction, but by the time you put both into full circuits
with current limiting resistors for the bipolar trans and bias resistors for
the tubes, where there is current there is voltage drop and where there is
voltage there is current flow, so I wouldn't characterise it as the primary
distinction.
The structural/configuration difference between the two is more significant, if
one looks at schematics of the two gate designs.
It is not simply a swap of transistors for vacuum
tubes, they are
fundamentally different approaches to gate design, at least if we're
talking about the common families of RTL ICs (that I've seen internal
schematics for).
Well, for one thing, isn't RTL generally bipolar junction transistors?
Vacuum tubes are operationally more like FETs than they are like BJTs.
(A BJT is current control, current output, whereas a FET or vacuum tube
is voltage control, current output.)
29 Aug
29 Aug
3 p.m.
Paul Koning wrote:
I found some work on the net (and did some of my own
based on that) on
Spice models for vacuum tubes. It's not a standard offering in Spice
packages but it's easy enough to define a model, and given a data
sheet with characteristics curves you can quickly match the model to
the data sheet.
(The circuit I'm modeling is being very uncooperative, but as far as I
can tell the tube models are not at fault in this.)
Spice might be a handy way to verify an existing schematic and to
discover how to tweak it for another tube and/or another plate
voltage.
DI Gerhard Kreuzer wrote:
to make use of my spice tool set was indeed the
first idea, but I
dont have enough time yet so I didnt answer.
Its no problem, there are still models for tubes available and its
not to hard to make new models for the originally used tubes.
...
Sorry, that I am short?with time, but I am very interested on this
project and as I can spare some hours, i will join, if possible.
It could be interesting if someone made up
a SPICE model of the ASM to check some resistor values. I have no experience
using SPICE and am not acquainted with how flexible it is, but it would then be
interesting to exercise the model to find out how reliable the design would be,
that is, what degree of resistance and tube variation is tolerable.
I have the set of resistor values from the appendix of "The First Electronic
Computer", as well as a set of values from my own attempt at calculating them.
Either or both might provide a starting point for input to the model.
...sheesh, I just went looking for those values, found them .. but I also found
a spreadsheet model of the entire ASM that I forgot I had created. I
measured the characteristics of an actual 6C8 in a simple circuit and used the
resultant values in a lookup table in the spreadsheet to mimic the transfer
function of the tubes. The model, at least, appears to work, but my R-network voltage
calculations were rather simplified.
I'm sure a SPICE model would be much better, and may show up problems.
5878
days inactive
5886
days old
4 comments
3 participants
participants (3)
-
gerhard.kreuzer＠liftoff.at -
hilpert＠cs.ubc.ca -
mouse＠Rodents-Montreal.ORG