BRX Built-in Serial Port to ASCII Device Hanging Up on Power Up

[Mike Nash]

I have seen some problems with failure to resume communications after PLC power down and back up. Switching from Remote to Stop, Run (and back to remote) will restart the comms again. I thought initially I had resolved the issue by unchecking the Program Retentive box. It got better but still hangs sometimes. By hanging up, I mean no TX RX LEDs on BRX and I don't remember seeing the program stepping through the stages.

RS-232 BRX to FA-Isocon.
RS-485 4-wire with common FA-Isocon to ASCII device.
Grounds good except FA-Isocon has none. No wires over 3 feet. Zero comms issues except on power up.
Stage programming for the communications in MyComms.
Main has ST1 Run MyComms program.
StreamIn and StreamOut in Stage - only one stage active at a time.
Works perfectly except sometimes after power cycling.

I don't know exactly where the hangup is, I had too many other things I had to get worked out first and time ran out for now.

My general question is, what's the best way to get the MyComms program and serial port all happy on power up?

It may be a race type situation based on what powers up when (separate circuit breakers for the PLC, FA-Isocon and 3rd party device), but I'm not sure.

Sorry this is so disjointed, but I am working off of memory only at this point.

[BobO]

I am unaware of any issue like you are describing, but that doesn't they don't exit. We do use custom serial comms extensively in house, as all of the manufacturing testing for BRX is built with BRX, and those test sets are frequently power cycled.

If the receive side isn't coded correctly, you can definitely get stuck. Is the BRX the master or slave? Slave can be a bit tougher, but garbage data can mess with either side, especially on startup. If you want to send us your project (support@hosteng.com), we're happy to look through it.

[Mike Nash]

Thanks Bob. I wasn't trying to imply a software/firmware/hardware issue. I suspect I am not handling things correctly. Since the system is all packed up and ready to ship I can't play with it any longer at the moment.

BRX is master.

I have exactly the same setup on my desk, except all the components are different (cookie cutter, don't you know) that recovers just fine. But it is an H2-DM1E, F2-Unicon and same family of the third party ASCII comms device, but different. It recovers every time just fine. I'll see if I can't strip some of the extraneous logic out of the program for the BRX and send it your way. It will probably be tomorrow though.

Thanks.

[BobO]

In most cases things like this aren't bugs, but I've been in this business too long to not allow that possibility.

The most common issues I've had with serial comms are subtle logic errors that allow the state machine to get hung somewhere. The worst kind are the ones that work 99%, but in the event of an error, or garbage data, or whatever, it's goes off in the weeds and hangs. Reframing and error recovery in serial comms can be frustrating.

[ADC Product Engineer]

In most cases things like this aren't bugs, but I've been in this business too long to not allow that possibility.

The most common issues I've had with serial comms are subtle logic errors that allow the state machine to get hung somewhere. The worst kind are the ones that work 99%, but in the event of an error, or garbage data, or whatever, it's goes off in the weeds and hangs. Reframing and error recovery in serial comms can be frustrating.

I nearly 30 years of doing PLC's, this is by far the most common thing.  It's nearly always that one bit that wasn't included, or the extra one that wasn't supposed to be, or a reset that didn't or ...  It's almost always a logic error.  I wish I could say that I don't make them anymore or that I don't spend hours in the weeds when the error is right in front of my face, but I do.  I do tend to blame myself first anymore and the hardware second.  I will spend more time now trying to prove that it isn't my error than I ever did fixing stuff in the field. 

To err is human.  To really foul things up you need a computer.