How to overclock the modem?????
ZXNet echo conference «real.speccy»
█│■
█┼│ All!
In short, subzh!
I connected a hayes modem to 9600, according to Shepelev’s scheme, but it doesn’t handle more than 2400.
What should I do????
Sorry, they brought fresh beer...
July 22, 00, somewhere at 09:05
Grigory Klimenko in his letter to All wrote:
GK> In short, subzh!
GK> I connected a hayes modem for 9600, according to Shepelev’s scheme, but it doesn’t handle more than 2400
GK>. What should I do????
What, does he write CONNECT 2400?
Come again...
Alexander
^^^^^^^^^^^^^^ but this is in vain.
Kondratieff rules.
GK> pulls.
are you sure that the line allows you to connect at higher speeds?
GK> What should I do????
see above F-))))
For this Au Revoire, CompoWellcome. [Team GS]
1. Disassemble Shepelev's circuit and assemble it according to Kondratiev's circuit. ;)
GK> can’t handle more than 2400.
GK> What should I do????
2. Configure the modem using AT commands. You have set a limit on the maximum
connection speed, unless, of course, the modem is really 9600.
Sincerely, Felix.
[I ZX]
Convert to Kondratiev's scheme. Only, COMPLETELY - without long wires and a lot of interference. Then, if the modem and line allow, there may be a connection up to 33600.
Good luck! Evgeny. (Aka 2:5057/18.73) ■ The night is shorter than the day... The day will kill me... The world of illusions burns down in it. (C) Aria.
AZ> What is it, it writes CONNECT 2400?
It writes depending on the set port speed, but it downloads the maximum
at 4800.
▌┘│ ■ ▐ ┬─┐▌ ┬──
Sorry, they brought fresh beer... ▌│└─│┌┬┐ ▌ ▌_│▌-┤▌_┬
GK> It writes depending on the set port speed, but it downloads
GK> to a maximum of 4800.
Acorp or what? RockWell?
So tell him uh... s95=255 in my opinion and he will write the real connection speed.
Come again...
Alexander
GK>> 2400 not
ND> ^^^^^^^^^^^^^^ but this is in vain.
ND> Kondratieff rules.
That's for sure. I have a controller lying around right now made by ARC, but
that’s not the point, the circuit is fundamentally Kondratiev’s very little
modified, so with this device a person sat on 33600 and no glitches.
Denis
Once upon a time Fri 28 Jul 2000, Denis Rybakov and Nicolas Davidov thought on the topic:
How to overclock a modem?????
[...]
ND>> Kondratieff rules.
DR> That's for sure. I have a controller made by ARC lying around right now, but no
what kind of controller is it? :-/
DR> this is the essence, the scheme is fundamentally Kondratiev’s, very slightly changed,
'quite a little' is how much? draw.
DR> so with this device a person was sitting on 33600 and no glitches.
well, what do you think I’m sitting on? ;-)
ND> 'just a little' is how much? draw.
So, instead of logic, the selection on ID7 is simply collected, etc. The operating principle
is the same as the original :) ... I just can’t wash it yet :(
Once upon a time Sun 30 Jul 2000, Igor Turashev and Nicolas Davidov thought on the topic:
How to overclock a modem?????
DR>>> this is the essence, the circuit is fundamentally Kondratiev’s quite a bit
DR>>> modified,
ND> a. then you don’t have to.
Well, whatever you want :)
ND> btw, why put id7 there?
And something that doesn’t conflict with Soundrive and the programmer, and make fewer connections
;)
PS: Soap obviously doesn’t come to you.........
All the best, Nicolas!
Alexander aka Arc of RLDG.
IT>> and
IT>> are not convenient :)
ND> don’t mock me ;)
Don’t tell me that it’s convenient on a modem... ;)
A long time ago Thu 3 Aug 2000, Igor Turashev and Nicolas Davidov thought on the topic:
How to overclock a modem?????
ND>> btw, why put ID7 there?
IT> So to make ports on it, you don’t need logic, practically, and decryption
but on logic simpler.
IT> tougher.
does it need to be tougher?
AA> And somehow it doesn’t conflict with Soundrive and the programmer, and
AA> there are fewer connections to make ;)
well, if you do everything together then it’s clear, but if it’s just momed, then it’s simpler
with small logic.
AA> PS: Soap obviously doesn’t come to you.........
it looks like that, and it’s only similar with your region. I’ll find out, but in the meantime try
write to me at 2:5085/26.51.
FK> 2. Configure the modem using AT commands. You have a limit set to
FK> the maximum connection speed, unless, of course, the modem is really 9600.
The modem is really 9600, and in Melon the port speed is set to 9600, but if
you set the port speed to 2400, or 4800, then it goes much faster.
Also in Melon 1.90 there is a funny thing, at the port speed 9600, the modem works normally, but at 2400, or 4800, after the modem detects that the line is busy, it hangs until the break is pressed. And I have to redial the number, and this glitch is only
in version 1.90.
PS. What specific commands need to be set for the modem to come to life?
ND> but in terms of logic it’s simpler.
Bullshit, the main thing is that it works ;) Ark just does everything in ID7 ;)
IT>> tougher.
ND> does it need to be tough?
Well, just in case, perhaps it might come in handy later? ;)
IT> But I feel sorry for myself, and the momed is not titanium in the least, but I have only one :)
okay, don’t excuse yourself :-D
IT> p.s. Today I’m sending you a test to another address...
throw it, throw it, I’ll let you know when I call.
For this Au Revoire, CompoWellcome.
2:5085/26.51
ND> apparently he has a lot of them :). btw, it's like GS - eight pieces are crammed.
Probably :) But why so many in GS? Or so, for quantity? :)
Well, you said...
ND> well, if you do everything together then it’s clear, and if only momed, then on
ND> fine logic is easier.
Have you heard about delays that are measured in nanoseconds?
Bye, Yuri.
[I ZX] [Chemists] [For full address] [Ski poles]
/_500:322/33@ZXNet_/ /_dang...@bigfoot.com_/ /_dang...@mail.ru_/
/_2:462/93.33@FidoNet_/
... Proved yourself - secure it!
IT>> Bullshit, the main thing is that it works ;) Ark just does everything like ID7 ;)
ND> apparently he has a lot of them :). btw, it's like GS'e - eight pieces
Once upon a time, Fri 11 Aug 2000, Igor Turashev and Nicolas Davidov were thinking on the topic:
How to overclock a modem?????
IT>>> Bullshit, the main thing is that it works ;) Ark just does everything in ID7 ;)
ND>> apparently he has a lot of them :). btw, it's like GS - eight pieces are crammed.
IT> Probably :) But why so many in GS? Or so, for quantity? :)
well, not really for the quantity... there is a crowd of internal ports + RAM sampling is done on
them.
YT> Have you heard about delays, which are measured in nanoseconds?
yes, yes, of course, two logic elements of the 1533 series introduce simply crazy delays
interfering with the normal operation of the circuit and one ID7 will bring you salvation...
don’t make my slippers laugh, sir :-)
A long time ago Sat 12 Aug 2000, Yuri Teslenko and Nicolas Davidov were thinking on the topic:
How to overclock a modem?????
ND>> apparently he has a lot of them :). btw, it's like GS'e - eight pieces are crammed.
YT> But this is in vain :) Instead of 8 ID7 - 4 ID3 you can put in.
you can, but not necessary. You should have looked at his diagram first before giving advice
... and I can imagine what size the board would have been then...
FK>> is really set to 9600.
GK> The modem is really set to 9600, and in Melon the port speed
GK> is set to 9600, but if you set the port speed to 2400, or
What about 19200?
GK> 4800 then it runs much faster.
Run S-Terminal_v1.1 (an alternative terminal /tr-dos, nmi used/ for the
MK circuit) and find out the UART type of your modem. Results in the studio. Has your ROM BASIC_48
been corrected to work with interruptions from the modem according to the MK scheme? If not, then
S-Terminal and Dickie Mailer will not work.
GK> Also in Melon 1.90 there is a joke: at port speed
GK> 9600, the modem works normally, but at 2400, or 4800, after
GK> the modem detects that the line is busy, it hangs until the
GK> break is pressed. And I have to redial the number, and such a glitch
GK> only in version 1.90.
1.90 I did not use at all, only the latest branded version 1.80+.
GK> PS. And what specific commands need to be set in order for the
GK> modem to come to life?
Declassify the name of your modem and throw (here or to me) its profiles at
at&v (or whatever they are called up for you?). Options: at&v, at\s, ati4.
Aga, and even better pzu flash and on tire - soldered minimum-rulez!
Good luck, Yuri!
ND> it is possible, but not necessary. You should have looked at his diagram first before giving advice
ND>... and I can imagine what size the board would have been then
ND>...
Have you seen 1533ID3? They are narrow. So the fee would even be reduced.
Bye...
ND> yes, yes, of course, two logic elements of the 1533 series introduce simply crazy
ND> delays that interfere with the normal operation of the circuit, and one ID7 will bring you salvation
ND> salvation... don’t tell my slippers, sir :-)
Of course, it all depends on the circuit. Still, it is illogical to build “cascades” from logic with total
delays of hundreds of ns.
Bye, Yuri.
ND> I was there just yesterday - there’s nothing from you.
ND> in general, through our NC, I found out that problems with soap in region 5052 are only
ND> at
ND> your node, because your local NC answered the test quite quickly. figure it out
ND> on the spot.
Ok, I’ve slowly started to figure it out, but I’m handing this matter over to Ark, because I’m
leaving myself :)
Message Return, CompoWellcome. at: 2:5085/1.149
2:5085/26.51
Once upon a time Tue 15 Aug 2000, Wlad Sorokin and Yuri Teslenko were thinking on the topic: How to
overclock a modem?????
ND>>> apparently he has a lot of them :). btw, it's like GS'e - eight pieces packed.
YT> Of course it all depends on the scheme. Still, building “cascades” from logic with
YT> summary
naturally depends on the circuit. in this case we were talking about the connection diagram of
momed, but there it somehow doesn’t matter.
YT> delays of hundreds of ns are illogical.
everything is usually done from improvised material, and not from what is needed ;-)
Once upon a time Tue 15 Aug 2000, Igor Turashev and Nicolas Davidov thought on the topic:
How to overclock a modem?????
ND>> well, not really for quantity... there’s a crowd of internal ports + sampling
ND>> RAM
ND> well then what's the matter? take the original GS circuit and sculpt your own.
I’ll sculpt it, but there’s no one to build the board, (in the financial sense)
although there is a simplification - a board from AOH, on its basis you can glue devices,
there are even four eight-bit ports!!!! Rulez!
the only thing is that the belly will have to be corrected a little, but it doesn’t matter...
Well, basically an idea, but I don’t have a complete diagram, just a skeleton...
Good luck, Nicolas !
[ Sandra ] [ ZX ] [ -SpaceSoft- ] [ Punks Not Dead ]
WS> I’ll build it, but there’s no one to build the board, (in terms of finances) although there is
on the plane ;)
WS> simplification - the board is from AOH, on its basis you can glue devices, there are even
WS> ports there are four eight-bit ports!!!! rulez! The only thing is that it's a bitch
WS> you'll have to tweak it a little, but it doesn't matter...
nu-nu, you know better...
WS> damn, in principle it's an idea, but I don't have a complete diagram, just a skeleton...
if necessary -> into the soap.
Bye, Yuri.
[I ZX] [Chemists] [For full address] [Ski poles]
/_500:322/33@ZXNet_/ /_dang...@bigfoot.com_/ /_dang...@mail.ru_/
/_2:462/93.33@FidoNet_/
... Proved himself- secure it!
Once upon a time Wed 23 Aug 2000, Yuri Teslenko and Nicolas Davidov thought on the topic:
How to overclock a modem?????
ND>> everything is usually done from improvised material, and not from what is needed
ND>> ;-)
YT> It’s especially annoying when flashing tricky ROMs and PLM, and sometimes ROM as
YT> PLM.
why? On the contrary, I like such circuit solutions :), and
the keyboard is expanded on RT4'x.
YT>> as a PLM.
ND> why? On the contrary, I like such circuit solutions :), there
ND> and the keyboard is expanded to RT4'x.
They have a very long read delay time. And sometimes they wear off on their own.
ND>> over there and the keyboard is expanded on RT4'x.
KF> Their reading delay time is very long.
Completely normal, for RT4A - 45 ns.
Good luck!
Vitalic.
... Cabbage is always an appetizer: it’s not a shame to serve it, and it’s not a shame to eat it sorry
Is 30ns a lot?
KF> And they wash themselves
KF> sometimes.
And they warm up like a stove!
All the best.
Roman.
[ZX]
Once upon a time, Sun 27 Aug 2000, Kirill Frolov and Nicolas Davidov were thinking on the topic:
How to overclock a modem?????
ND>> why? On the contrary, I like such circuit solutions :), there
ND>> my keyboard is expanded to RT4'x.
KF> They have a very long read delay time. And they sometimes erase themselves.
In a keyboard this is not critical.
For this Au Revoire, CompoWellcome. aka: 2:5085/80.3
2:5085/26.51
For decryptors, IMHO rulez! (if not at 10GHz of course) ;-)))
KF> are erased sometimes.
Yes? Approximately how often? As far as I know, in them, after all, during
programming, the extra conductors are physically destroyed? How do they
manage to recover?
With respect, Dmitriy.
VS> Quite normal, for RT4A - 45 ns.
For small logic, it’s definitely 5 times less.
RM> 30ns is a lot?
It depends on what you compare it with. A regular TTL has less logic.
KF>> Their read delay time is very long.
VS> Quite normal, for RT4A - 45ns.
Don’t trust reference books. ;) 70ns they all (RT4 and RT4A). RT4A differs only in a
larger percentage of programmability. +Their output is OK: whatever load you put
that’s the front you’ll get.
KF> Small logic is definitely 5 times smaller.
Where does such confidence come from? ;)
LA3, LE1 - 20, LL1 - 22 ns. This is for 555.
If, say, in some decoder there are 3-4 of them in series, then consider for yourself that it is faster.
In your circuit, for example, LH1 + IR23 = 20 + 42.5 = 62.5 ns. This without everything
else, if you also take into account the 7FFD decoder and KP11 - it will cost a hundred nan
definitely.
29 Aug 00 17:12, Nicolas Davidov wrote to Kirill Frolov:
KF>> Their reading delay time is very long. And they themselves
KF>> are sometimes erased.
ND> in the keyboard this is not critical.
And in the RAM decryption circuit?
Once upon a time Thu 31 Aug 2000, Kirill Frolov and Nicolas Davidov thought on the topic:
How to overclock a modem?????
KF>>> They have a very long read delay time. And they erase themselves
KF>>> sometimes.
ND> what does ^^^^^^^^^^^^^^ have to do with it? in the original letter it was mentioned
ND> it was the keyboard...
Do I think I remember what the letter is about, if there is no quota, no subject
corresponding?
The subject says - modem. And there are 3 clock cycles for everything, plus if IOGE is used
then
you need to decrypt everything in half a clock cycle.
VS> Where does such confidence come from? ;)
Firstly, it’s not 45ns, but closer to a hundred...
VS> LA3, LE1 - 20, LL1 - 22 ns. This is for 555. If, say, in some
VS> decoder there are 3-4 of them in series, then consider for yourself that they are faster.
What kind of fool would put so many of them? For this purpose there are more other
miposchemes. No need to build a barn.
VS> In your circuit, for example, LH1+IR23 = 20+42.5=62.5 ns.
For some reason it seems to me that IR23 is not excluded when using RT4.
VS> This is without everything else, if you also take into account the 7FFD decoder and KP11 -
VS> will definitely go over a hundred nans.
KP11 what? How does it relate to the decoder? Have you even seen the diagram of the spectrum?
VS>> - it will definitely come out for a hundred nan.
KF> KP11 what? How does it relate to the decoder? Have you even seen the diagram
KF> of the Spectrum?
Hmm... Vitalik has seen the diagram of the Spectrum ;)
And, besides, he knows how to use reference books and not talk nonsense.
Lower your tone, please. I wouldn’t like to organize demonstrative moderations here.
[I ZX] [Team Sashki-rulez] [Team Guardian] [Team Kelly]
Hasta la vista.
Tim Kelly
And all children are born brave...
... tim_...@mail.ru 500:622/1@zxnet http://timkelly.newmail.ru
RM>> 30ns is a lot?
KF> It depends on what you compare it with. A regular TTL has less logic.
1533ID7 - 30ns. And even though single elements have almost three times less, they
make long chains if you want to collect what is stitched into the RTshka.
All the best.
Roman.
[ZX]
On August 29, 2000, you wrote here:
[about RT cards]
KF>> sometimes they get erased.
DN> Yes? Approximately how often? As far as I know, in them, after all, during
DN> programming, the extra conductors are physically destroyed? How
DN> do they manage to recover?
The conductor burns out and explodes. And where is the garbage deposited?..
So that they do not recover, even artificial thermal aging is supposed to be done.
But practically I have not encountered this problem, everything works.
All the best.
Roman.
[ZX]
^^^
VS>> - it will definitely come out for a hundred nan.
KF> What is KP11?
Which one do you have after IR23!
KF> How does it relate to the decoder?
Who said it does? Does the conjunction “and” mean something to you?
KF> Have you even seen the diagram of the spectrum?
:-)
You will be disappointed if you think that this is inaccessible to mere mortals,
Good luck!
Vitalic.