Page 1 of 1
ZX24a fault with unterminated serial cable?
Posted: 02 January 2007, 12:41 PM
by stevech
At times, I've seen the ZX24a with an 8 ft low cost store-bought serial cable exhibit a problem. When the far end of this cable is not connected, the ZX tends to intermittently get into a condition where my program's "heartbeat" LED blinker (a ZBasic task) does not blink and the application is not running. I thought perhaps the serial input was junk at times, and the application was going into a loop of some kind. But the LED blinker quits, which could mean that the operating system failed. Or that the RESET line is getting noise.
When this happens, If I power cycle the module, it often does not correct itself. If I unplug the unterminated serial cable, it always corrects the problem.
The DC input to this module is rather low, if that matters.
So I wonder if 60Hz hum is getting onto the RESET wire and into the ZX input circuit. I could look with a scope, but grounding the scope to the module will change things.
Re: ZX24a fault with unterminated serial cable?
Posted: 02 January 2007, 12:52 PM
by dkinzer
stevech wrote:[...]the RESET line is getting noise.
The ATN line does have a pulldown resistor on it so I would not suspect it to be the cause of the problems. However, on the current version of the ZX-24/a board the Rx line does not have a pulldown. You might try adding a one on pin 2 and see what effect it has. Something in the range of 4.7K to 10K ought to be sufficient.
Posted: 02 January 2007, 16:32 PM
by stevech
thanks. will do.
Posted: 02 January 2007, 17:33 PM
by DocJC
Several thoughts...
What do you mean the dc power into the module is low? Are you referring to your power supply? Speaking of this, any cold solder joints or flaky bread boards?
If your scope has two channels, you may be able to view your input signal as the difference between the two inputs, in differential mode, floating, without being tied to ground. Is the transmitting device tied to physical ground, or battery powered? Are you powering the remote device with V+ and ground via the same ground used for your RS-232 ground?
For troubleshooting RS-232 it is usually easiest to run at the slowest possible baud rate you can to get the system working, then start increasing your speed if you need to.
I think it may help to determine if you have noise being read as an unexpected signal, causing a software problem, or if there is another cause for the lock ups.
What baud rate are you using?
Last thought when you mentioned you needed to unplug the serial cable... Is either circuit getting parasitic power through the serial line / Max drivers? i.e. You are not truely powering them down unless you unplug the serial cable. This can give incredibly erratic behavior.
Are you using true RS-232 driver chips, or serial ttl signals? The former, obviously, is designed to drive the cable, while the latter is not.
Keep us posted. Good Luck.
JC
Posted: 02 January 2007, 17:37 PM
by DocJC
Recall, also, that how your read the que is important. If you specify a specific number of characters to read in, GetQueue will not return until it gets them all! Did you lock the task for any purpose? this can also make this appear to be locked up.
JC
Posted: 03 January 2007, 8:45 AM
by stevech
Thanks DocJc... The problem I'm having is simple - My (large) program runs fine for weeks on end - until I unplug the far end of the serial cable from a PC. Some time later, the VM hangs up as I described. Probably noise induced by the long "antenna" and the lack of RS232 in the ZX.
I'm sure I can achieve a work-around as Don suggested.
The DC input to the ZX's regulator is less than I'd like, due to the source of DC that I'm using - common to another part of the system.