City #42 25 марта 2000

Железо

Iron - Cashe memory scheme for the Spectrum.

(C) V. Kazakov


  Today I bring to your attention one of shemok cache, sent me 
a VTS, which in turn took it from the CD M.M.A 'Maxx'a. 

  The author, unfortunately I do not know.

  Text and diagrams bring no change in the original.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~

  So much for the scheme cashe:

                    KD522

  TURBO>


  A14 1

  A15


  / RES oR T BLOCK_ROM

  A7 D

          10k / C 18 Jan. pin.

  +5 V o oS oo PN10 / CS

    Magic / Magic +

  / NMI /

                                          K1533LL1

  A2 January 1 K1533TM2

  / RD


  / IORQ

  / MREQ 20 pin. PN10 / OE

  / WR pin 21. PN10 / WR


  BLOCK_ROM signal goes to 20 or 22 output memory (if they have 
not instituted the "land") through an OR element, mixed with 
the old signal, which went to the ROM. 

  Address inputs and data bus connected to PN10
corresponding signals on the processor.

  If not TURBO, the diode can be left.

  Button Magic / Magic + allows vyvalivatsya of Magic is not in 
TR-DOS, and the program, which is in the cache. 

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~

  I draw your attention to our loyal readers
error in the scheme of hours Mr.Gluk 'and that
was cited CITY # 30. The scheme signals A14 and
A13, going to the elements LE1 be interchanged.
Also, the clock can go unevenly due to the fact that
printed guides, walking on the board of the data bus, divorced 
too close to the oscillator clock. 

  Noticed these errors Alexey Volyk / Phantom Family,
from which I just recently got a letter.
In addition, Alex is sent to clock driver circuit
Mr. Gluk'a for IS-DOS. This driver, along with his
sources to Alasm'e, as well as CMOS Commander 0.8
for setting hours, I decided to give a supplement to city42.z.
PS. According to Alexey, driver hours can be a bit
fail when passed to a system of due date ..