mouse
ZXNet echo conference «real.speccy»
Can we talk about the mouse
We need a Kemston mouse circuit that uses not 9 wires, but 4. And does not use
the BB55A chip.
Thank you.
Until next contact All !
And how do you understand this?.. Clint Eastwood
26 Nov 01 17:33, Alexander Fisun wrote to All:
AF> You need a Kemston mouse circuit that uses not 9 wires, but 4.
Takogo does not exist. A passive mouse has 9 wires.
KF> Takogo does not exist. A passive mouse has 9 wires.
Well, firstly, there is a scorpion controller for the active mouse, and secondly, here
a year ago the active mouse controller circuit flew at its peak.
Sincerely, Evgeniy.
KF> Takogo does not exist. A passive mouse has 9 wires.
I say it happens. Nik-a Patlaenka has the same. Just a schematic diagram
no.
Until the next contact Kirill!
KF> Takogo does not exist. A passive mouse has 9 wires.
But I have a Zone controller and the mouse has 4 wires... :P So there's no need to say that this doesn't happen. Happens. But I don’t have a diagram.
ZXNet: 500:8462/1 ~/ <~ http://zs80.chat.ru/
Fido: 2:5057/21.93 /_ _> ftp://zx-museum.org.ru/zs/
Hемедленно нажми на RESET, Sergey!
01 Dec 01 11:44, Sergey Zotov wrote to Kirill Frolov:
SZ>> wires... :P So
KF> THE PASSIVE MOUSE.
Why the hell should you cripple the mouse... you need to install a regular mouse from the PC, for example...
[Wizard Warz] Yurik aka Jerri / Alien Factory /I \
[No RAR allowed] [SoHm] [Conceptual Minimalism] \ZX/
KF>>> Takogo does not exist. A passive mouse has 9 wires.
SZ>> But I have a Zone controller and the mouse has 4 wires... :P So
11 Dec 01 00:58, Yuri Potapov wrote to Kirill Frolov:
KF>> A PASSIVE MOUSE.
YP> why the hell should you cripple the mouse... you need to install a regular mouse from the PC, for example...
It twitches. And in general, this is a pussy mouse! It has no place in the spec!
A pure Spectrum mouse should be in the spec!
YP>> PC
YP>> for example...
KF> She twitches. And in general, this is a pussy mouse! She has no place
KF> in the spec!
KF> The spec should have a pure Spectrum mouse!
or an Amiga one. There is no difference;-)
JseveN from the 4th Dimension...
11 Dec 01 15:35, Alexander Fisun wrote to Kirill Frolov:
KF>> WITH A PASSIVE MOUSE.
AF> What is the difference between a passive mouse and an active one?
An active one, when recording coordinates in counter registers, can itself move
within the mousepad.
SZ>>> But I have a Zonovsky controller and the mouse has 4 wires...
KF>> IT’S A PASSIVE MOUSE.
YP> why the hell should I damage the mouse... I need to install a regular mouse from the PC
YP> for example...
It’s not that simple. To process the signals of the PC subject, a microcontroller or a microprocessor is needed. Oh, that’s enough for passive logic.
Such things... Here.
...Still its hard,hard 2 c fragile lives,shattered dreams...
YP>> PC
YP>> for example...
KF> She twitches. And in general, this is a pussy mouse! She has no place
KF> in the spec!
KF> A pure Spectrum mouse should be in the spec!
Master, you should have a girl (c)
December 12, 2001 01:21, Kirill Frolov wrote Yuri Potapov:
KF>>> THE PASSIVE MOUSE.
YP>> why the hell should you cripple the mouse... you need to install a regular mouse from the PC
YP>> for example...
KF> She twitches. And in general, this is a pussy mouse! It has no place in the spec!
KF> The spec should have a pure Spectrum mouse!
A pure Spectrum mouse - is it taken from an Amiga or pussy, but ancient and Russian?
Or is it the same pussy mouse, but also with a crippled cord and torn out
innards?
See you later Kirill!
... Best student - dead student
AF> What is the difference between a passive mouse and an active one?
A passive mouse without a processor, signal processing is handled by the interface using
small logic and counters.
13 Dec 01 09:48, Oleg Maxsimuk wrote to Yuri Potapov:
OM> Not everything is so simple. To process the signals of the PC subject, you need a
OM> microcontroller or microprocessor. And, for passive and logic,
OM> is enough.
Absolute nonsense.
YP>> PC
YP>> for example...
OM> Not everything is so simple. To process the signals of the PC subject, you need an
OM> microcontroller or
OM> microprocessor. Oh, for passive and logic it’s quite enough.
but then you can connect any mouse...
it’s time for me to change the mouse, I’ll go and get a brand new one, never worn,
maybe even a cordless one...
and here you suffer, the mouse is gutted
16 Dec 01 21:18, Yuri Potapov wrote to Oleg Maxsimuk:
YP> and here you suffer, gutting the mouse
But what a pleasure it is to gut a PUSSY mouse!
And the passive one doesn’t twitch, unlike some...
KF> But what a pleasure it is to gut a PUSSY mouse!
KF> And the passive one doesn’t twitch, unlike some...
that’s why it’s passive... the mouse then
The active one, yes... resists, squeaks, tries to bite...
horror in one word
OM>> Not everything is so simple. To process the signals of the PC subject, you need
OM>> a microcontroller or microprocessor. A, for passive and logic
OM>> microcontroller or microprocessor. And, for passive and logic, it’s quite
OM>> enough.
YP> but then you can connect any mouse... It’s time for me
YP> to change the mouse, I’ll go and get a brand new one, never worn, maybe even
YP> without cords... and then you’ll suffer, gutting the mouse
So I have nothing against active ones I don’t have any creeping things...
OM> So I have nothing against active creeping reptiles...
OM> Such things... Here.
well, come up with a powerful, cunning, smart controller on a PPLM, for example... and
so that without problems
after all, it’s done in Sprinter
OM>>> Not everything is so simple. To process the signals of the PC subject, you need
OM>>> a microcontroller or microprocessor. Ah, for passive and logic
OM>>> it’s quite enough.
KF>> Absolute nonsense.
YP> well, come up with a powerful, cunning, smart controller on a PPLM, for example
YP> ... and so that without problems
So a circuit on a peak controller is flying somewhere... Why is this PLM?
KF>>> Absolute nonsense.
OM>> Justify it, smart guy!
KF> Both the mouse and the computer have iron logic, but no
KF> controller.
Oh, well? We weren't talking about the guts of mice, but about the process of processing
signals coming from them... If someone doesn't catch up, put it on your fingers.
Look at the pinout of the PS/2 connector:
1. Data
2. NC
3. Gnd
4. VCC
5. Clock
6. NC
We have 2 wires, Data and Clock... How are you going to process them using logic? This means: logical elements AND, OR, inverters,
triggers, counters... Of course, if Monsieur Kirill is a subtle pervert, then he can
assemble a circuit on 30 publicly available logic boxes. elements, but it is much
easier to install a microcontroller.
Unfortunately, I don’t have a pinout of the passive creature’s connector at hand, but I’ll try
to remember from memory...
1. X
2. Y
3. Direction X
4. Direction Y
5. LMB
6. RMB
7. VCC
8. GND
Everything else without comment...
Once Monday December 24, 2001 at 00:31:51, Oleg and Yuri were arguing about
mouse
YP>> well, come up with a powerful, cunning, smart controller for
YP>> PPLM for example
YP> so that without any problems, with one firmware, you can add wheels, rocking chairs,
YP> tails, wings, legs, buttons, etc. to it
She's waiting! It will slow down in demos. Mazda.
[Wizard Warz] Yurik aka Jerri / Alien Factory /I \
[No RAR allowed] [SoHm] [Conceptual Minimalism] \ZX/
YP> so that without any problems, with just one firmware, you can add wheels, rockers,
YP> tails, wings, legs, buttons, etc. to it
Well, at the peak, also build it up as much as you want.
YP> the keyboard slows down... but the mouse doesn’t
WAIT is generated there. Peak, even the AVR type is unable to react so quickly
to an interruption.
KF> WAIT is generated there. A peak, even an AVR type, is unable to respond so quickly
KF> to an interruption.
AVR is not a peak. Is it really possible that a device operating at a frequency of 20 MHz with a time
execution time of one command of 4 clock cycles (jump commands do not count) cannot cope with this?
Such are the things...
... The truth is in the out there...
KF>> respond quickly to an interruption.
OM> AVR is not a peak.
A peak means small (from pico).
But whether it is an AVR or a microchip is not important.
OM> Surely a device operating at a frequency of 20 MHz with an execution time
OM> of one command of 4 clock cycles (jump commands do not count) will not be able to cope with this
OM> ?
Can't. The response time to an interrupt (several commands) + processing
interrupts
will be longer than the read cycle from the port of the Z80 (without turbo). But it is necessary
in parallel
to process the keyboard (or mouse). Imagine what the Speck program IN
A,(MOUSEPORT)
does in a loop - even if the peak has sufficient speed to respond
to every
request without a brake, it will not have time to process bytes from the mouse. Here we need
peaks
at 100 MHz.
KF> He can’t. The response time to an interrupt (several commands) + processing
KF> interrupt
KF> will be longer than the read cycle from the port of the Z80 (without turbo). But you need
KF> to process the keyboard (or mouse) in parallel
KF>. Imagine what the Speck program
KF> IN A,(MOUSEPORT)
KF> does in a loop - even if the peak has sufficient speed to
KF> respond to every
KF> request without a brake, it will not have time to process bytes from the mouse. Here
KF> peaks are needed
KF> at 100 MHz.
I repeat once again, registers with independent writing and reading are used. Zonov's controller contains IR26. But this kind of garbage won't work on the keyboard, since
you'll have to combine outputs, etc. In short, the diagram will turn out to be on the floor of "Leningrad";)
JseveN from the 4th Dimension...
OM>> cope with this?
KF> He won’t be able to. Interrupt response time (several commands) +
KF> interrupt processing will be longer than the read cycle from the port of the Z80 (without
KF> turbo). But you need to process the keyboard (or mouse) in parallel.
KF> Imagine what the spec program IN A, (MOUSEPORT) does in the loop --
KF> even if the peak has sufficient speed to respond to every
KF> request without a brake, it will not have time to process bytes from the mouse. Here
KF> we need peaks at 100 MHz.
Okay, let's do the math... On the Spectrum, the IN A command is executed in 12-16 clock cycles. i.e.
(roughly) 1/218750 of a second. At the peak the team: 1/5000000 i.e. During the execution of an IN command on the Z80, the peak can execute 22.8 commands. This, of course, is all at
the maximum, but Comrade Kirill probably forgot that PIC has a bunch of nice
things such as, for example, interrupts... Specifically, the 6th peak output is 16F84
RB0/INT.
Now Comrade Kirill, let’s estimate the speed of polling the mouse for speckies, usually it does not
exceed 50 Hz. And finally, let's calculate the exchange rate of the PS/2 port with the mouse
(maximum 20 kHz) i.e. 250 PIC commands.
Specifically for you, here is a modest listing of interrupt handling on PIC:
ORG 04h
BTFSC PORTB,0
GOTO PORT_X
BTFSC PORTB,1
GOTOPORTY
PORTKEY MOVF BACKUP_KEYS,0
MOVWF PORTA
RETFIE
PORTY MOVF BACKUP_Y,0
MOVWF PORTA
RETFIE
PORTX MOVF BACKUP_X,0
MOVWF PORTA
RETFIE
I.e. Interrupt processing cycle max. 10*4 clock cycles 20000000 MHz.
And this doesn’t suit you? No question... Let's put three 8-bit registers, and we will
write the date there in advance...
Should we still flame or is that enough?
KF>> speed to respond to each request without a brake, then
KF>> it will no longer have time to process bytes from the mouse. Here we need peaks at
KF>> 100MHz.
VK> I repeat once again, registers with independent writing and
VK> reading are used. The Zonov controller contains IR26. But this kind of garbage won't work on the keyboard, since you'll have to combine outputs, etc. In short, the scheme
VK> will turn out to be on the floor of "Leningrad";)
not at all. controller+RAM+pack of multiplexers. I did it myself, it worked
perfectly. Now it’s just lying there ;)
Another option is programmable logic. $4 for a mikruhu is acceptable IMHO.
All the best.
Dmitry. [ZX] [Quake]
np: The Beatles - The Night Before
OM> Okay, let's do the math... On the Spectrum, the IN A command is executed in 12-16
OM> clock cycles. those. (roughly)
To be precise, the command itself is 11 clock cycles, and the cycle on the bus is only 2.5 clock cycles.
That is 0.71 µs. Even if you subtract the time for the signals to settle down on the bus,
then you are left with about half a clock cycle less.
OM> 1/218750 of a second. At the peak the team: 1/5000000 i.e. during the execution of
OM> the IN command on the Z80, the peak can execute 22.8 commands.
Less than 15 clock cycles or 4 commands of commands is realistic.
OM> This is of course all at the maximum, but Comrade Kirill probably
OM> forgot that the PIC has a bunch of nice things such as, for example,
OM> interrupts...
Well yes...
OM> Specifically 6 peak pin 16F84 RB0/INT.
What is the interrupt response time? At a minimum, finish the current
command
(how many clock cycles does the longest command take at the peak?), and push the
address onto the stack.
OM> Now, comrade Kirill, let’s estimate the speed of polling the mouse on a speckie,
Crazy. 11 Spec cycles per cycle.
OM> usually it does not exceed 50 Hz.
Do we need it to work normally or always?
OM> And finally, let's calculate the exchange rate of the PS/2 port with the mouse (maximum 20
OM> kHz), i.e. 250 Peak commands.
Does Peak have a serial interface?
OM> Specifically for you, here is a modest listing of interrupt handling on
OM> PIKE:
OM> ORG 04h
OM> BTFSC PORTB,0
OM> GOTO PORT_X
OM> BTFSC PORTB,1
OM> GOTO PORTY
OM> PORTKEY MOVF BACKUP_KEYS,0
OM> MOVWF PORTA
OM> RETFIE
OM> PORTY MOVF BACKUP_Y,0
OM> MOVWF PORTA
OM> RETFIE
OM> PORTX MOVF BACKUP_X,0
OM> MOVWF PORTA
OM> RETFIE
I don’t remember how everything slows down at the peak, but you can immediately see that you can’t keep up.
OM> That is Interrupt processing cycle max. 10*4 clock cycles 20000000 MHz.
^^^^^^^^^^^^^^
Well, if you overclock the peak to such a frequency,
then
it will emulate the entire spec for you.
OM> And that doesn’t suit you? No question... Let's install three 8-bit registers,
OM> and write the date there in advance...
OM> Is it still enough to flame or is it enough?
Add an address decoder to those registers (the peak cannot), and the differences from the
pure hardware
version are as follows: one peak for $4 or 4 counters for $1.
The version with programmable logic is moreреальный.
DL> not at all. controller+RAM+pack of multiplexers. I did it
DL> myself, it worked
DL> perfectly. Now it’s just lying there ;)
What if the circuit is in the echo? it just lies there anyway
VK>>> such garbage will not
VK>>> work, since you will have to combine outputs, etc.
VK>>> In short, the circuit
VK>>> will turn out to be on the floor of “Leningrad”;)
DL> the archive is 400 kilos, it contains a 4-meter .bmp and one .trd.
DL> something real people will be happy with ;)
DL> archive is 400 kilos, it contains a 4-meter .bmp and one .trd.
DL> the real people will be happy ;)
yes. Can you describe it in general terms? What is the elemental base and what is
connected to what.
JseveN from the 4th Dimension...
All the best.
Dmitry. [ZX] [Quake]
VK> yes indeed. Can you describe it in general terms? What element base and what
VK> is connected to what.
is plugged into the bus. no extras no wires needed.
the mouse is passive. keyboard - a la AT.
Z...@9.8MHz, 537RU25, 2764, KP11*3, all sorts of TM8, TM9, logic. a total of 15 cases.
handling is completely transparent for the Speccy - both for the mouse and the keyboard.
if no key is pressed, the native keyboard port functions in the Speccy
together with all tape recorders, if pressed, the port is replaced.
the built-in Z80 fills the 256-byte table in the built-in RAM, into the same RAM has
access to the Speccy processor (transparent time division up to double
turbo in Speccy). two banks inside the RAM are used, one is available for the Speccy,
the other is filled with the built-in Z80, after which the banks are changed. All 256 addresses are used and
correctly processed to allow simultaneous
reading of rows. just in case, keyboard events are synchronized with the INT Speccy signal in order to avoid collisions. The mouse is scanned by software with a frequency of about 20 kHz, the data for the mouse goes into the same RAM and is read in the same way. Among the bells and whistles there is the possibility of WAIT and RESET for the Speccy processor. It is also
possible to “slip” any operation code into the latter if the current command
is selected from ROM. in the keyboard program there are smallprograms like "reset with transition to TR-DOS", "reset to 48th mode with open/closed
pages" and the like. the button is pressed, Reset goes to Speccy, then the necessary instructions are fed there one by one. turns out to be rulez ;)
DL>>> not at all. controller+RAM+pack of multiplexers. I did it myself
DL>>>, it worked
DL>>> perfectly. now it’s just lying there ;)
YP>> what if the circuit is in the echo? it just lies there anyway
np: http://www.mp3.ru - Zemfira - Pochemu