Mailing List Archive

At 04:50 PM 6/9/2003 -0400, John Klimek wrote:
>I built myself a DB9-to-RJ11 cable and have been having some problems.
>
>On the BACK of my DB9 connector, it appears to last the pin numbers as
>follows:
>
>5 4 3 2 1
> 9 8 7 6
>
>However, when I searched for RS232 pin assignment, I found the following:
>
>1 2 3 4 5
> 6 7 8 9

There is no one correct answer, since male and female connectors are mirror
images (and of course, the wiring side of a DE-9 is the mirror image of the
pin side ... I don't know which of these you are calling the "BACK" of the
connector or which side you are describing the pinouts with respect to).
This URL -- http://www.rhinorobotics.com/stwire9-25.html -- gives both sets
of pinouts ... viewed, as is so customary that they don't bother to say it,
from the pinout side, NOT the wiring side, of the connector.

Assuming a standard serial port at the PC end, it requires a female
connector on the cable, so the correct pinouts are the first of the two you
list above.

>However, my cable SOMEWHAT worked. It was able to change channels, but it
>didn't receive a "Success" signal after sending the command.

One interpretation of this is that you have TX wired correctly but not RX.
Since TX is pin 3, it goes to the same pin with either of your pinout
choices, but RX, pin 2, changes. In this case, you would also have SG
miswired ... I've seen odder things than a serial cable without a ground
return working, but that might cause some, as you say, "very weird results"
... especially since I don't recall what handshaking line, if any, this
cable uses.


>I rewired my cable (using incorrect (?)) using a 50" (might be longer)
>cable, and I'm getting very weird results:

I assume you mean to write 50' (50 feet), not 50" (50 inches).

50 feet is long, but not unreasonably long for a standard serial cable.
Back in the days when I used a lot of serial connections, I used cables up
to 100 feet routinely (to connect serial terminals to multiuser PCs ... I
*said* it as a long time ago). Of course, what is generally true may not be
true in a specific instance, so you might want to check on whether the
DirecTV device has some wimpier serial implementation than the standard.

>1) When plugged into one computer, the satellite receiver kept turning on
>and off every few seconds.
>2) When plugged into the other (Win XP, using DTVCON), it somewhat worked.
>It would change channels, but sometimes it would change to the wrong channel
>(is my cable too long?) and I was still unable to receiver a "Success"
>response from the satellite. I got a ton of Buffer errors.
>
>Sorry for the long post.
>
>I'm wondering if a 50" or 75" cable is too long for the cable and is my
>wiring incorrect?

Probably not too long. But if you are worried about this, test with a short
cable (cables aren't that hard to make if you know how, as you clearly do).

You don't actually *say* how you wired the cable, so I can't say whether
what you did is right.

>I'm using a continuity tester to make sure the signal is passing through the
>cable, so I know that isn't the problem.
>
>Thanks for any help. After I get my cable working complete, I'm going to
>write a very detailed HOW-TO with pictures to help out anybody else who
>might be having problems.

If you are the guy who posted on this a week or so ago, I **think** I
recall two things:

1. The DirecTV end was an RJ22, not an RJ11.

2. The Web sites about wiring this cable have at least 2, maybe 3,
suggested wirings. If you have RX, TX, and SG all wired correctly, but
whatever handshaking line the DirecTV device expects to use is wrong, that
could cause intermittent errors.

(If you post again, please include a URL for the cable-wiring diagram you
actually followed.)

http://www.pcmx.net/dtvcon/ -- this is the diagram that I tried to follow.

If you turn the phone connector (RJ22) upside (so you can see the gold
pins), I ordered them like this (top to bottom):

1 (white)
2 (white)
3 (red)
4 (black)

Then, I cut a 75' (I think its 75', it appears longer than 50') RJ11 cable
just to make my RJ22 cable longer. Now, I connected these two cables like
this:

RJ22 --> RJ11
1 - (white) --> Green
2 - (red) --> Red
3 - Not used
4 - (black) --> Black

Now, I cut an older cable that was used to connected an external modem to a
desktop. One end was DB25M and the other was DB9F. I cut off the 9-pin
female connector. It has 9 pins which are holes (I'm assuming this means
DB9F).

Following the diagram, I connected as follows:

RJ11 --> DB9F
1 - (white) --> 5 (Brown)
2 - (red) --> 2 (Red)
4 - (black) --> 3 (Orange)

I used a multimeter (to check continuity) to check which hole on the DB9F
connector went to which colored wire.

Now, before I added my "extension" (RJ22 --> RJ11), I used these hole (pin?)
assignments for the DB9F connector:

5 4 3 2 1
9 8 7 6

The cable transmited, but DTVCON always gave a "Timeout" error when waiting
for a response.

Now, I tried reversing my DB9F hole assignments:

1 2 3 4 5
6 7 8 9

The cable transmits, but again recevies a "Timeout" from DTVCON.

My "reversed" hole assignments I think I have confused with a DB9M
connector.


Sorry for this being so long but I'm trying to be as through as possible so
you could help me =)

Thank you so much.


----- Original Message -----
From: "Ray Olszewski" <ray@comarre.com>
To: "Discussion about mythtv" <mythtv-users@snowman.net>
Sent: Monday, June 09, 2003 8:06 PM
Subject: Re: [mythtv-users] DirecTV (RS232-to-RJ11 cable) -- I think I
figured out my problem!


> At 04:50 PM 6/9/2003 -0400, John Klimek wrote:
> >I built myself a DB9-to-RJ11 cable and have been having some problems.
> >
> >On the BACK of my DB9 connector, it appears to last the pin numbers as
> >follows:
> >
> >5 4 3 2 1
> > 9 8 7 6
> >
> >However, when I searched for RS232 pin assignment, I found the following:
> >
> >1 2 3 4 5
> > 6 7 8 9
>
> There is no one correct answer, since male and female connectors are
mirror
> images (and of course, the wiring side of a DE-9 is the mirror image of
the
> pin side ... I don't know which of these you are calling the "BACK" of the
> connector or which side you are describing the pinouts with respect to).
> This URL -- http://www.rhinorobotics.com/stwire9-25.html -- gives both
sets
> of pinouts ... viewed, as is so customary that they don't bother to say
it,
> from the pinout side, NOT the wiring side, of the connector.
>
> Assuming a standard serial port at the PC end, it requires a female
> connector on the cable, so the correct pinouts are the first of the two
you
> list above.
>
> >However, my cable SOMEWHAT worked. It was able to change channels, but
it
> >didn't receive a "Success" signal after sending the command.
>
> One interpretation of this is that you have TX wired correctly but not RX.
> Since TX is pin 3, it goes to the same pin with either of your pinout
> choices, but RX, pin 2, changes. In this case, you would also have SG
> miswired ... I've seen odder things than a serial cable without a ground
> return working, but that might cause some, as you say, "very weird
results"
> ... especially since I don't recall what handshaking line, if any, this
> cable uses.
>
>
> >I rewired my cable (using incorrect (?)) using a 50" (might be longer)
> >cable, and I'm getting very weird results:
>
> I assume you mean to write 50' (50 feet), not 50" (50 inches).
>
> 50 feet is long, but not unreasonably long for a standard serial cable.
> Back in the days when I used a lot of serial connections, I used cables up
> to 100 feet routinely (to connect serial terminals to multiuser PCs ... I
> *said* it as a long time ago). Of course, what is generally true may not
be
> true in a specific instance, so you might want to check on whether the
> DirecTV device has some wimpier serial implementation than the standard.
>
> >1) When plugged into one computer, the satellite receiver kept turning
on
> >and off every few seconds.
> >2) When plugged into the other (Win XP, using DTVCON), it somewhat
worked.
> >It would change channels, but sometimes it would change to the wrong
channel
> >(is my cable too long?) and I was still unable to receiver a "Success"
> >response from the satellite. I got a ton of Buffer errors.
> >
> >Sorry for the long post.
> >
> >I'm wondering if a 50" or 75" cable is too long for the cable and is my
> >wiring incorrect?
>
> Probably not too long. But if you are worried about this, test with a
short
> cable (cables aren't that hard to make if you know how, as you clearly
do).
>
> You don't actually *say* how you wired the cable, so I can't say whether
> what you did is right.
>
> >I'm using a continuity tester to make sure the signal is passing through
the
> >cable, so I know that isn't the problem.
> >
> >Thanks for any help. After I get my cable working complete, I'm going to
> >write a very detailed HOW-TO with pictures to help out anybody else who
> >might be having problems.
>
> If you are the guy who posted on this a week or so ago, I **think** I
> recall two things:
>
> 1. The DirecTV end was an RJ22, not an RJ11.
>
> 2. The Web sites about wiring this cable have at least 2, maybe 3,
> suggested wirings. If you have RX, TX, and SG all wired correctly, but
> whatever handshaking line the DirecTV device expects to use is wrong, that
> could cause intermittent errors.
>
> (If you post again, please include a URL for the cable-wiring diagram you
> actually followed.)
>
>
>
> _______________________________________________
> mythtv-users mailing list
> mythtv-users@snowman.net
> http://lists.snowman.net/cgi-bin/mailman/listinfo/mythtv-users
>

First, please do not apologize for "being long". Troubleshooting cabling
problems is always a matter of detail, and I'd much prefer to wade through
some irrelevancy than to hit "I think I got that part right" in the middle
of a report. And in facf none of the detail you provide is irrelevant, anyway.

At 08:39 PM 6/9/2003 -0400, John Klimek wrote:

>http://www.pcmx.net/dtvcon/ -- this is the diagram that I tried to follow.
>
>If you turn the phone connector (RJ22) upside (so you can see the gold
>pins), I ordered them like this (top to bottom):
>
>1 (white)
>2 (white)
>3 (red)
>4 (black)

In looking at the URL, I can't really see which way they expect you to
orient the RJ22 connector. Were I to guess, I'd make the same guess you did
... but that still eaves it a guess, albeit a shared one.

>Then, I cut a 75' (I think its 75', it appears longer than 50') RJ11 cable
>just to make my RJ22 cable longer. Now, I connected these two cables like
>this:
>
>RJ22 --> RJ11
>1 - (white) --> Green
>2 - (red) --> Red
>3 - Not used
>4 - (black) --> Black

Note that between these two descriptions, the location of the (red) wire
moved from position 3 to position 2. Is this just a typo or does it reflect
a wiring error?

If it is a wiring error ... as I read the wiring diagram at te URL you
reference, this is the pin for RX (that is, the line the PC receives data
on). Miswiring it would cause replies from the DirecTV device not to reach
the PC, no matter how you wired the DE9 end.


>Now, I cut an older cable that was used to connected an external modem to a
>desktop. One end was DB25M and the other was DB9F. I cut off the 9-pin
>female connector. It has 9 pins which are holes (I'm assuming this means
>DB9F).
>
>Following the diagram, I connected as follows:
>
>RJ11 --> DB9F
>1 - (white) --> 5 (Brown)
>2 - (red) --> 2 (Red)
>4 - (black) --> 3 (Orange)

Another notation problem. Pin 1 on the RJ11 (above) was Green, not (white).
This time, though, I assume you are just still using the RJ22 color codes,
not making a wiring error.


>I used a multimeter (to check continuity) to check which hole on the DB9F
>connector went to which colored wire.
>
>Now, before I added my "extension" (RJ22 --> RJ11), I used these hole (pin?)
>assignments for the DB9F connector:
>
>5 4 3 2 1
> 9 8 7 6

There are pins inside the holes, so I still call them pinouts. If you are
viewing the connector from the "outside", this one is the right pinout set
for a female connector.

>The cable transmited, but DTVCON always gave a "Timeout" error when waiting
>for a response.
>
>Now, I tried reversing my DB9F hole assignments:
>
>1 2 3 4 5
> 6 7 8 9
>
>The cable transmits, but again recevies a "Timeout" from DTVCON.
>
>My "reversed" hole assignments I think I have confused with a DB9M
>connector.

Me too.


>Sorry for this being so long but I'm trying to be as through as possible so
>you could help me =)
>
>Thank you so much.
[old stuff deleted]

Ok, here are my findings so far.

DB9F pin assignment is definitely as follows:
5 4 3 2 1
9 8 7 6

RJ22 pin assignment seems to (definitely?) follow: (when looking at the
bottom, pin-side)
Pin-1
Pin-2
Pin-3
Pin-4

Following the pin assignment, I've connected:
(RJ22-to-DB9F)
1 - to - 5 [ground]
2 - to - 2 [receive]
4 - to - 3 [transmit]

Here is what happens:

1) If I DO NOT connect the RECEIVE line (2 - to - 2), I can transmit to the
receiver PERFECTLY. The receive changes channels, etc, etc and works
perfectly fine. Of course I still get "a Timeout" response from the receive
because the RxD line is not connected. However it will still change
channels.

2) If I DO NOT connect BOTH the RECEIVE and the GROUND line, whenever I
transmit to the receive it will change to WEIRD channels and do weird
things.

3) If I connect ALL THE WIRES, as soon as I plug the cable into the
computer and the receive, my satellite receiver will turn itself on and off
constantly.

I believe my problem is DEFINITELY located in the receive line.

If I can transmit fine (with TxD and Gnd connected only), then I think its
safe to assume my RJ22 and DB9F pin assignments (and connections) are
definitely correct. However once I hookup the RxD line, I immediately get
problems ranging from receiver self shutdown/reboot or I get "Timeout"'s
when I send

Also, I measured my cable and it's 50' (I thought it looked longer than
that). So, if I can transmit fine then I'm thinking I have one of two
possible problems:

1) The satellite receiver is not strong enough to send a signal back to the
computer over the 50' of cable. Can I test for this any easy way? It's
hard to rewire my cables espicially if I had to chop the length back down.

or

2) The diagram is screwed up and I'm actually suppoed to connect RJ22 Pin 3
to DB9F Pin 2.

Has anybody else built a cable like this that can possibly help?

Ray, thank you so much for helping me. If you have any other ideas please
let me know. I know I'm very to getting this to work.

Thanks!



[entire original message thread follows]
> First, please do not apologize for "being long". Troubleshooting cabling
> problems is always a matter of detail, and I'd much prefer to wade through
> some irrelevancy than to hit "I think I got that part right" in the middle
> of a report. And in facf none of the detail you provide is irrelevant,
anyway.
>

Thank you very much.



> At 08:39 PM 6/9/2003 -0400, John Klimek wrote:
>
> >http://www.pcmx.net/dtvcon/ -- this is the diagram that I tried to
follow.
> >
> >If you turn the phone connector (RJ22) upside (so you can see the gold
> >pins), I ordered them like this (top to bottom):
> >
> >1 (white)
> >2 (white)
> >3 (red)
> >4 (black)
>
> In looking at the URL, I can't really see which way they expect you to
> orient the RJ22 connector. Were I to guess, I'd make the same guess you
did
> ... but that still eaves it a guess, albeit a shared one.
>
> >Then, I cut a 75' (I think its 75', it appears longer than 50') RJ11
cable
> >just to make my RJ22 cable longer. Now, I connected these two cables
like
> >this:
> >
> >RJ22 --> RJ11
> >1 - (white) --> Green
> >2 - (red) --> Red
> >3 - Not used
> >4 - (black) --> Black
>
> Note that between these two descriptions, the location of the (red) wire
> moved from position 3 to position 2. Is this just a typo or does it
reflect
> a wiring error?
>
> If it is a wiring error ... as I read the wiring diagram at te URL you
> reference, this is the pin for RX (that is, the line the PC receives data
> on). Miswiring it would cause replies from the DirecTV device not to reach
> the PC, no matter how you wired the DE9 end.
>
>
> >Now, I cut an older cable that was used to connected an external modem to
a
> >desktop. One end was DB25M and the other was DB9F. I cut off the 9-pin
> >female connector. It has 9 pins which are holes (I'm assuming this means
> >DB9F).
> >
> >Following the diagram, I connected as follows:
> >
> >RJ11 --> DB9F
> >1 - (white) --> 5 (Brown)
> >2 - (red) --> 2 (Red)
> >4 - (black) --> 3 (Orange)
>
> Another notation problem. Pin 1 on the RJ11 (above) was Green, not
(white).
> This time, though, I assume you are just still using the RJ22 color codes,
> not making a wiring error.
>
>
> >I used a multimeter (to check continuity) to check which hole on the DB9F
> >connector went to which colored wire.
> >
> >Now, before I added my "extension" (RJ22 --> RJ11), I used these hole
(pin?)
> >assignments for the DB9F connector:
> >
> >5 4 3 2 1
> > 9 8 7 6
>
> There are pins inside the holes, so I still call them pinouts. If you are
> viewing the connector from the "outside", this one is the right pinout set
> for a female connector.
>
> >The cable transmited, but DTVCON always gave a "Timeout" error when
waiting
> >for a response.
> >
> >Now, I tried reversing my DB9F hole assignments:
> >
> >1 2 3 4 5
> > 6 7 8 9
> >
> >The cable transmits, but again recevies a "Timeout" from DTVCON.
> >
> >My "reversed" hole assignments I think I have confused with a DB9M
> >connector.
>
> Me too.
>
>
> >Sorry for this being so long but I'm trying to be as through as possible
so
> >you could help me =)
> >
> >Thank you so much.
> [old stuff deleted]
>
> _______________________________________________
> mythtv-users mailing list
> mythtv-users@snowman.net
> http://lists.snowman.net/cgi-bin/mailman/listinfo/mythtv-users
>

Well ... from what you describe, I have two ideas, with the second one kind
of out on the fringe.

1. Try connecting RX to RJ22 pin 3, as you speculate.

2. Consider the possibility that you are reading the pin positions on the
RJ22 upside down. (You'll recall I agreed with your guess about how to read
this, but I also reminded you that my doing so just made it a shared guess,
not correct.) Then you have TX and SG reversed, and RX on the (supposedly)
unused pin. RS232 signalling is just wacky enough that swapping TX and SG
might still let TX work, but RX wouldn't work in this setting no matter
what you do.

If neither of these ideas pans out, I'm left with suggesting you look for
one of those old RS232 troublshooting boxes, that had red and green LEDs to
detect which pins were active. (I don't know where to find these today;
mine is probably 10 years old.)

Good luck. I'd like to see this procedure get well documented.

At 11:30 PM 6/9/2003 -0400, John Klimek wrote:
>Ok, here are my findings so far.
>
>DB9F pin assignment is definitely as follows:
>5 4 3 2 1
> 9 8 7 6
>
>RJ22 pin assignment seems to (definitely?) follow: (when looking at the
>bottom, pin-side)
>Pin-1
>Pin-2
>Pin-3
>Pin-4
>
>Following the pin assignment, I've connected:
>(RJ22-to-DB9F)
>1 - to - 5 [ground]
>2 - to - 2 [receive]
>4 - to - 3 [transmit]
>
>Here is what happens:
>
>1) If I DO NOT connect the RECEIVE line (2 - to - 2), I can transmit to the
>receiver PERFECTLY. The receive changes channels, etc, etc and works
>perfectly fine. Of course I still get "a Timeout" response from the receive
>because the RxD line is not connected. However it will still change
>channels.
>
>2) If I DO NOT connect BOTH the RECEIVE and the GROUND line, whenever I
>transmit to the receive it will change to WEIRD channels and do weird
>things.
>
>3) If I connect ALL THE WIRES, as soon as I plug the cable into the
>computer and the receive, my satellite receiver will turn itself on and off
>constantly.
>
>I believe my problem is DEFINITELY located in the receive line.
>
>If I can transmit fine (with TxD and Gnd connected only), then I think its
>safe to assume my RJ22 and DB9F pin assignments (and connections) are
>definitely correct. However once I hookup the RxD line, I immediately get
>problems ranging from receiver self shutdown/reboot or I get "Timeout"'s
>when I send
>
>Also, I measured my cable and it's 50' (I thought it looked longer than
>that). So, if I can transmit fine then I'm thinking I have one of two
>possible problems:
>
>1) The satellite receiver is not strong enough to send a signal back to the
>computer over the 50' of cable. Can I test for this any easy way? It's
>hard to rewire my cables espicially if I had to chop the length back down.
>
>or
>
>2) The diagram is screwed up and I'm actually suppoed to connect RJ22 Pin 3
>to DB9F Pin 2.
>
>Has anybody else built a cable like this that can possibly help?
>
>Ray, thank you so much for helping me. If you have any other ideas please
>let me know. I know I'm very to getting this to work.
[old stuff deleted]

Ray Olszewski wrote:
...
> If neither of these ideas pans out, I'm left with suggesting you look
> for one of those old RS232 troublshooting boxes, that had red and green
> LEDs to detect which pins were active. (I don't know where to find these
> today; mine is probably 10 years old.)

If you're in the Bay Area, try Halted Specialists (HSC) at
Lawrence and Central or, of course, Weird Stuff Warehouse.

> Good luck. I'd like to see this procedure get well documented.
>
> At 11:30 PM 6/9/2003 -0400, John Klimek wrote:
>
>> Ok, here are my findings so far.
>>
>> DB9F pin assignment is definitely as follows:
>> 5 4 3 2 1
>> 9 8 7 6
>>
>> RJ22 pin assignment seems to (definitely?) follow: (when looking at the
>> bottom, pin-side)
>> Pin-1
>> Pin-2
>> Pin-3
>> Pin-4
>>
>> Following the pin assignment, I've connected:
>> (RJ22-to-DB9F)
>> 1 - to - 5 [ground]
>> 2 - to - 2 [receive]
>> 4 - to - 3 [transmit]
>>
>> Here is what happens:
>>
>> 1) If I DO NOT connect the RECEIVE line (2 - to - 2), I can transmit
>> to the
>> receiver PERFECTLY. The receive changes channels, etc, etc and works
>> perfectly fine. Of course I still get "a Timeout" response from the
>> receive
>> because the RxD line is not connected. However it will still change
>> channels.
>>
>> 2) If I DO NOT connect BOTH the RECEIVE and the GROUND line, whenever I
>> transmit to the receive it will change to WEIRD channels and do weird
>> things.
>>
>> 3) If I connect ALL THE WIRES, as soon as I plug the cable into the
>> computer and the receive, my satellite receiver will turn itself on
>> and off
>> constantly.
>>
>> I believe my problem is DEFINITELY located in the receive line.

Well, this may sound dumb and simplistic, but here goes.

Maybe you should only connect TX and GND. If you think
of the data port as a substitute for an IR device,
there should be no out bound data from the cable box.
If, as you say, you can change channels, that may be all
you need.

If pin two had random noise, that may cause the serial port
on the Linux box to do weird things like re-negotiate baud
rate, send HUPs, long breaks or whatever. Try connecting
just the two pins and set the terminal parameters to assume
that the port is connected even though there are no signals
on RX. No HUP, no flow control, etc.

-- bjm