HP48
|
|
Null Modem Adapter
|
|
Gender Changer
|
|
Modem
I am using Term2 to communicate, but its not working correctly. What is
wrong???
dave
>HP48
> |
> |
>Null Modem Adapter
> |
> |
>Gender Changer
> |
> |
>Modem
>
>I am using Term2 to communicate, but its not working correctly. What is
>wrong???
The cable that should be used to connect the serial port of your computer
to the modem is a serial cable with 1-1 connections.
You are using a null modem cable, effectively switching TX and RX around.
Naturally it will not work. You might get lucky by hooking up two
null-modem cables in tandem, as this would cancel out the switch of TX and
RX.
Better luck next time
Pete
In article <54a8s7$7...@newsbf02.news.aol.com>, pma...@aol.com (Pmaage) writes:
> The cable that should be used to connect the serial port of your computer
> to the modem is a serial cable with 1-1 connections.
> You are using a null modem cable, effectively switching TX and RX around.
Hold on a bit, let's review:
Unless I manage to get this all backwards, my recollection is:
o To clarify what equipment takes what role, RS-232 devices are classified
as either DTE (Data Terminal Equipment) or DCE (Data Communications Eqp.)
o Modems are DCE (communications), terminals are DTE (terminals).
Connecting DTE <-> DCE is "straight thru" (no crossover) between
RX: data received by modem (from the phone line),
going thru connecting cable Modem (DCE) -> Terminal (DTE)
TX: data transmtd by modem (into the phone line),
going thru connecting cable Modem (DCE> <- Terminal (DTE)
Connecting DTE <-> DTE or DCE <-> DCE requires a "crossover"
between TX and RX, so that when one device "talks" the other "listens".
The computer (PC) acts like a terminal, hence is DTE.
The HP48 can directly talk to a PC without "crossover,"
so the HP48 is configured as DCE, which is the same as a modem.
For the HP-48 (DCE) to talk to a modem (DCE) rather than
to a computer (DTE), it would appear to need an extra "crossover"
as provided by a "null modem" adapter, no?
The other reasons that modems balk at communicating are that they
may expect a DTR (Data Terminal Ready) signal; if the "null modem"
routes DSR (Data Set Ready) back into DTR, then this may already
be taken care of, but if not, other responses have suggested how
to tell your modem to stop looking for DTR.
Another approach to getting TX vs. RX correctly connected is to
try it one way; if it doesn't work, insert or remove one "null modem"
(or reverse the wiring) and try again; this practical approach
eliminates the need for mind-boggling abstract thinking :)
Getting back to the original question, see whether the computer
can talk both to the modem and to the HP48; when both of those
are working with the same parameters (baud rate, etc), then
try HP <-> modem and see what happens.
The next time a new cabling arrangement is designed, *all* equipment should
be designed so that any "female" connector means "I am listening on pin 2
and talking on pin 3" and any "male" connector means the reverse; all
"male" <-> "female" extension cables should connect "straight through"
and any "male" <-> "male" or "female" <-> "female" connectors or cables
should "cross over" (no wisecracks invited); then it will be impossible
for any cabling arrangement to fail to communicate properly, unlike people.
-----------------------------------------------------------
With best wishes from: John H Meyers ( jhme...@mum.edu )
>> The cable that should be used to connect the serial port of your
computer
>> to the modem is a serial cable with 1-1 connections.
>> You are using a null modem cable, effectively switching TX and RX
around.
>
>Hold on a bit, let's review:
>
>Unless I manage to get this all backwards, my recollection is:
>
> o To clarify what equipment takes what role, RS-232 devices are
classified
> as either DTE (Data Terminal Equipment) or DCE (Data Communications
Eqp.)
Right
> o Modems are DCE (communications), terminals are DTE (terminals).
Right
> Connecting DTE <-> DCE is "straight thru" (no crossover) between
> RX: data received by modem (from the phone line),
> going thru connecting cable Modem (DCE) -> Terminal (DTE)
> TX: data transmtd by modem (into the phone line),
> going thru connecting cable Modem (DCE> <- Terminal (DTE)
>
> Connecting DTE <-> DTE or DCE <-> DCE requires a "crossover"
> between TX and RX, so that when one device "talks" the other
"listens".
Right
> The computer (PC) acts like a terminal, hence is DTE.
> The HP48 can directly talk to a PC without "crossover,"
Aha, can it ? Why is my HP cable crossing TX and RX, then ?
> so the HP48 is configured as DCE, which is the same as a modem.
>
> For the HP-48 (DCE) to talk to a modem (DCE) rather than
> to a computer (DTE), it would appear to need an extra "crossover"
> as provided by a "null modem" adapter, no?
No.
That's why I wrote he should either use a straight-through connection
or two nullmodems which would effectively cancel each other.
If you have a G-series manual, look up page 27-8:
Do you note that the arrow outbound from the HP is labelled RX ?
The RX is the designation on the PC side.
On the HP-side the data is outbound, and that's TX to me.
I think I made an error in this when I sent somebody the
pin out of the HP, sorry for any inconvenience caused.
CU
Pete
dave
JHM> The HP48 can directly talk to a PC without "crossover," so the HP48
JHM> is configured as DCE, which is the same as a modem. [...]
How do you know that there's no crossover in the cable ?
JHM> The next time a new cabling arrangement is designed, *all* equipment
JHM> should be designed so that any "female" connector means "I am
JHM> listening on pin 2 and talking on pin 3" and any "male" connector
JHM> means the reverse; all "male" <-> "female" extension cables should
JHM> connect "straight through" and any "male" <-> "male" or "female" <->
JHM> "female" connectors or cables should "cross over" (no wisecracks
JHM> invited);
Did you ever notice that this is exactly how serial ports are defined ? Any
modem (DCE) comes with a female connector - any terminal (DTE) with a male
connector. All male/female cables (like the computer-modem connection) are
wired straight through and all male/male connections require a nullmodem
But you won't find many female/female connections since it's highly improbable
that two DCEs manage to talk to each other... ;)
JHM> then it will be impossible for any cabling arrangement to fail to
JHM> communicate properly, unlike people.
As reality shows... ;)
Problem is that your above argumentation is quite right for equipment using
standard serial connectors. Alas the HP48 has a simple 4pin connector far from
any standards for serial ports so you can't come up with "nullmodem" or
anything related. It's just a cable where RxD meets TxD and whether a
or not is needed simply depends on which side of the HP48 you are facing... ;)
--
joan.
>ok, so can anyone tell me what wires to switch and how to connect
>everything? Do i need any more adapters??
OK, what do you already have ?
I'm assuming a modem and a HP48.
Now you probably have a cable to attach the HP to a PC
and a cable to attach the modem to a PC, right ?
Now you need a nullmodem adapter
and a way to plug these things into each other.
DB9 to DB25 converters, gender changers,...
It works, just the terminal emulators on the HP are, umm,...,
well, they still need development.
CU
Pete
In article <19961024185...@rhein-neckar.netsurf.de>,
jo...@rhein-neckar.netsurf.de writes:
> How do you know that there's no crossover in the cable ?
You can't connect the HP48 to anything else until you have first
plugged in its cable, which finally brings the HP48's serial
connections out to a DB9 connector. At that point, the HP48 is clearly
configured as DCE (the same as a modem), so you need a "crossover"
(such as a "null-modem adapter") to connect the HP48 to a modem.
I can not assign any meaning to the phrase "crossover in the cable"
because the HP48 connector end follows no standard and does not connect
to any other equipment through any standard cabling until its wires
come out into a standard connector, at which point it takes on the
meaningful characteristics of DCE or DTE, according to which of the
two standard signal pins of the connector it "talks" and "listens" on.
I believe you said the same thing youself towards the end of your post;
it's just that you (and Pete Maage) are looking at the HP48 by itself
and trying to decide whether the HP48 is then DTE or DCE, but at that
point it has only a non-standard connector, so this has no meaning
(I will address more of Pete Maage's viewpoint later on).
Once you plug in the cable, the HP48 with cable is now something that
can connect to other serial devices using standard connectors, and if
you want to know which way the HP48 acts for this purpose, you'd better
consider the HP48 to be acting like a modem (DCE), even though there's
no telephone line coming out the other side of it!
> JHM> The next time a new cabling arrangement is designed, *all* equipment
> JHM> should be designed so that any "female" connector means "I am
> JHM> listening on pin 2 and talking on pin 3" and any "male" connector
> JHM> means the reverse; all "male" <-> "female" extension cables should
> JHM> connect "straight through" and any "male" <-> "male" or "female" <->
> JHM> "female" connectors or cables should "cross over" (no wisecracks);
[then any two devices would automatically always connect properly]
Joan wrote:
> Did you ever notice that this is exactly how serial ports are defined ? Any
> modem (DCE) comes with a female connector - any terminal (DTE) with a male
> connector. All male/female cables (like the computer-modem connection) are
> wired straight through and all male/male connections require a nullmodem.
I must disagree with you about this "definition"! "Any terminal" does not
always come with a male connector; they do if they are IBM PC clones or
DEC VT-100's, but they don't if they are Wyse or other general brands;
the choice of connector gender is arbitrarily made by manufacturers,
but there is no choice about which pin number has to be which signal.
The fact that neither "gender changers" nor "male-male" cables include a
"crossover" as a standard feature should be a clue to the fact that serial
cables and connectors do not in any way adhere to my suggestions above.
If, by the way, a female connector implies "like a modem," then one look
at the connector on the HP48 cable should tell you that the HP48 is itself
"like a modem" (which happens to be true, but not because of this fact).
The meaning of the pin numbers for the DB25 or DB9 connectors is unrelated
to whether they are male or female; all DB25's meeting RS-232 standards
must assign TX to pin 2 and RX to pin 3, whether male or female, while
all DB9's, male or female, must have RX on pin 2 and TX on pin 3.
[DB9's were actually invented for a totally different RS-422 signalling,
but we are speaking here about when they are used instead for RS-232]
Now we come to another point of possible disagreement: some of us like to
think that TX and RX are "relative" terms, according to whether the device
closest to this connector is "talking" or "listening" on that wire; however,
in the CCITT (I think) definition, TX means that side of the conversation
which a modem would be transmitting to the *telephone* line, and RX means
that side of the conversation which a modem would be receiving from the
*telephone* line. Note that by the time the demodulated signals end up
at the DB25 or DB9 connector on the *other* side of the modem, it turns
out that TX is the *input* to the modem from the terminal, and RX is
the *output* from the modem to the terminal! From the terminal's
point of view, TX is once again the terminal's output (from the keyboard),
and RX is the terminal's input (to be displayed on the screen). However,
if the same connector type is being used throughout all these devices,
the pin numbers are not assigned according to whether the device talks
or listens on that wire, nor what gender the connector is, but according
to a constant *absolute* reference to the direction of travel of data
on the *phone*line* side of the *modem*; because of this, some devices
can be plugged directly into each other, while others require a
"crossover" to be inserted between them, causing confusion
(plus cost and endless frustration).
If you have a G/GX User's Guide, turn to around page 27-7 and gaze at
HP's diagram of the HP48 cable connections. HP has labeled as TX the
line on which the HP48 listens, and as RX the line on which the HP48
talks. So which way is the HP48 behaving: like a terminal, or like a modem?
Those devices on which TX (at the DB25 or DB9) is the device's *input*
are in fact like modems (DCE), so the HP48 behaves just like one of those;
it needs a "crossover" (e.g. a null-modem adapter) to talk to another
DCE (modem or HP48), but no such adapter or crossover to talk to
a terminal, a PC, or a serial printer (all of which are DTE).
The "serial cable adapter kit" which HP sold for use with modems and
printers included two separate adapters: one for modem, one for printer;
it would have been impossible (without a switch) to use one adapter for both.
There may also be different extra connections in these adapters, but they
must basically differ in whether TX,RX are passed straight thru or crossed.
Now, the newsgroup FAQ, and even Donnelly (in 1st printing of 2nd edition
of "HP48 Handbook") seem to manage to get "TX" and "RX" mixed up at times
(Donnelly in the bottom illustration of the HP48 cable end on page 216);
the words "input" and "output" are sometimes written on diagrams without
saying from which device's point of view we mean these words (and which
of the two "sides" of the device in the case of a modem); illustrations
of connectors sometimes don't make clear whether you are looking into
the end of the cable or into the end of what the cable plugs into!
All of the above can be sources of confusion; the only absolutely clear
illustration I have seen is the one in the G/GX User's Guide, where you
see exactly what connects to what, you see arrows on the wires to show
signal direction (talker->listener), and if only the Mac connector had
been illustrated as well, everyone would have been taken care of.
The "RJ" series of telephone connectors and cables come close to being the
kind of connector arrangement I suggested earlier; any properly assembled
male-male cable does in fact reverse all the (2, 4, 6, or 8) wires, and
so does any female-female connector ("gender changer") between two cables,
so that any arrangement with "like" ends is a "crossover" while any
arrangement with "opposite gender" ends is "straight thru."
You will note that complete circuits are always defined by a pair of wires
which map to each other by a crossover, e.g. the "middle pair" or "outer pair"
of wires, or one of the pairs in between (for 6- and 8-wire versions).
However, there is no "signal ground" or "shield" on a common 4-wire cable;
you could use 6-wire cable, wasting one whole pair on each of the above,
but this would use more wires than should be necessary. The real virtue
of the "RJ" cabling system was to maintain correct *polarity* (which
wire is *positive* and which is *negative*) in telephone circuits,
which used to be more important than it is today, whereas with
*serial* circuits we are instead interested in which wire is for
talking and which for listening, with an extra wire being required
for a common "signal ground," and yet another for "shield."
If Dave Arnett had allowed a 4-wire "RJ" connector to be stuck on the
serial port of the HP48, don't you suppose that the "signal ground"
and "shield" wires might also have become reversed whenever
TX and RX were reversed, just by the nature of these connectors?
This is one more reason why a 4-wire "RJ" might not have been
a wise choice for the HP48.
Now that everyone is well-grounded in RS-232 theory, the practical side
remains as always: "If at first you don't succeed, just reverse pins 2 & 3";
like much of what we learned in class, we can pretty much forget the theory
as far as practical life is concerned :)
>ok, so can anyone tell me what wires to switch and how to connect
>everything? Do i need any more adapters??
read the manual. you need 3 wires. RD TD GND.
tom
## You keep pressing you car's cigarette lighter button, wondering
## why a missile doesn't blast the jerk in the car ahead of you ?
If you want to connect the HP to a MODEM you need to make a null modem
cable. The reason is that the HP cable is a null modem and you need to
cross again the cables so the serial one is straight
(1 null modem + 1 null modem = 1 straight cable).
Also you may have to short pins 6 and 20 (if using a DB25 connector).
The pin 6 and 20 are DTR and DSR, modems usually need those signals
and the HP doesn't provide them. You have to short them in the connector
that plugs into the modem.
I tried it and it works!
73's de Francesc, EB3AMV
>If you want to connect the HP to a MODEM you need to make a null
>modem cable. The reason is that the HP cable is a null modem and
>you need to cross again the cables so the serial one is straight
>(1 null modem + 1 null modem = 1 straight cable).
right, wrong! if you want to connect the HP to a modem, you DO NOT need
a null-modem cable, since the modem is already a modem. you need a null
modem INSTEAD of a modem. the modems themselves change receive with transcv.
since what the first modem sends is what the second receives.
the 'normal' serial cable IS a null modem, cause there is NO real modem
in between the two HPs. you're right, when you say that two null modems
are like a straight cable, but what should your modem think of you, if
you connect the HPs 'output' with the modems 'output' ?
>Also you may have to short pins 6 and 20 (if using a DB25
>connector). The pin 6 and 20 are DTR and DSR, modems usually need
you could use some AT-commands to override that. AT&Dx AT&Cx i think.
>I tried it and it works!
no chance. i'm in modems for about 12 years. (because of the nullmodem,
with the shortening of 6,8,20,22 you're right.)
tom
## come down to my house and stick a stone in your mouth.
## (Supervixen; Garbage)