Regarding the file "24to96.zip", which "explains" how to increase the speed
of a 2400 baud modem to 9600:
THIS FILE IS EITHER A HOAX OR ABSOLUTELY THE WORST TECHNICAL DOCUMENT EVER
WRITTEN. DON'T TRY IT!!!!!!
In my opinion, it originated as a hoax, but since I know of two cases where
it has resulted in the destruction of perfectly good 2400 bps modems, I
thought I would write up this "rebuttal", and encourage all to ignore the
24to96 file, as well as remove it from the BBS's it has shown up on.
The problem with this type of "joke" is, unless the reader is fairly
technical and knows something about chips and especially about modems, it
almost sounds as though it would work! That has been true in the two cases
I know of. The instructions sounded plausible to a couple of newcomers to
electronics, and caused them to open up their modems, and when they found a
slight match between their modems construction and the 24to96 instructions,
they jumped in and hacked away. The fact that these files were retrieved
from BBS's with solid reputations for policing the quality and authenticity
of the files gave the experimenter's confidence they would not have felt had
they retrieved the data from the National Enquirer! In addition, overly
enthusiastic uploaders have added testimonials to the explanation of the file
(maybe to enhance the joke?) such as "I've tried this and it really works!".
If they have, they should contact me, and we will start up another company to
build 9600 modems this new and revolutionary way which will surely blow all
competitors away, since the chips listed are 1/6th the cost of the next
cheapest 9600 bps chip set.
Lest you wonder why I think I know so much about this, my credentials:
degree: BS Electronics Engineering/Mathematics, AAA ranked University
20+ years engineering experience
7 years modem design experience
have designed 40+ modems using the chips in the 24to96 file
founder of three modem companies now in business
[haven't bragged that much in 20 years!]
Specific problems in "24to96" (other than horrible english and grammar):
(1) The INTEL part numbers called out are not INTEL part numbers. They are
probably manufacturer's part numbers for a particular Hayes clone modem, but,
they could just as easily be spare tires for a volvo motorcycle!
(2) The SC11005 chip called out in the instructions is also not an INTEL
number; it is a SIERRA number. Specifically, it is a SIERRA V.22 bis modem
filter (don't recall all the exact specs).
(3) They instruct you to remove the SIERRA SC11005 chip and replace it with
an SC11020. In the first place, the SC11020 is essentially a custom
microprocessor (not a filter of any sort), in the second place, the SC11020
has about twice as many pins as the SC11005! Not to mention it uses
different power supply connections and voltages and may possible blow up if
you could get it rigged in just right. An experienced technician would
notice that right off, but again, some have just enough experience to feel
they can do this, but do not realize what is going on until they have already
hacked the SC11005 (or some other part) out of the modem (if they even FOUND
one in the first place!).
(4) To the best of my knowledge, there is no way to upgrade a V.22bis modem
to any kind of 9600 bps modem no matter what you do. It is not a matter of
"bandwidth".
(5) I tried unsuccessfully for 4 years to buy chips from US Robotics. Why
would they sell cheap upgrades instead of expensive HST modems?
(6) "Optional Lap-M error correction".
This is the only part of the instructions that won't hurt your modem,
though it is definitely NOT V.42 anything. You may recognize this as the
"noise filter" circuit explained in a number of BBS files over the past
couple of years.
Since Lap-M error correction is a SOFTWARE function, it goes without
saying that you cannot add Lap-M error correction to a modem simply by adding
a series RC network between tip and ring on the phone jack!
To add Lap-M error correction requires the addition of software requiring
thousands of man hours to write and debug, and a microprocessor capable of
executing the software in real time. At 9600 bps this is no small
requirement!
(7) A 9600 bps modem requires support circuitry at least 10 times better
(higher quality, lower distortion, better echo-canceling, etc.) than 2400 bps
modems. Dropping a 9600 bps modem chip set into a 2400 design would barely
function and would probably fail miserably in actual use, provided of course
such a thing were even possible.
**************************************************
Now, if you had already figured out this file was a joke, Congratulations!
You are officially designated a technical sophisticrat! But, remember, most
modem users are not technical and will not be able to see through this as
easily as you did. If two people who knew me and could call for help have
ruined modems over this irresponsible document, how many others did also, and
could not call for help? (BTW, another guy mentioned to me he was going to
try it and I talked him out of it--that makes three I know of). This is a
reckless joke, and I consider it a "HARDWARE VIRUS" document; those who
originate and promote such things are no better than those who write and
spread viruses.
If you had not seen through this document, I am glad you took the time to
download and read this. Please, NO MATTER HOW MUCH SENSE THIS ALL MAKES TO
YOU, DO NOT TRY TO DISASSEMBLE AND UPGRADE YOUR MODEM USING THESE
INSTRUCTIONS! In fact, even if you managed to add V.42bis to your modem
somehow, unless you purchase a license from IBM, UniSys and British Telecom
to use it, IT WOULD BE PIRATED SOFTWARE AND YOU WOULD BE A SOFTWARE PIRATE.
V.42/bis is owned and licensed by IBM, UniSys and British Telecom; their
license fees are substantial and are included in the cost of a V.42/bis
capable modem.
If you still want to try it, at least find a friend or solicit the help of a
student or ??? They won't be able to do the upgrade either, but maybe THEY
can convince you this is baloney!
I don't usually "go public" this directly. When I first read this file I
thought it was intended to be serious. When I read it again I knew it was
obviously a joke. When people started taking it seriously and suffering from
it I became angry. I ought to actually publish instructions for upgrading
from 2400 bps to 9600 bps. I think it would parallel the recipe I once
learned for cooking common lake Carp (a "trash" fish in the U.S.):
Clean the Carp
Lay it out flat on a redwood shingle
Generously season with salt, pepper and dill
Place in a 400 degree oven and cook for 3 hours
To eat, throw away the carp and eat the shingle!
The modem instructions would be:
Remove all interior components from your 2400 bps modem
Retain the case, power switch and lights
Put the 9600 bps sub-section you purchased into the case
Screw it down and connect up the switch and the lights
Throw out the 2400 bps modem
Well, that's about it. I've done my bit for the modem world, saved it from
a terrible fate, right?
Well, maybe not...............
Larry Holmes
Wednesday, November 21, 2007
How to Connect a PC to a VideoCrypt Decoder
How to connect a PC to a Videocrypt decoder
-------------------------------------------
1994-04-11
The easiest way to connect your PC to a decoder is to use the card slot
as an interface and connect it with a voltage converter (MAX232) and a
TTL open collector driver (74LS07) to the RS-232 serial port. This way,
you don't even have to open the decoder.
WARNING: In order to build the adapter described below, you will at
least require some digital electronics experience. If
you don't understand, how the described circuitry works,
better don't use it! Errors might in the worst case damage
both your PC and your TV system.
The chip cards used by the Videocrypt pay-TV decoders follow exactly
the specification ISO 7816 (you might find this international standard
in a local library, if you are interested). Also, the protocol is the
asynchronous half-duplex T=0 protocol with active low reset and inverse
convention as defined in the standard.
According to ISO, a chip card is 85.60 mm long, 53.98 mm high, 0.76 mm
thick and the edges are rounded with a radius of 3.18 mm. It has eight
defined contact areas (C1 - C8 in the diagram below), each of which is
at least 2 mm wide and 1.7 mm heigh:
______________________________________
/ \
| |
| |
| C1 C5 |
| C2 C6 |
| C3 C7 |
| C4 C8 |
| |
| |
| |
\________________________________________/
These contacts have the following purpose:
C1 VCC Supply voltage (+5 V, max. 200 mA)
C2 RST Reset signal
C3 CLK Clock signal
C4 - reserved
C5 GND Ground
C6 VPP Programming voltage
C7 I/O Data input/output
C8 - reserved
The following table gives the precise location of the contact areas.
These areas are only minimum areas, the actual contacts might be larger
but must of course be properly isolated from each other.
In the following table,
A represents the maximum distance between the card's left
edge and the contact area's left edge,
B represents the minimum distance between the card's left
edge and the contact area's right edge,
C represents the maximum distance between the card's top
edge and the contact area's upper edge,
D represents the minimum distance between the card's top
edge and the contact area's lower edge.
A B C D
-----------------------------------------
C1 10.25 12.25 19.23 20.93
C2 10.25 12.25 21.77 23.47
C3 10.25 12.25 24.31 26.01
C4 10.25 12.25 26.85 28.55
C5 17.87 19.87 19.23 20.93
C6 17.87 19.87 21.77 23.47
C7 17.87 19.87 24.31 26.01
C8 17.87 19.87 26.85 28.55
Older card systems had these contacts located higher (distance from the
top between 9.07 mm and 18.39 mm, distance from the left identical). As
some decoders support both contact area alternatives, make sure that
this old contact area is properly isolated or you'll produce a short
circuit when inserting your card. You might have noticed, that the
contacts are arranged in the usual 0.1 inch (= 2.54 mm) system (i.e.
like the pins of a 8-pin DIL chip).
You can produce your card adapter by making a PCB with contact areas at
the above listed locations. The PCB must have precisely the thickness
and width of a real card, but it may be longer, so that you can locate
the interface electronics on the part which remains outside the slot.
Cards are inserted in most decoders with the contacts on the bottom
side, but check this on your system. Normal PCBs are about 1.3 mm think
and won't fit into the slot. Either you get a PCB which is about 0.8 mm
thick or you make it thinner, e.g. by using a sander machine. Perhaps
you find also ready to use test cards with connectors instead of
producing your own or you simply open the decoder and clamp contacts to
the resistors near the card slot (not recommended: there are unisolated
230 V parts inside the decoder, this may kill you if you are not very
carefull!!!).
The adapter will only need the card contacts I/O, GND, RST and VCC. On
the RS-232 side, only the following contacts will be used:
Sub-D 25-pin Sub-D 9-pin
---------------------------------------------------------
TxD 2 3 transmit data
RxD 3 2 receive data
CTS 5 8 clear to send
DSR 6 6 data set ready
GND 7 5 ground
DCD 8 1 carrier detect (here: reset)
DTR 20 4 data terminal ready
The pins DTR, DSR and CTS are not actually needed, they are just
connected together in the adapter, so that defined levels are available
on them because some software might need this.
The following components are necessary for the adapter
1 PCB or test card
1 IC Maxim MAX232
1 IC 74LS07 (or only a 7407)
4 capacitors 22 uF
1 female Sub-D connector (9 or 25-pin)
The MAX232 converts the RS-232 levels (about +10 and -10 V) to TTL
voltage (0 and +5 V) and vice versa without requiring anything else
than +5 V power supply. This chip contains two TTL->RS-232 and two
RS-232->TTL drivers and needs four external 22 uF capacitors in order
to generate the RS-232 voltage internally. The adapter electronic gets
its power supply from the decoder's VCC line or you can use an external
5 V supply if you wish.
The card slot's RST line is connected using one of the TTL->RS-232
drivers in the MAX232 to DCD, so that the software and the decoder can
easily resynchronize in case of a protocol error.
The I/O line is a bidirectional half-duplex asynchronous TTL level
serial port that is operated in a Videocrypt system with 9600 bits/s.
We can connect this line to a MAX232 TTL input driver (which is
connected to RxD and sends bytes to the PC) in order to receive data
from the decoder. The TxD line's signal is converted in the MAX232 to
TTL level and is connected with an open collector TTL driver to I/O.
This open collector driver (one of six in the 74LS07) has a high
impedance output during idle state and 1 and is connected to GND during
a 0 on it's input. As there is already a pull-up resistor to +5 V on
I/O in the decoder, this circuitry guarantees, that the adapter is in
high impedance state if the TxD line is idle and delivers the correct
voltage if the PC sends bytes and the decoder is in reception mode. As
we don't connect totem-pole or tristate outputs to I/O, a short circuit
should be impossible in the adapter.
The following diagram describes the whole interface:
+-------------+
+-----------|1 V 16|----o +5V (VCC)
+| +| |
=== +5V o-||-|2 MAX232 15|----o GND (card & RS-232)
| | |
+-----------|3 +---14|----o DCD +-<-o DTR
+ | | | |
+---||---|4 | +-13|- +->-o DSR
| | | v | |
+--------|5 | +-12|- +->-o CTS
+ | | |
GND o-||-|6 +-<-11|----o RST
| |
RxD o----|7 ---<--- 10|-------------------+----o I/O
| | |\ |
TxD o----|8 --->--- 9|--------------| |--+
+-------------+ 1|/ 2
74LS07
(also connected to 74LS07:
pin 7=GND, pin 14=VCC)
Pay attention to the polarity of the capacitors (marked with a + in the
diagram next to each capacitor)!
As a side effect of this simple interface design, every byte sent by
the PC is at the same time also received by the PC. Consequently, you
can test the circuit with a terminal emulator by switching of local
echo: if you still see every typed immediately character on the screen,
the interface should be all right. Software must be capable of dealing
with this echo from the interface. As specified in the ISO standard,
the decoder activates VCC only shortly before a reset and deactivates
VCC if an answer-to-reset packet isn't received in time after the reset
signal. If no external 5 V supply is used, the software might have to
wait a few milliseconds after the reset before starting with the
answer-to-reset, in order to allow the capacitors to load up and
provide a stable operation of the MAX232.
A few final hints:
If you have a larger distance between the PC an the decoder, then
locate the adapter electronic near the decoder, because the RS-232
interface is much more suitable for long cables than the TTL signals.
Cables of 12 m length have sucessfully been used and you shouldn't have
problems with RS-232 cables up to 15-30 m length. (If you need much
longer cables, you should use RS-422 line drivers, e.g. the Am26LS31
and Am26LS32 from AMD, which allow over 1 kilometer cable length.)
You can also use this adapter circuit to allow a PC to listen to the
data traffic between a decoder and a real card. Just connect the real
card and the adapter parallel to the decoder and don't let the PC
software transmit anything. Suitable card slots are available for
little money from various manufacturers (e.g. Amphenol). Videocrypt
uses the inverse convention data format, i.e., you have to reverse and
invert the bits in each byte in the PC software in order to get the
correct byte value. For more details about the protocol, check ISO
7816-3.
There are many alternative ways to build this interface. E.g. instead
of a MAX232, an LT1081 from Linear Technology could be applied or the
74LS07 could be replaced by two open collector inverters in the 74LS05
and a 2.2k pull-up resistor between them, etc.
Normally, both the RS-232 and the decoder slot should not be harmed by
short circuits, but be careful. Also try to avoid electrostatic voltage
(e.g. generated by walking on a suitable carpet) near the interface,
because discharges cause easily decoder or PC crashes and could
theoretically even harm the hardware (the MAX232 is a CMOS chip as are
some of the chips in the decoder). And please doublecheck everything I
have written here, because I don't want to be responsible if anything
goes wrong just because I wrote something wrong. DON'T USE THIS DESIGN
IF YOU DON'T UNDERSTAND IT.
-------------------------------------------
1994-04-11
The easiest way to connect your PC to a decoder is to use the card slot
as an interface and connect it with a voltage converter (MAX232) and a
TTL open collector driver (74LS07) to the RS-232 serial port. This way,
you don't even have to open the decoder.
WARNING: In order to build the adapter described below, you will at
least require some digital electronics experience. If
you don't understand, how the described circuitry works,
better don't use it! Errors might in the worst case damage
both your PC and your TV system.
The chip cards used by the Videocrypt pay-TV decoders follow exactly
the specification ISO 7816 (you might find this international standard
in a local library, if you are interested). Also, the protocol is the
asynchronous half-duplex T=0 protocol with active low reset and inverse
convention as defined in the standard.
According to ISO, a chip card is 85.60 mm long, 53.98 mm high, 0.76 mm
thick and the edges are rounded with a radius of 3.18 mm. It has eight
defined contact areas (C1 - C8 in the diagram below), each of which is
at least 2 mm wide and 1.7 mm heigh:
______________________________________
/ \
| |
| |
| C1 C5 |
| C2 C6 |
| C3 C7 |
| C4 C8 |
| |
| |
| |
\________________________________________/
These contacts have the following purpose:
C1 VCC Supply voltage (+5 V, max. 200 mA)
C2 RST Reset signal
C3 CLK Clock signal
C4 - reserved
C5 GND Ground
C6 VPP Programming voltage
C7 I/O Data input/output
C8 - reserved
The following table gives the precise location of the contact areas.
These areas are only minimum areas, the actual contacts might be larger
but must of course be properly isolated from each other.
In the following table,
A represents the maximum distance between the card's left
edge and the contact area's left edge,
B represents the minimum distance between the card's left
edge and the contact area's right edge,
C represents the maximum distance between the card's top
edge and the contact area's upper edge,
D represents the minimum distance between the card's top
edge and the contact area's lower edge.
A B C D
-----------------------------------------
C1 10.25 12.25 19.23 20.93
C2 10.25 12.25 21.77 23.47
C3 10.25 12.25 24.31 26.01
C4 10.25 12.25 26.85 28.55
C5 17.87 19.87 19.23 20.93
C6 17.87 19.87 21.77 23.47
C7 17.87 19.87 24.31 26.01
C8 17.87 19.87 26.85 28.55
Older card systems had these contacts located higher (distance from the
top between 9.07 mm and 18.39 mm, distance from the left identical). As
some decoders support both contact area alternatives, make sure that
this old contact area is properly isolated or you'll produce a short
circuit when inserting your card. You might have noticed, that the
contacts are arranged in the usual 0.1 inch (= 2.54 mm) system (i.e.
like the pins of a 8-pin DIL chip).
You can produce your card adapter by making a PCB with contact areas at
the above listed locations. The PCB must have precisely the thickness
and width of a real card, but it may be longer, so that you can locate
the interface electronics on the part which remains outside the slot.
Cards are inserted in most decoders with the contacts on the bottom
side, but check this on your system. Normal PCBs are about 1.3 mm think
and won't fit into the slot. Either you get a PCB which is about 0.8 mm
thick or you make it thinner, e.g. by using a sander machine. Perhaps
you find also ready to use test cards with connectors instead of
producing your own or you simply open the decoder and clamp contacts to
the resistors near the card slot (not recommended: there are unisolated
230 V parts inside the decoder, this may kill you if you are not very
carefull!!!).
The adapter will only need the card contacts I/O, GND, RST and VCC. On
the RS-232 side, only the following contacts will be used:
Sub-D 25-pin Sub-D 9-pin
---------------------------------------------------------
TxD 2 3 transmit data
RxD 3 2 receive data
CTS 5 8 clear to send
DSR 6 6 data set ready
GND 7 5 ground
DCD 8 1 carrier detect (here: reset)
DTR 20 4 data terminal ready
The pins DTR, DSR and CTS are not actually needed, they are just
connected together in the adapter, so that defined levels are available
on them because some software might need this.
The following components are necessary for the adapter
1 PCB or test card
1 IC Maxim MAX232
1 IC 74LS07 (or only a 7407)
4 capacitors 22 uF
1 female Sub-D connector (9 or 25-pin)
The MAX232 converts the RS-232 levels (about +10 and -10 V) to TTL
voltage (0 and +5 V) and vice versa without requiring anything else
than +5 V power supply. This chip contains two TTL->RS-232 and two
RS-232->TTL drivers and needs four external 22 uF capacitors in order
to generate the RS-232 voltage internally. The adapter electronic gets
its power supply from the decoder's VCC line or you can use an external
5 V supply if you wish.
The card slot's RST line is connected using one of the TTL->RS-232
drivers in the MAX232 to DCD, so that the software and the decoder can
easily resynchronize in case of a protocol error.
The I/O line is a bidirectional half-duplex asynchronous TTL level
serial port that is operated in a Videocrypt system with 9600 bits/s.
We can connect this line to a MAX232 TTL input driver (which is
connected to RxD and sends bytes to the PC) in order to receive data
from the decoder. The TxD line's signal is converted in the MAX232 to
TTL level and is connected with an open collector TTL driver to I/O.
This open collector driver (one of six in the 74LS07) has a high
impedance output during idle state and 1 and is connected to GND during
a 0 on it's input. As there is already a pull-up resistor to +5 V on
I/O in the decoder, this circuitry guarantees, that the adapter is in
high impedance state if the TxD line is idle and delivers the correct
voltage if the PC sends bytes and the decoder is in reception mode. As
we don't connect totem-pole or tristate outputs to I/O, a short circuit
should be impossible in the adapter.
The following diagram describes the whole interface:
+-------------+
+-----------|1 V 16|----o +5V (VCC)
+| +| |
=== +5V o-||-|2 MAX232 15|----o GND (card & RS-232)
| | |
+-----------|3 +---14|----o DCD +-<-o DTR
+ | | | |
+---||---|4 | +-13|- +->-o DSR
| | | v | |
+--------|5 | +-12|- +->-o CTS
+ | | |
GND o-||-|6 +-<-11|----o RST
| |
RxD o----|7 ---<--- 10|-------------------+----o I/O
| | |\ |
TxD o----|8 --->--- 9|--------------| |--+
+-------------+ 1|/ 2
74LS07
(also connected to 74LS07:
pin 7=GND, pin 14=VCC)
Pay attention to the polarity of the capacitors (marked with a + in the
diagram next to each capacitor)!
As a side effect of this simple interface design, every byte sent by
the PC is at the same time also received by the PC. Consequently, you
can test the circuit with a terminal emulator by switching of local
echo: if you still see every typed immediately character on the screen,
the interface should be all right. Software must be capable of dealing
with this echo from the interface. As specified in the ISO standard,
the decoder activates VCC only shortly before a reset and deactivates
VCC if an answer-to-reset packet isn't received in time after the reset
signal. If no external 5 V supply is used, the software might have to
wait a few milliseconds after the reset before starting with the
answer-to-reset, in order to allow the capacitors to load up and
provide a stable operation of the MAX232.
A few final hints:
If you have a larger distance between the PC an the decoder, then
locate the adapter electronic near the decoder, because the RS-232
interface is much more suitable for long cables than the TTL signals.
Cables of 12 m length have sucessfully been used and you shouldn't have
problems with RS-232 cables up to 15-30 m length. (If you need much
longer cables, you should use RS-422 line drivers, e.g. the Am26LS31
and Am26LS32 from AMD, which allow over 1 kilometer cable length.)
You can also use this adapter circuit to allow a PC to listen to the
data traffic between a decoder and a real card. Just connect the real
card and the adapter parallel to the decoder and don't let the PC
software transmit anything. Suitable card slots are available for
little money from various manufacturers (e.g. Amphenol). Videocrypt
uses the inverse convention data format, i.e., you have to reverse and
invert the bits in each byte in the PC software in order to get the
correct byte value. For more details about the protocol, check ISO
7816-3.
There are many alternative ways to build this interface. E.g. instead
of a MAX232, an LT1081 from Linear Technology could be applied or the
74LS07 could be replaced by two open collector inverters in the 74LS05
and a 2.2k pull-up resistor between them, etc.
Normally, both the RS-232 and the decoder slot should not be harmed by
short circuits, but be careful. Also try to avoid electrostatic voltage
(e.g. generated by walking on a suitable carpet) near the interface,
because discharges cause easily decoder or PC crashes and could
theoretically even harm the hardware (the MAX232 is a CMOS chip as are
some of the chips in the decoder). And please doublecheck everything I
have written here, because I don't want to be responsible if anything
goes wrong just because I wrote something wrong. DON'T USE THIS DESIGN
IF YOU DON'T UNDERSTAND IT.
Subscribe to:
Posts (Atom)