Monitoring Rs232 data

[Tha fios agaibh]

I'm trying to diagnose a serial communication problem between a host computer and client printer. It's a RS232 full duplex connection from a DB9 com port on the PC through a crossover cable to the DTE printer. All connections and com settings look normal.

I thought I'd try reading the data stream coming from the PC with my laptop running Putty terminal software. When I tried connecting my female to female crossover cable, I get nothing at all on the screen. I know I'm using the right com port because it passes a loop back test.

I'm wondering if I've got the right cable for this, or I'm doing something else wrong?

 

[Bluejets]

Don't know if crossover is essential.........
see here......
https://www.lammertbies.nl/comm/cable/rs-232-serial-printer

 

[Tha fios agaibh]

Thanks Bluejets. I know for sure the printer needs a crossover cable but perhaps I should try a straight through cable since this test is really going to my laptop?

 

[Harald Kapp]

Try a software debugging approach as for example the Free Serial Port Monitor. As fas as I understand it the tool analyzes the data being exchanged over an existing COM port without the need for additional hardware.
Only drawback: The website mentions only Windows older and up to 8.1, I have no idea whether it will work on Windows 10 .
I found another port monitor here.

 

[Tha fios agaibh]

Thanks Harald, but I should be able to monitor the serial port with putty. I'll try a straight through cable next.

I also considered making up a sniffer 3 connector cable.

 

[Harald Kapp]

Of course its up to you. But out of sheer curiosity: why does it have to be putty?

 

[hexreader]

First thing is to check that wiring is correct, ideally using breakout box with LED indications. Connect One side at a time and check that no LED is lit for both sides, and that every expected LED is lit on one side or the other. i.e. no output pin is driving another output.

The cheaper way to do this is to use a multi-meter on Volts range to check that no pin shows voltage on both sides. e.g. pin 3 should be approx -6 to -12 Volts (lets call it -9V) on PC connector, but open circuit on device side (assumes D9 connector). Pin 2 should be -9V (ish) on device connector, but open circuit (less then 1V typically) on PC side. This test should be done with connectors unplugged.

Connect common lead of multi-meter to pin 5 of the connector being tested (assuming 9-pin D connector)

Same goes for handshaking signals if used. Always plus or minus 9V (ish) one side and floating other side, for each pin.

Next level of diagnostic requires oscilloscope. Do you own one?

Putty or similar only tells you when you have a working link. It is unlikely to give meaningful clues on robustness, and is a very poor tool to diagnose non-working links.

 

[Tha fios agaibh]

No luck with straight through cable either.

I'm open to other software, it's just a pain getting approval installing it on a company computer.

 

[hexreader]

Software is a very poor tool for diagnosing bad wiring. It is generally only useful for verifying good wiring, and even then can give misleading results if you do not fully understand the limitations.

 

[Tha fios agaibh]

Software is a very poor tool for diagnosing bad wiring. It is generally only useful for verifying good wiring.

Thanks hexreader. I like your idea of using a breakout box with LEDs.

I just wanted to use a terminal program on my laptop just to verify that the data is being transmitted from the host computer. Not even to necessarily analyze it.
Once transmission is verified, the problem is most likely on the printer side.

Cabling is about 25ft and I'm not too worried about its integrity. I've verified continuity and it passed a loopback test.

I don't want to get into the level of analyzing with a scope. Jjust seeing a blip of the Tx,Rx, leds would probably suffice in order to isolate the problem.

If I dont see the lights blink its likely a software or com port problem on the PC.

 

[hexreader]

" I've verified continuity and it passed a loopback test."

Be aware that this test will pass even if baud rate is completely wrong or if Tx/Rx is crossed

 

[Tha fios agaibh]

True, except i would see the crossed Tx/Rx by checking continuity.

I believe I never saw the data with my laptop because the proper bits weren't set such as clear to send.
If I use a spy cable I presume it would then work.

I still like the idea of a breakout box. I'll either make or buy one.