Za Rulem #00

Iron - fix incorrect wiring diagrams musical coprocessor AY.

I wondered how a question: why me
do not go to certain programs, games and demos
who are my friends? After
it was found that these shoals - not
defect rewrite, there was nothing left,
other than sinning on the computer.
 On my computer, you can tell in some detail. This was once a 
48kB Spectrum - one of the many clones, as said

Blast Off. It had AY-muzprotsessor,
but somehow clumsy connected.
 Where do I get it? Here:
 I did not have the Spectrum, and very much wanted.
 I learned that the area is Kalinina
firm "meson" and that there is such a dude
Sergei Ogaltsov, which seems to be
to Speck. I phoned him and, in
general, ordered a 128K Spectrum. He said
that would be complex to be called "Forum". I did not care, 
just to work. In general, Ogaltsov remade 48kB of 128K.  After 
some two months, I finally had the opportunity to buy this 
creation Russian masters. The next day I

took him back because ALASM, loaded into the computer, did not 
respond to an attempt to put symbols. , *, A debugger at all

when you press any button on the published lusty cries and 
displays instead of the character set of pixels resembling that

anything but the desired character.
 The next day I again took the Spectrum
home - now it was the original firmware 128kb Speck. However, 
in this keyboard has ceased to be expanded, not

Russian has become embedded font. But who is he
they need?
 Now I could steer at full speed, but
than dolshe I taxied, the more discovered
programs that I did not go.
 Initially it was part of "a bottle" in
LYRA II MEGADEMO. It has long been the only prog that dumps 
after load, but when it became to Novosibirsk

reach every nove, challenges me to score.
 Journal Funeral 1 1 / 2 of Delirium Tremens,
Melan # 01 (of the same newspaper) - cbrasyvalis at
output image table rabchego, Netwalk from
FaceOff reset after the first level, and in Nenwalk from the 
Style Group is not were "authors", after the inscription "in" 
hung demo Ecstatic, and very original -

music was played to the end and the white
strip of curb in the area of ​​the editor, the message "out of 
memory". Anamnesis discharged after the first scan

brain, Workstop fell out in the 48 basic
with a glowing inscription "stop", Shibentsiya
after the first levels are set across the screen attributes 
ink7, paper 7 and when you try to guess the next mercilessly

hung. Wormworld bismuth after selecting "start
game ", a demo of Funtasy Oops cbrasyvalas the third rule of 
demo-making organization.  Now I know that this collection of 
bend those who work in the mode IM2 without constructing a 
table in memory of 256 bytes, which is the address of the 
treatment program interrupts.

 At first I thought that I had some
RUshka "broken." But by writing a small files of it was found 
that "Che's memory put that otteda and take. "

 In some book I read that in a loop
confirm the termination of the bus must contain B # FF, because 
at this point it All devices must "let go", and since

it is "pull" to +5 v.
 I decided that my bus did not pull in
5, and to understand complex. Digging in the ligaments
MGTF'a I found that the bus to pull 5 and I have a competent 
ruler of the eight RU7, ie 256kb physically. Yet it was evident

Ogaltsov that delved into the connection AY.
Oh well.
 I began to experience a different comp tests. But all of them 
except the dough and ESE1.4 test before the game Sq.Head, 
stated that data bus I is stable. How I hacked later, test 
Sq.Head gave that interrupt vector = # 00 because it just

reads this number to 255 port - port attributes, and assured 
that I have but there is this port. (In this test, it appears 
that in a computer can not be both a vector and # FF port 
attributes).  So I remained ignorant about the parameters

my computer, yet, I have not got TEST4.22
- This is the coolest test that I know
until now. Here he is just what I gave: "Bus
data is unstable, in a series confirm
peryvaniya clogged, possibly wrong
connecting muzprotsessora.
 I immediately dismantled his computer and pulled
AY from the socket. On a miracle - anything that I listed above 
- went, and how!  Two views could not be - I have not properly 
connected AY. We had to do something. First I drew a diagram on 
which it is connected. Here it is:



             360 Ohm

                           / WR 1

        / IORQ 1 of 27

                      about 1O *

               1o


                                 1
A1 1 of 29

           o A14 1o
A15 1o


 In general, do not want problems - do not collect
so ever.
 We just recently got acquainted with Bones
aka DEBOSH and I decided to drive for the scheme
AY connection to him - I saw him
two whole disc with modifications to the Spectrum.
 That such a scheme has brought me from him:



                           / WR 1

                                     about 27

                     / IORQ

                             1

            A1 1

                         *


  A15 1

                             1

  A13

                  / M1 1O

                                   1

                                     of 29

                       A14 1O


 In this scheme has been analyzed and
And A13 / M1, and I bought into it, besides,
this scheme was not hard to make out already
existing one. I picked up a scalpel and a soldering iron.
 AY silent, like the dead. The test does not find him, and the 
author of the article and this scheme assured so that all 
connected AY them in the city.  In short, I decided to collect 
the same scheme, as a Scorpion 256 Turbo + 1996.

 I must say that this scheme worked,
data bus is stable, and vaasche
Everything was OK That this scheme, if someone does not
know:


                                    n

   A0 1 A0 DC / WR 1 A0 DC o11 p

   A1 2 A1 A14 2 A1 of

   A5 3 3 A2 A2 A15 A N

                                o9 o and

  / M1 E1 +5 v 6 6 E1 where
IORGE 5oE2 GND 5oE2 g e

       4oE3 o10 4oE3 o7 e

                                    n

                                    and

                                    e
DOS 9 '12'

         o8 o11
A1 October 13


       n FOOT

       p AY8910 AY8912

       about 1 '

       Dr. o3 -> 27 -> 18

       o 2

       L -> 28 -> 19

       x 4 '

       e o6 -> 29 -> 20

       H 5

       and

       e

 You can use only 3 chip
If you can find the right signals in his
card companies. I did a fourth output of the first
ID7 grounded, so that the two elements and is not
remains free, and the signal from IORQE
I - / IORQ, enabled through a resistor
360 ohms. Just a 28 foot AY8910 I just
filed +5 v, and then suddenly sampling will not be
play. The scheme of Scorpio on the sixth leg
first decoder comes M1, but izlazav whole scheme, I did not 
find the item which inverts / M1, must be a typo.


 And yet. On AY, I hang such potentials:
AY8910 AY8912 potency.

  40 3 +5 v

  25 17 +5 v

  28 19 +5 v

  June 1 GND

  24 did not know his leg GND
One 5 and one GND - this is food, and the rest - I do not know.

DC-decoders ID7 555.1533 series. Elements and is not - 555LA3.

Eighth leg ID7 and seventh leg LA3 - GND
16th leg ID7 and 14th leg LA3 - +5 v

P.S. Big Fuck to: GLOB.TEST and TEST IM2.
The first tests the stability of the long bus
and provides that it is stable, although in reality it is not. 
Second, even contains a description of the test itself, and

quite competent, and entitled to
lives separately from the program, because
program does not work. Especially after
reconnect AY and half tested
bus - not got any bit unstable.
Interestingly, these tests are generally tested for
unstable company or not? And it
were at the computer (or as ugly written
programa), if the defect was detected Bus
Only after "warming up" your computer?? On
pisyuke, or what? From under the emulator? Maybe
to setup the emulator even has a point "A stable data bus + / -?