I'd rather check the setup twice instead of blindly swap these signals.klricks wrote:Try swapping TX & RX from GPIO to the RS232 board.
You don't mention which Operating System is running on your RPi, or whether you have disabled the default console session which runs on the GPIO uart under Raspbian.why does the USB rs232 converter work perfectly and not the GPIO one?
With latest USB driver update (released approx. three weeks ago) the USB communication seems rock-solid. I performed some freehand stress testing by transferring large files in parallel via Ethernet and WiFi + communicating to one USB I/O board with several millions query-response data packets without a single error.fhood wrote:I hope that it will have a more reliable connection than the USB.
I wrote about native USB communication, not wrapped communication established on top of it (eg. your RS-232).fhood wrote:I still see weird errors ...THe errors are one in 500 runs of the script I am using but they are still there.
In this context it is an abbreviation for 'oscilloscope' - a device that gives a trace similar to the writer's avatar.fhood wrote:I am not entirely sure what you mean by scope.
fhood wrote:I acknowledge that wiring problems are usually at fault but as I mentioned I can talk successfully to myself by wiring pins 2 and 3 on the tty to rs232 board. I have also switched tx and rx multiple times. The device in question (I just talked to the guy who built it) does not require or use hardware flow control.
rx to rx
vcc to 3V3
tx to tx
ground to ground
I am running rasbpian, which I downloaded in 2013 and have updated regularly.
The hardware in question is a set of 8 relays connected to a processor that interprets hex commands (baud rate: 38400). I would like to get this working because I hope that it will have a more reliable connection than the USB. I use these pis for long term system automation.
Assuming the changing the default console uart refers to removing sections of the cmdline.txt, and the other file (cant remember what its called) than yes I have done that properly. AS I MENTIONED I can communicate with myself, so it sends and receives properly on some level.
fhood wrote:I tried crossing the rx tx pins by hand and that showed no improvements so I doubt a null modem will help.
For a loopback test (TX-RX) the local ground is used. It doesn't indicate if the local ground is connected to the remote ground.fhood wrote:Yup remember I can send and receive data when I connect pin 2 to pin 3 so the connections are correct insofar as I can discern.
I didn't think there was a difference between a null modem and swapping RX/TX in your case. There are no handshake lines so swapping TX/RX should be the same as using a null modem.fhood wrote:SOLVED: needed a null modem, turns out there is a difference between crossing the rx and tx wires and a null modem. Props to whoever suggested it.
ON THE RS232 side, there is no difference between either swapping TX/RX (pin 2 &3) or adding a NULL modem adapter.joan wrote:I didn't think there was a difference between a null modem and swapping RX/TX in your case. There are no handshake lines so swapping TX/RX should be the same as using a null modem.fhood wrote:SOLVED: needed a null modem, turns out there is a difference between crossing the rx and tx wires and a null modem. Props to whoever suggested it.