City #05 21 марта 1999

Железо

Iron - is offered to the wiring diagram megabyte SIMMa as RAM to Spektpumu.

          Saransk. Republic of Mordovia

 Free computer magazine for ZX SPECTRUM

         (Hm. .. Free? And do not expect)



                # 5 of 22.03.99g



--------------------------------------------------
                PLANET SHELEZYAKA
--------------------------------------------------

  As promised last issue of the newspaper, I suggest
wiring diagram megabyte SIMMa as RAM
to the Spectrum. Article courtesy of AP Petukhov of PHT 
g.Kovrov. 

(C) AXLR 1998

Wiring diagram 1Mb 30pin SIMM to computers such as

          "Pentagon" ("Leningrad")


         LL1 LL1 KP11

A1 1 1 3 4 1 6 3 D0 A0
/ IORQ May 2 LL1 D1 6 A1

                            D2 10 A2 Y0 4
A13 LL1 9 Jan. 8, 1913 A3

                     10 Y1 7
/ WR 13 January 1911 D5 2 B0
DD61: 5 January 1912 D6 B1 Y2 9

                            D7 11 B2

                      DD62: 14 September 1912 B3 Y3

          PROFI

         S1 / bit2 # EFF7 15 ° OE 1K

           / 1 S

         PENTAGON +5 B

                           1K


                            +5 B

                                     IR22


                                  3 D0 Q0 2

                                  4 D1 Q1 5

                                  7 D2 Q2 6

                  D0 .. D7 with Z80 8 D3 Q3 9

                                 13 D4 Q4 1912

                                 14 D5 Q5 1915

                                 17 D6 Q6 16

                                 18 D7 Q7 1919


                        DD51: Jan. 4 ° OE

    DD61: 1 +5 B 11 C

                        by DD18: 5

     TM8 LE1 KP12

 D0 Q0 ° 4 3 2 1 ° January 1915 ° SB

                3
 5 D2 10 B0

       Q1 ° 6 B14 11 B1
12 D3 June 1 ° April 1912 B2 YB ° 9

                5 A8 13 B3
 9 C Q2 ° 11

  1 ° R LE1 6 A0

                      B15 5 A1

               January 11 ° 13 4 A2

               March 12 A9 A3


          LI1 1 ° SA YA ° 7


   / RES 'C14 14 S0

   / C23 2 S1

   S2

                SIMM 30pin-1Mb


            5 In 1 Vcc

         DD45: June 2 / CAS

                 3 D0

         DD18: April 4 A0

         DD18: May 7 A1

                  6 D1

         DD18: July 9 A2

         DD18: August 12 A3

          total 9 Gnd

                 10 D2

         DD19: April 1911 A4

         DD19: July 12 A5

                 13 D3

         DD19: September 1914 A6

         DD19: December 15 A7

                 16 D4

                17 A8

                18 A9

         nothing 19 A10

                 20 D5

         DD51: April 1921 / we

         total 22 Gnd

                 23 D6

         nothing 24 NC

                 25 D7

         nothing 26 ParOut

         DD45: 27 March / RAS

         personal 28 / ParCAS

         nothing 29 ParIn

          +5 B 30 Vcc


         D0, D1, D2, D3, D4, D5, D6, D7 to


         3, 4, 7, 8,13,14,17,18 feet

           respectively.

DD64: a file for DD18: 11
by DD16: 5 and DD16: 11 apply common.
Swap: DD18: 2 (A8) and DD19: 13 (C34),

                  DD16: 2 (B11) and DD17: 13 (C35).
Old signals from the legs to go new - cut.
All Products Series 1533. Possible but not recommended for 
replacement. Instead 1533KP12 can put 1533KP2. Diode VD -

any low-power, for example KD522.
Memory chips RU5 (7) must be removed from the board
or block, if the power of BP allows them
leave). If you have any problems with regeneration (the 
Pentagon after all) can make such a thing: to IR22 (see above) 
to signal to the DD51: 4 passed through the element LI1:


                      LI1

             K43
DD51: 4 'to 1 foot to IR22
                                 treatment (see above)

                 ===

                    10 .. 100pF


  Mikrik S2 (and LI1) must be set if you are going to the 
hardware oklyuchat upper memory area. And sometimes it is 
enough, and S1, is if the prog does not understand all 
standards for extended memory.  Vit2 port # EFF7-a software 
lock memory above 128K.Esli you have in a computer, this port 
is absent, Instead, the signal was fed to a common wire.


  Signals the address and data bus of the processor are marked
as A0 .. A15 and D0 .. D7, respectively (be careful!).

  Still need to remove the lock memory above 48K.
Removed is simple: by DD64: 11 cut bit5
# 7FFD and this DD64: 1911 served a common wire.

  I have this 48y latches do not have one and a half years and 
no once there was no problem with it. True, I've seen several 
old 48x toys that do not work without it. If some of these toys 
as luck would have turned out to be your favorite:), do not 
worry - it's solved too easily. It can be used in

As S2 dual mikrik. Then it will be as follows:

               S2.2

                 /

      DD62: 15 / in DD64: 11


                     1K


Or replace the item mikrik log.I:

                LI1

DD62: January 15 '3 in DD64: 11
bit2 # EFF7 2


  Now when you reset bita2 port # EFF7 zero memory
above 128 will be opened and 48ya latch-lock. And when you 
install it in the unit will be closed access to memory above 
128K and 48ya latch can be used. 

  If the comp is a CMOS-hours and installed flash ROM "MadRom", 
then all these mikriki can be removed and make software 
management, from CMOS_SETUP. 

  For details, see help for the firmware.


  For the "Leningrad":
Instead of B14 at KP12 serve with DD19: 4
instead of B15 at KP12 serve with DD19: 7
instead of C14 on feeding KP12 / RAS with DD2: 9
instead of C23 on feeding KP12 / H1 with DD34: 2
instead of DD61: 1 Alert with 1st leg KP11 (revision to 128K)
instead of DD62: 9 lodge signal 9th ​​leg TM9 (revision to 128K)
instead of DD51: 4 to signal with DD9: 8
instead of DD18: May apply for DD15: 3
instead of DD64: a file with the 12th leg KP11 (revision to
128K)
 by DD15: 6
instead of DD16: 5 and DD16: 11 to apply for land DD18: 3 and
DD18: 6
 * On 15yu leg KP11 (revision to 128K) is recommended
apply the land, freeing it (leg) from the pre-
H1
 * With the 9th and 12th legs of the same KP11 recommended to 
remove resistors.


  Must still make a revision:

       LL1 LI1

                                  usually such eleA14 January 1 
'3 mA DD9: 13 element LL1 can A15-2 found in control

                                  Leray drive
DD34: 2 (/ H1)

  Refinement of this is needed to program, writing in
domain addresses # 0000 .. 3fff not spoil 128yu memory.

  Data bus SIMM apply for a DD32, address bus for
SIMM taken with the same foot former RU5 (7) - is
legs 5,6,7,9,10,11,12,13
/ RAS SIMM to take DD2: 9
/ CAS takes something like this:

                            LN1


              DD3: 3 (H0) January 1 ° 2 on the 2nd pin SIMM


  As the inverter can juzat remaining element LE1. And KP11-x 
swap does not need anything, in Leningrad, such zamorochek with 
no regeneration. 

  Leningradovskaya refinement "memory expansion up to
128k. "There is a view it is one in which management CAS'ami by 
using KP11 and LA3. 

   In principle, this is the analogue of the Pentagon scheme
128k management, but L-ray used Z-state KP11.

  If on finalizing the Leningrad-clone made
different scheme, you should understand it.

  As can be seen, the basis of this shemki - Revised Scheme
Pentagon-128 to 512K by V.M.G (from Spectrofon-17).
If SIMM is placed on the 512y complex, most likely,
need to add only a small number
chips and do not need to regenerate
permutation.

  As the expansion ports are used to 1Mb
Profi-port (# DFFD, mask # 07) and the Pentagon-1024 (# 7FFD,
mask # 07).

  Instead, 1Mb 30pin SIMM can be inserted 2Mb-30pin
(Tested). Only need to apply for land 19y pin
SIMM'a (A19). Of course, access is possible
Only the youngest of megabytes.

  For a similar system can be connected and 4Mb30pin SIMM (put 
another TM8, KP12 and LE1), the problem only in the absence of 
real ports and software for this memory. However, in the STS6.2 
and ISDOS possible memory use under any ports, would

driver.

                   *
 I can ask the most tricky in such cases
questions (and ask): yes Why do you need this MB? In general, 
soft under it have? 

  Yeest! with regard to pros, it's a different story, but for a 
megabyte at # 7FFD most "home" cFosSpeed ​​- TRDOS5.12f by 
Reanimator. For those who have So far in the tank, to remind:

This Doce instead of drive "D" is formatted RamDisk
at 892k!

  All prog using to access the disk
# 3D13 working on ramdiske much faster. Naturally, the program 
should not spoil 512yu memory. 

  I personally quite enough of that to (bystro!)
Tossa-mail with a real drive to virtualnyy.Ili,
say, ZIP ... Or assemble with podgruzki.

  A lot more then ...

  Numbering chips given for the scheme, the Pentagon,
in which the address signals of the processor are marked
as A1-A16. A scheme to Leningrad written "Drafted
in Kharkov ".
P.S. At Leningrad shemka works with mid-summer,
at the Pentagon, somewhere in late summer, and to date
the moment everything is ok:)