RS232 Shield issues

Hello,

I’m running into what I think is a mismatch between actual port settings on the FEZ Domino and what’s realized on the RS232 Shield that’s attached to it.

In plugging the RS232/Domino equipment to my PC, I get consistent and reliable communications. When I remove the DB9 from the PC and attach it to my TNC, I can’t receive any data, but I do see data transmitted. When the TNC is connected to the PC, I can receive and transmit reliably and consistently.

Further, with a small oscilloscope attached, I see absolutely no signal coming from the TNC when its attached to the RS232 Shield from GHI Electronics. I do see signal when it’s attached to the PC.

Technical Summary:

  1. TNC has only 3 wires, GND, RX and TX. Signal is standard RS232 ±12V DC swing on communication signal.

  2. TNC communicates great with PC - no issues. O-Scope shows standard ±12VDC RS232 signals both ways.

  3. PC to Shield/Domino working both ways, no issues. DB9 connector straight-thru with all 9 pins.

  4. TNC Receives data FROM Domino/Shield and transmits it over the air. However, Shield/Domino not seeing any data from TNC when received by TNC. O-Scope shows no data incoming.

This is rather puzzling - I tried calling GHI to see if the RS232 shield is hardware strapped for Hardware RTS/CTS or DTR/DSR flow control, but they directed me here ???:-\

Anyone got a clue by four?

In the RS232 world there are two types of equipment; DCE and DTE.

DCE == Data Communications Equipment (modem)
DTE == Data Terminal Equipment (PC)

You can connect a DCE to a DTE with a normal RS-232 cable.

I suspect that the TNC and the RS-232 shield are both DCEs.

To connect them, you will need a modem eliminator cable. This is a cable with crossed wires instead of the straight through wires of the normal cable.

Is called null modem. You can find it for under $1 on amazon

There is no flow control on most RS232 shields as most UARTs aren’t presented with full pins. On a Domino if you NEED flow control you can use UEXT.

Having said that, first I am not sure what your TNC device is - more info there might help, I assume we’re talking Terminal Node Controller?

Often times, UART signal issues, whether shifted to RS232 or not, are caused by lack of GND connectivity - make sure that’s a good connection, check the cables have good connector integrity etc.

Have you confirmed the signal looks correct on each side of the RS232 chip? If you can see it on the input (TNC output) of the chip but you can’t see it on the output (Fez) side, then the chip is not detecting the signal change correctly. If you can track multiple channels on your o-scope, can you see signal on both sides at once ?

Hi Guys,

@ Mike: Yes, I’m aware of DTE/DCE issues and already have a null modem cable connected between the Shield/Domino and the TNC (Terminal Node Controller, which is also a DCE.)

@ Brett: Yes, I continuity tested the ground between radio/TNC and the FEZ/Shield. That was my first assumption…seeing that the (Kenwood TH-D7A) Radio/TNC is battery operated. However, note that there IS TX signal from the domino/shield and I see the data train on the scope going TO the radio/TNC. If the grounds weren’t there, I wouldn’t get a TX signal.

Recap:
Domino -> Shield -> TNC Transmit = GREAT
TNC -> Shield -> Domino Receive = Nope

TNC -> PC = Great both ways
Domino -> = Great both ways

If I had no null modem connection, neither would work. Pins are good, continuity is good. It’s really looking like I need CTS to be tied high…(running to lab to try)

ADDITIONAL INFO:

  1. Tying CTS to DTR didn’t accomplish anything, as DTR isn’t high when the RS232 Shield turns on.
  2. Interesting to note that when the Domino transmits a packet to the TNC, the TNC echoes it back to the Domino and the data is received. If you leave all the cables plugged in, and remove the TNC, this doesn’t happen. (The pins are not shorted together in or out of the TNC - the TNC must be on for the echo to come back to the RS232/Domino.)

…Waiting for Rod Serling to show up.

Giving up on the RS232 shield. I suspect there’s another type of issue going on here after reviewing the pin-outs for a typical MAX232 chip implementation.

I’ll use signal level conversion and drive the UEXT pins. Dollars to donuts I can get it to work there. I think there may be more than one reason why the RS232 shield has been discontinued… ???