More RS232 issues

Projects General Guidance
1

Hi all,

I'm attempting to communicate with a digital balance using the RS232 protocol. I have the appropriate cable from the manufacturer. I have been able to test this on a computer using a USB -> RS232 adapter (9600 baud, no flow control, 1 stop bit, no parity, 8-bits) and a hyperterm-like program called gtkterm. I get good input and output. I have also purchased an RS232 to TTL logic level board that uses the MAX3232 IC. http://www.mikroe.com/downloads/get/1509/max3232_manual_v100.pdf

I am attempting to interface with a Due which takes 3.3V TTL signals (thus the decision to go with the MAX3232). So I wire up the board 3.3V from the Due to VCC, Gnd to Gnd, RX to RX1, TX to TX1, which is Serial1 on the Due (very similar to a Mega in this respect.) I wrote a simple sketch to echo anything received from the MAX3232 board to the Serial port for debugging and anything sent to the Serial port to be transmitted to the MAX3232 board. Serial1 begins at 9600 baud.

I connect the MAX3232 board up to my USB to RS232 adapter. Using gtkterm with the same settings as I used for the digital balance, and I can send and receive characters between gtkterm and the serial monitor. I can also short pins 2 and 3 on the DB9 port on the MAX3232 board and get a loop back to the serial monitor.

In order to connect the digital balance to the MAX3232's DB9 port, I have to use a gender bender because the digital balance is female and the MAX3232 DB9 is female. So I do that, and I get no love. No input or output.

No problem says I, I will simply switch the TX and RX going into the Due. No love yet again.

Ah ha, the new element is the gender bender! I grab the multimeter and check the pins on the gender bender. 1 goes to 1, 2 to 2, etc all the way to 9. So there is good integrity there and no surprises.

To summarize:
Balance to computer, check.
MAX3232 to computer, check.
Loopback through the Due, check.
Loopback through the MAX3232 board, check.
Balance to MAX3232, no good.
Swap the TX1 and RX1 on the Due, still no good.
Gender bender integrity, check.

Needless to say, I am stumped. Am I missing something here?

2

Your RS-232 module is connected as a DCE (let's say slave, pin 2, RXD, is going to T1out, while pin 3, TXD, is going to R1in), so it cannot fulfill the PC role. I don't know if you can re-route that or if you have to buy another module that uses the MAX232 the other way around.

3

I think I see what you're saying. There's a 1K resistor and a connection to VCC on the T1in side which is why I can't just flip the wires on the Arduino side of things. I will attempt to reroute the DB9 connection.

4

You can use what's called a null-modem cable, that's doing this for you.

5

I also thought of that, but it would add more cabling than I would like and the need for yet another gender bender. It is no wonder USB took over with headaches like this.

6

It is no wonder USB took over with headaches like this.

You have the same headaches with USB (you have a master and some slaves there too), it's just more clear there because you have different connectors for the master (USB-A) and slave (USB-B) sides of the connection. But a lot of people (especially in the Arduino community) are not aware that there are differences between the two functionalities and want to control USB devices over the USB connector the Arduino has.

7

I have manually routed around the pins as discussed and have finally achieved communication. Thanks very much for your help! If I had a do-over I'd just get the MAX3232 breakout board from Sparkfun.