1.44 0n 3.5 drive

One day on December 21, 98 at exactly 00:49:53 Pavel Fedin
wrote to Oleg Sazonov on the topic '1.44 0n 3.5 drive':

PF> The hole responsible for the HD-ness of the floppy disk is sealed. That's
PF> all.

but only, the disks will be 720 kb...

With best wishes, Dmitry.

Not seeing a moderator nearby, I decided to insert a word:

RM> The turbo mode on the computer must be constantly turned on! It turns off only during
RM> the time
RM> of disk spin-up (it turns on later by the index pulse). And there should be
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Into the garden. An extremely complex scheme. Let the program itself D00Mate.

RM> done
RM> brake - lengthening /IORQ to normal duration, otherwise the VGshka will not
RM> have time to send data.

??? This doesn't happen. But #FFs from VGshka are read with a short iorq.
It turns out that the bits are bent.

RM> There is no software for this yet, but it will be a promising idea.

For isdos?

With best wishes, Kirill Frolov.

PF>> The hole responsible for the HD-ness of the floppy disk is sealed. That's all.

DS> but only, the disks will be 720 kb...

Yes. But AFAIK, our controller is completely unable to read 1.44 MB.

Bye, with a bunch of vichez, Pavel Fedin.

... UI125:/0EPJFAGAECCVNGLFHGiGDGFFFPAFfBCKBRRDQCBKHIEBCCKC

■ 13 Dec 98 at 19:03:00 Oleg Sazonov addressed to
All the following words:

[...]
OS> Tell me, pls, a scheme for finalizing the spec so that
OS> can be used on a 3.5 drive
OS> 1.44 disks kb.

there is, but only for the Pentagon, because its circuit can be found
in any pack of waste paper :)

2ALL: However, if they give me any other circuit
of the controller, there will be a subcircuit for it.

a turbo mode is used, if it’s not there, we fly by... or
set the turbo.

with support I slow everything down ;( .
But never mind, I’ll soon finish assembling a new computer board and I’ll get to work on it.

2Vladimir_Klymus: You asked who needs the “MS#” source code.
Throw it in. Let’s try to make msdos-1.44 in it too. Of course,
the sources will then return to the author. Or, maybe, on their own? ─────═════════════╗

(c)1998 AXLR
──────────────────────────────── ───────────────────────────

So, what needs to be done to make a regular controller
work with high-density disks? If the computer already
has a turbo mode installed, then all the modifications take literally
an hour, and if not, well, here we offer an already modified
turbo circuit forPentagon-assemble.
So, even if you already have a turbine, you will have
to finish it a little. Actually, there are two improvements here: one for
turning on the turbo when DOS is running, the other generates /wait
when the processor accesses the ports.
Special mention should be made about the latter. It has been noticed that some
KR1818VG93 copies, namely 1993 and later
years of production, do not work well in the turbo. and the quality of exchange with
the disk does not seem to be affected (they haven’t tested it specifically),
but who, VGshku, knows.

It would be best to provide a complete turbo diagram for the Penta
gon, but with modifications:

*100pF
LA3 ─┬─ TM2 free elements TM2 and
┌──┐ === ┌──┬──┐ LA3 can be found on board
D3:2 ─────│ &°───┴─────13° R│ │ company
D3:3 ─────│ │ ┌────12│ D│ _│
└──┘ │ ┌──11/ C│ Q°8─────> D5:2
D10:12 ───────────┘ │ ┌10° S│ │
D15:5 ──────────────┘ │ └──┴──┘
+5V

DD1 +5V
┌──┬──┐ │
D45:3(3.5MHz)─┬──────────────────────2│A0│ │ █K47
│ +5V ───5│A1│ │ │
│ ├──┤Y0│4──┴───>Z80:6
D2:9 (7MHz)──────────────────────┬───3│B0│ │ K10
│┌─────6│B1│Y1│7─███──>Z80:24
│ │ │ ├──┤ │
│ +5В DD3.1 ┌──────1│ S│ │
│ │ ┌──┬──┐ ││ │ ┌15°OE│ │
│ ├─1° R│ │ ││ │─┴─ └──┴──┘
│ ┌──2│ D│ _│ ││ │
└──────3/ C│ Q°6┘│ │
│└─4° S│ │ │ │
│ └──┴──┘ │ └───────────────┐
│ └───────────────┐ │
DD4.2 └──────────────────────────┐ │ │
┌──┐ │ │ │
/bit7#FF───4│ 1│6────┐ │ │ │
D76:9 ─────5│ │ │ │ │ │
└──┘ │ │ │ │
+5В DD5.1 │ DD4.3 "NORMAL" │ │ │
│ ┌──┬──┐ │ ┌──┐ S1 │ │ │
HLD──┬──1° R│ _│ └─13│ 1│11───── \ │ │ │
│├─2│ D│ Q°6────┬─12│ │ ┌─ \──────┘ │ │
/IP─────3/ C│ │ │ └──┘ │ │ │
│└─4° S│ │ │ DD6.1 │"TURBO" │ │
│ └──┴──┘ │ ┌──┐ │ │ │
│ └──1│ &│3────┘ │ │
└──────────────────2│ │ │ │
└──┘ │ │
│ │
/IORQ─┬──────────────┐ │ │
│ DD7 │ │ │
│ ┌──┬──┐ │ DD4.4 DD6.2 │ │
└14│ R│ 0│ │ ┌──┐ ┌──┐ │ │
+5В ──11°WR│ 1│ └─10│ 1│8 ───────4│ &│6──┘ │
┌─5/+1│ 2│6────── 9│ │ ┌─5│ │ │
│ ├──┤ 3│ └──┘ │ └──┘ │
│ │ │ │ │ │
│ └──┴──┘ │ │
└─────────────────────────────────────────┘
│
DD6.3 DD3.2 │
┌──┐ ┌──┬──┐ │
D13:26 ──10│ &│8─────13° R│ _│ │
D49:11 ── 9│ │┌─────12│ D│ Q°8──────┘
└──┘│ ┌───11/ C│ │
D13:28 ────────┘ │ ┌10° S│ │
D50:13 ──────────┘ │ └──┴──┘
+5B

Сигнал HLD берется с 28й ноги ВГ93, /IP - с ВГ93:35,
/IORQ - с Z80:20, а /bit7#FF - см. ниже.

Для турбы:

1. Отключить D5:2 и D13:28 (D5:13)
2. Отключить D15:1 от +5В
подать на D15:1 сигнал /RFSH (D13:28)
3. Отсоединить D13:6 от D1:8 так,чтобы осталась связь между
D13:6 и резистором, соединяющим его с +5В.

Кстати,AY-3-8910(не YAMAHA!) также плохо работает в
турбе, как и ВГ93, такthat this modification will also solve the problem
of the brake AY.
For normal operation of the UZH O turbo, change the diode
VD2 in the controller circuit. If it is not replaced, most
turboloaders in the turbo will fail, since they use jp #3d30 to access
DOS routines. In the turbo, this same
VD2 greatly slows down the switching of ROM pages (DOS/BASIC).
As a result, the effects are familiar to everyone: the launched one hangs and/or
craps on the disk...

Here is a diagram for replacing the VD2 diode (according to the Pentagon scheme
controller):

As it is:

DD77(LL1) DD76(TM2)
┌──┐ VD2 ┌──┬──┐
│ 1│11 ──│<──┬───┬─────13│ │ │
│ │ ─┴─ │ │ │ │
└──┘ C3 ─┬─ │ │ │ │
─┴─ █ R13 └──┴──┘
│
DD78(AG3) │
┌──┬──┐ │
│ │ │12──┘
│ │ │
│ │ │
└──┴──┘

And how it should be:

DD77(LL1) LI1 DD76(TM2)
┌──┐ ┌──┐ ┌──┬──┐
│ 1│11 ───┬─────────13│ &│11───13│ │ │
│ │ ─┴─ ┌──12│ │ │ │ │
└──┘ C3 ─┬─ │ └──┘ │ │ │
─┴─ │ └──┴──┘
│
DD78(AG3) │
┌─────┐ │ Free element LI1 is in the circuit until
│ │12──┘ turbo operation.
│ │
│ │
└─────┘

and here, in fact, is the scheme for finalizing the controller
BETA-DISK-TURBO in BETA-DISK-TURBO-HD :)

BQ1: 16 MHz DD9
┌──┬──┐
┌──────┤█├───────┐ DD86:6───2│A0│ │
│ DD8.1 DD8.2 │ DD8.3 DD93:14──5│A1│Y0│4──> on DD76:3
│ ┌──┐ ┌──┐ │ ┌──┐ › instead of 2MHz
│ └──┘ │ └──┘ │ └──┘ DD93:13──6│B1│ │ under revision
│ K43 │ K43 │ ├──┤ │ TURBO-VG
└──███──┴───███──┘ ┌─────1│ S│ │
│ ┌─15°OE│ │
DD5.2 │ ─┴─ └──┴──┘
+5V ┌──┬──┐ │
├─13° R│ Q│9──────┴───── to the HD/DD microphone in
D7(Z80:13)───12│ D│ _│ disk drive (see below)
D88:9 ───────11/ C│ Q°8── /bit7#FF
└─10° S│ │
└──┴──┘

Chips:
DD1: 1533KP11, DD2: 1533LE1, DD3: 1533TM2, DD4: 1533LL1,
DD5: 1533TM2, DD6: 1533LI1, DD7: 1533IE7, DD8: 1533L 1,
DD9: 1533(555)KP11

Any microcircuit, except DD8(L 1) and DD9(KP11)
* LH1 can be installed in the 531 series - this is even more stable.
can
be replaced with a similar one in the 555 series, this will be reflected
only in the load onPSU.

Old signals going to DD76:3, 2 MHz on TURBO-VG
- CUT OFF!

The signal from DD5:9 modifications going to the HD/DD microphone in the disk
gadfly can not be buffered (I have never buffered it, and
I have a 1m cable to the disk drive, and everything is OK). The HD/DD
mic itself is located on the motor board, just above the “HD”
hole in the disk (it is usually sealed). and the common wire usually goes to one leg of the microphone
, and the control
signal for the microcircuit is removed from the other. Here we give the signal
“HD/DD” to the second leg. Yes, now we have to remove the stickers
from the disks (on 3.5" - stickers! Insanity..). If you are going to use
720K 3.5" disks (there are no “HD” holes on them), in advance
pass the “HD/DD” signal through a 50..100 Ohm resistor, and
only then feed it to the microphone in the drive.

Actually, in The Pentagon would not need this 16 MHz generator, but because... Almost all Pentagon drivers have a 556PT4 PLL
, so the circuit is given taking into account its presence in the
controller. If there is no PLL, then DD9:2 needs to be supplied
DD76:6, DD9:3 - DD86:6, and then you need to cut off DD76:6
from DD92:3.

How does it all work together?

The density mode (normal/high) is switched
by software, bit7 of the #FF controller port: bit7=0 - normal,
bit7=1 -high density. Fortunately, it so happens that
all programs that work with the disk always keep this bit
in the same state: zero. Having made the modification, you can
safely run the old programs, they will work with 360K-720K disks just as before
.
I must say, without a turbo it doesn’t work, because... the exchange rate
increases exactly twice (2 MHz
is supplied to VG93 instead of 1 MHz) - now it is more than 60K/sec. By turning on
the turbo mode when executing a program from DOS ROM,
we ensure that the Z80 still has about 112
clock cycles per exchange cycle and it, the Z80, works with the VG93 just as
at normal density.
and high density disks can fit up to 18 512 byte-sec
tors (1.44M format) or up to 10 1024-byte (non-standard
but also 1.6M format). Thus, MSDOS (PS/2) HD disks are easily read/written.

After the turbo is modified, it will always turn on during
disk operations, but only after the disk has had time
to spin up. If the exchange with the disk is at normal
density, the turbo mode can be turned off with the microphone
“TURBO/NORMAL” (interestingly, this was needed for one
only program: FUT by MI&DI). If the turbo was
turned on, it turns off while the disk is spinning up, and then
turns on again.
When positioning for modification TURBO-VGalways goes
2 MHz, unfortunately, at a higher frequency the VGshka
does not work.
According to rumors, one of the two reserve contacts
of the 5.25" drive connector should receive a control signal
for the engine rotation speed. But in my Mitsumi 3.5" both
reserve ones do not go anywhere. So, I sent the "HD/DD"
signal through the 2nd backup pin (numbered as 4th on the
disk drive connector) - and it works fine for me, and on the PC.

P.S. All designations of the elements are given according to the Pentagon diagram
in which the Z80 address signals are designated A1-A16.
Unfortunately, there is no way to do without removing the drive casing,
but all the alterations in it are just one wiring.

P.P.S. The new turbo mode scheme was provided to me
by Roman Milyukov (2:4616/8.52) for which many thanks to him!

╚═════════════------ End of file HDdoc .W ------═════════════ў

Goodbye, Oleg & All.

[ Team ZX-HSC'84 ]

PF> Yes. But AFAIK, our controller is completely unable to read
PF> 1.44 MB.

No, it can’t, but if you have hardware, you can write a driver...
True, I won’t do this (or maybe I can’t), and I don’t need it ;)

With best wishes, Dmitry.

RM>> VGshka will not
RM>> have time to send data.

KF> ??? This doesn't happen. But #FFs from the VGshka are read with
KF> a short iorq.

:) So it simply does not have time to output data to the bus!
According to its specifications, data appears 350 ns after the start of the /CS signal.

KF> It turns out that the bits are bent.

How is that?