SUCKS! :(
ZXNet echo conference «real.speccy»
OG> PC keyboards. This did not help :( then it was decided to turn off
OG> other signals in this way... in short, this also did not help
OG> :|
What are the “others”?
OG> but the most interesting thing is that when I finally took out this
OG> controller the computer remained hanging :( I thought strangely and
OG> soldered all the places where there were breaks...
OG> as a result of which I got a mattress (this happens when
OG> you take out the ROM).
OG> The ROM is working properly and the processor too... having sniffed the ROM
OG> with an oscilloscope, I came to the conclusion that more than one signal is not coming to it:|
This cannot happen. There must be at least something on the address and on the data, otherwise
you wouldn’t have a “mattress”. address... in short, the pipe :(
Check what is on the 20th and 22nd legs of the ROM (is there 27512?). If at least one of
of them has a constant log.1, then check further
the circuit of this signal, i.e. what it is formed from.
OG> maybe, of course, I put a snot on the board somewhere... but it seems like everything
OG> I checked it carefully...
Rinse board from the flux and look at it again carefully using a magnifying glass.
p.s. I haven’t seen the Kai-1024 circuit;)
Sincerely, Felix.
[I ZX]
04 Sep 00 10:33, Oleg Golenkoff wrote to All:
OG> Damn! people! help fix KAY1024...
When a user climbs into the computer with a blunt hot object,
this is what happens... Kitchen creativity... Contact
qualified specialists... (C) Nemo.
OG> I bought a PC keyboard controller from SCORPION and in the dock it was said that in
OG> KAY it is inserted without any modifications. because the buses are the same in
OG> in general :(
There are several differences, but they don’t seem to affect the controller.
This clock frequency is 14MHz, +12V, and I don’t remember what yet.
OG> the most... without hesitation I decided (on the advice of a friend) to cut the
OG> IORGE track going to the bus connector with the PC keyboard controller. This did not
OG> help :( then it was decided to turn off other signals in this way...
OG> in short, this also did not help :| but the most interesting thing is that when I took out
OG> finally this computer controller remained hanging :( I thought strangely
OG> and soldered all the places where there were breaks... as a result of which I got
OG> a mattress (this happens when you take out the ROM). ROM - OK & processor
OG> too...
Well, I short-circuited a couple of tracks somewhere...
OG> having sniffed the ROM with an oscilloscope, I came to the conclusion that it doesn’t
OG> receive more than one signal:|address...
This can’t happen! And if the processor is turned off, then there will be no mattress!
OG> in short, the pipe :( maybe, of course, I put a sniffle somewhere on the board... but it seems
OG> I checked everything carefully...
What do you mean the signals do not reach the ROM? This does not happen.
There is a mattress - it means that on the address bus all the time 0x38, on the High-Z data bus.
See address signals on ROM CS and OE. One of them is in the unit all the time.
I’m not sure exactly, but most likely CS goes directly to GND according to the circuit, which means
all that remains is to check pin 22.
And here is the circuit:
#CSROM <----------------+
|
1533LL1 |
2 +---+ 1533ЛЛ1 |
#RD <----+ 1 |3 10+---+ | 600
1 | +----+ 1 |8 | +----+
#MREQ <----+ R19+------> +---+ +--+ | +----+ ROM, output 22.
D10 | +---+
| D10
+------+
1533ЛH1 |
#BLK <-----+ 1 |
4 | 11| 1 |10 |
#A14 <-----+ +---+
5| |
#A15 <----+
+---+ D32
D49
from the bus. And you most likely have stuck one
of the following: #RD, #MREQ, #BLK, #A14, #A15, #CSROM, #RDR.
Check with an oscilloscope, it should be like this (in the case of a mattress):
#A14 is always log. 0
#A15 is always log. 0
#RD pulses
#MREQ pulses
#BLK always logical. 0
#CSROM impulses
#RDR most likely you have a log here. 0 or GND
Logical zero differs from gnd by about half a volt,
so if you see less than expected somewhere, it’s definitely stuck on
the common wire.
Greetings, Kirill!
It’s been a long time since you measured the “mattress” with an oscilloscope. KF> All signals are taken from the bus. And you most likely have stuck one KF> Check with an oscilloscope, it should be like this (in the case of a mattress): KF> #A14 is always logical. 0 KF> #A15 is always log. 0 2OG: If everything is fine on the ROM, check the /RESET, /BUSRQ, /WAIT signals on the The protoss escaped. The Zerlings took off. It's time for me to call it a day. Ivan. - Divorce the scheme, big and small... [Sprinter-II] [Forth-CPU] [ZX] ... ivan...@mail.ru * http://st-rektal.chat.ru * http://www.atlant.ru/peters
On the address bus #38 is mixed with the running address of the stack, where it is written
return address.
KF> of the following: #RD, #MREQ, #BLK, #A14, #A15, #CSROM, #RDR.
Here there are jumping pulses, that is, a comb, then no, with a frequency of several hertz.
Similarly, with a frequency two times lower than on A14.
processor. In a normally operating computer, the first two should have a
logical one, the third should mostly have a one with short
impulses to zero (they may not exist in certain sections of the program).
IM> Likewise, with a frequency two times lower than on A14.
Exactly! This is regeneration!
IM> be a logical unit, on the third there is basically a one with
IM> short pulses to zero (they may not be in certain
IM> sections of the program).
There are no sections of the program there, it only writes to the stack.
The Protoss have fled. The Zerlings took off. It's time for me to call it a day. Ivan.
- Divorce the scheme, big and small... [Sprinter-II] [Forth-CPU] [ZX]
... ivan...@mail.ru * http://st-rektal.chat.ru * http://www.atlant.ru/peters
IM>>> be a logical unit, on the third there is basically a unit with
IM>>> short pulses to zero (they may not be in certain
IM>>> sections of the program).
KF>> There are no sections of the program there, it only writes to the stack.
The Protoss have fled. The Zerlings took off. It's time for me to call it a day. Ivan.
- Divorce the scheme, big and small... [Sprinter-II] [Forth-CPU] [ZX]
... ivan...@mail.ru * http://st-rektal.chat.ru * http://www.atlant.ru/peters
KF>> When is WAIT? When reading from a missing ROM there is no stack, when writing to
KF>> there is no stack either.
IM> If there is a stick caused by a soldering iron... The machine will stand up. And on the ROM
IM> there will be no signals left...
So it has a rustling percentage - “strips” into the memory are push-it..
btw, what about a white computer?
2minisoft: try to force control signals to the ROM
Always yours Combin8or/PHT
KF>> not on the stack either.
IM> If there is a stick, set with a soldering iron... The machine will stop. And there will be no
IM> signals left on the ROM...
So his mattress is hanging! And the mattress is an eternal RST 38 with 0x0038 inserted into
the stack throughout memory, you know it yourself...
12 Sep 00 20:47, Yuri Voynalovich wrote to Ivan Mak:
YV> btw, and kay is a white computer?
Kay-256 yes. Kay-1024 only in the turbo.
YV> 2minisoft: try to force control signals to the ROM
Then either colored squares will appear, or the computer will smoke.
The Protoss have fled. The Zerlings took off. It's time for me to call it a day. Ivan.
- Divorce the scheme, big and small... [Sprinter-II] [Forth-CPU] [ZX]
... ivan...@mail.ru * http://st-rektal.chat.ru * http://www.atlant.ru/peters
The Protoss have fled. The Zerlings took off. It's time for me to call it a day. Ivan.
- Divorce the scheme, big and small... [Sprinter-II] [Forth-CPU] [ZX]
... ivan...@mail.ru * http://st-rektal.chat.ru * http://www.atlant.ru/peters
IM> Are you sure? The mattress may arise at the first moment after switching on and
IM> remain, and the computer will stop. Glitches are an unpredictable thing...
The mattress appears when reading 0xFF instead of COP. And if the processor managed
to corrupt 48kb, then why should it continue to freeze?
options to the ass:
1) the csrom signals are planted
2) their forming signals are planted
(for example, 5v hangs on /RD)
2) the elements that form the csrom are dead
3) a simple break
4) the reason 1) there may be some
slot devices
5) or maybe a hoof problem threw back
6) someone extinguishes the bus when ROM is selected
and so on;)
to find out the reason, an oscilloscope is quite enough.
although all my life I have been repairing computers without it and even
without an Avometer;-)))
with just two simple probes: an LED and
a low-frequency amplifier.
ps question:
I observed the same thing glitch on completely different computers
(Profi and Pntheagon) - go to basic48 (usr 0), press
Enter a couple of times and fall out in TRDOS. Has anyone had this happen? and what's the matter?
According to my assumption, this is related to the access time to
ROM.
bYe, JTN/4D j...@mail.ru
Hi *Vassili*!
And it all started on 15-Sep-00 at 09:39:00, when Vassili Klimov
talked to All about SUCKS! :(
VK> to find out the reason, an oscilloscope is enough. Although all my life
VK> I have been repairing computers without one and even without an autometer;-))) with just two
VK> simple probes: an LED and a low-frequency amplifier.
The best substitute is a wire with a probe thrown through a resistor
on video. _Immediately_ almost any tricky signals are visible, and frequency periods can be easily calculated by eye ;) For the spectrum, this is sometimes
even better than an oscil, although sometimes it is exactly what is needed..
VK> ps question: I observed the same glitch on completely different computers
VK> (Profi and Pntheagon) - exit to basic48 (usr 0), press Enter
VK> a couple of times and fall out into TRDOS. Has anyone had this thread? and what
VK> is the matter here? according to my assumption, this is connected with the access time to
VK> ROM.
A painfully familiar glitch ;) One people had this in Leninist1, I then had a problem with the controller, but I never figured it out - the main thing for people was that the din worked.. Most likely the glitch is buried in the block checking the correspondence of the address of the readable COP for installing/resetting TRDOS, because without the
controller no glitches were observed..
VK> bYe, JTN/4D j...@mail.ru
Always yoursCombin8or/PHT
KF> Then either colored squares will appear, or the computer will smoke.
So not on top of the old ones! Although then I still wouldn’t smoke ;)
Always yours Combin8or/PHT
Once Yuri Voynalovich tapped on the keyboard to Ivan Mak:
YV> So his percentage is rustling - “strips” in the push-it memory..
like I found something :( iorq nope... let’s look further...
YV> btw, and kay white computer?
like only in turbo...
YV> 2minisoft: try to force control signals to the ROM
first I'll deal with iorq...
Minisoft avec NRJ^Ascendancy Creative Labs [I ZX] [I AMIGA] [I BEER]
ASCENDANCY HQ: ascen...@mail.ru http://www.ascendancy.homepad.com
ZX-Spectrum Banner NetWork: http://zxbn.narod.ru Let's Unite!
Minisoft avec NRJ^Ascendancy Creative Labs [I ZX] [I AMIGA] [I BEER]
ASCENDANCY HQ: ascen...@mail.ru http://www.ascendancy.homepad.com
ZX-Spectrum Banner NetWork: http://zxbn.narod.ru Let's Unite!
The Protoss have fled. The Zerlings took off. It's time for me to call it a day. Ivan.
- Divorce the scheme, big and small... [Sprinter-II] [Forth-CPU] [ZX]
... ivan...@mail.ru * http://st-rektal.chat.ru * http://www.atlant.ru/peters
VK> Enter a couple of times and you fall out in TRDOS. Has anyone had a thread like this
VK>? and what's the matter?
It was. But after adding the air conditioners to the controller board, all the glitches disappeared.
The glitches appeared after converting the computer to a common field and after installing another
ROM.
VK>, according to my assumption, this is due to the access time to
VK> ROM.
Imho you right.
██▓▓▓▒▒▒░░░ Sergey aka Cyber from Cobra Software.
█▓▓▓▒▒▒░░░ 500:322/62 500:322/67.16
... [I ZX] [Beer] [Storm] [Turbo Commander] [ZXNet] [FidoNet] [Offspring]
YV>>> 2minisoft: try to force control signals to the ROM
KF>> Then either colored squares will appear, or the computer will smoke.
YV>>> 2minisoft: try to force control signals to the ROM
KF>> Then either colored squares will appear, or the computer will smoke.
VK> (Profi and pentagon) - go to basic48 (usr 0), press
VK> Enter a couple of times and fall out into TRDOS. Has anyone had a thread like this
VK>? and what's the matter?
which LP2? neither Pro nor Penta have it :)
If you're talking about the Conder that delays switching, then
I still don't understand: how can you get into trdos from _pure_ 48 BASIC
by only pressing Enter?
KF> So what if he planted a blot on the surface or the ground?
Well then, what hasn’t burned yet won’t burn anymore..
Always yours Combin8or/PHT
KV> which LP2? neither in Profi nor in Penta it is not :) If you are talking about the conder
KV> which delays the switching, then I still don’t understand: how
KV> from _pure_ 48 BASIC you can get into trdos by pressing only Enter?
If the block checks whether the address of the readable cop is in
#3dxx, then maybe. Then maybe delivering
additional blocking air conditioners in that area will really help..
Always yours Combin8or/PHT
KV> which LP2? neither in Pro nor in Penta it is not :)
Eh... It turns out LA2.
KV> If you are talking about the Conder that delays the switching, then
KV> I still don’t understand: how from _pure_ 48 BASIC
KV> you can get into trdos by only pressing Enter?
Well, there are glitches at the moment of switching addresses slip by...