From: Bil Herd (bherd_at_ids-business.com)
Date: 2007-04-12 18:04:19
I don't know the context but it's kind of tough to make TTL OR gates
with diodes. You can't use them to pull low as the voltage drop across
them is greater than Vin (plus whatever you pulled low with). If you
use them to pull high then you will need a pull down resistor (for when
neither diode is pulling high) which has it's own problems. For example
a 1K pulldown resistor would require that the device driving it be able
to source 2.4 - almost 4 ma depending how how you interpret the stress a
chip sees... check the device driving it, you may find it's limited to
microamps) Unless your in a special case and have made a lot of
consideration (in other words I have done squirrelly things such as the
Z80 clock circuit and the monitor sync combiner on the 128) but only
after looking carefully at all of the specs.
You also have to look at sequential delays when stacking logic together,
not only for when you are trying to create a true condition but what
happens when the true condition starts to fall apart, I.E. do you create
lots of transitions on a cs line during intermediate conditions. (Some
applications tolerate this some hate it)
Recommend using a logic module that encompasses the state table of
whatever you are trying to do: unused cs lines can create simultaneous
And or Not-OR functions and pieces of chips like LS139's can accomplish
some small state tables when you combine the decode logic and the enable
logic. Some of the fun of design is trying to match a state table need
with a way to accomplish it. If it's easy to cut a pin on an existing
chip that is otherwise grounded such as an unused CS line, you might
find an easy way to qualify a condition with existing hardware.
If you follow the diodes with a transistor you can get away with some
things but lose some things (time and inversion plus RC components start
to come into play).
Sorry for the ramble, just haven't thought of this stuff in many years.
Bil
-----Original Message-----
From: owner-cbm-hackers@ling.gu.se [mailto:owner-cbm-hackers@ling.gu.se]
On Behalf Of Antitrack@networld.at
Sent: Thursday, April 12, 2007 9:10 AM
To: cbm-hackers@ling.gu.se
Subject: RE: 8kernal/8modules for c64-II ? OR-Gate with diodes
Zitat von ruud.baltissen@abp.nl:
> Hallo ATT,
>
>
> > Okay. Let me think about this again, Ruud:
> > IF A13=1 then Memorywanted=somewhere between e000-ffff (KernalROM)
> > ELSEIF A13=0 then Memorywanted=somewhere between a000-bfff
> > (BasicROM) ?
>
> Yep.
>
>
> > So if A13=0 then I must Chipselect. This is not an OR gate,
> > it is an AND gate, isn't it?
>
> No. Write down the 0/1-tables for both and you'll see that you need
the OR
> gate.
You are right, it is an OR gate that I need.
Basically, I'll take the A13 line coming out of the kernal socket,
and OR it together with the Chipselect-Line which also comes from
the kernal socket, and I feed the result of the OR-gate into the
STROBE-line of the 74LS157. So the resulting three A15, A14, and A13
lines
for the 27c1001 eprom will all be at 0. A16 will be 1, because A16 is
connected to the Chipselect-Line from Pin11 of the module port.
Now, some old electrician friend told me that I can use diodes (1N1418,
is this the right diode to be used ?) to substitute for an OR gate, but
I
cannot remember in what direction the diode must be placed...any hint?
Yours
ATT
--------------------------------------
Ein Service von http://www.networld.at
Message was sent through the cbm-hackers mailing list
Message was sent through the cbm-hackers mailing list
Archive generated by hypermail pre-2.1.8.