RS232 failure to send in XON/XOFF, hangs WAITING, thermally caused.

[gar]

050526-1913 EST USA

Has anyone had a failure of RS232 communication where:

1. Handshake is XON/OFF, the other parameters do not matter.

2. Nothing connected to RS232 connector, or even jumpering pins 4 and 5 and the voltage is +10 on 4 and 5. On HAAS 6, 8 , and 20 are do not care. Also in XON/XOFF pin 5 (CTS) is a do not care.

3. Put cursor on a program.

4. Press SEND. This should send the file.

5. What happens is that HAAS displays the message WAITING. This should not occur in XON/XOFF, but it should in RTS/CTS if pin 5 was negative.

6. What this tells me is that the input chip that receives the CTS signal is not at fault because the WAITING message never occurs in normal XON/XOFF operation.

7. The problem occurs when the circuitry is warm and cooling in the processor board area removes the problem. Has anyone else experienced this failure, and what was the specific cause?

.

[Al_The_Man]

If it IS a temperature condition, I would see if you can identify the RS232 Conditioning IC's adjacent to the port, if they are socketed then replacement should be easy, some recent types are the single ICL232, older IC's are 1488 & 1489 or variants of the these.
I know that hardware HS should not be necessary with XON XOFF but I always jumper 4-5 6-8-20 as a matter of course.
Al.

[miljnor]

i have had the same problem, but, It wasn't a heat issue. Some of the older haas have a bug in them. The bug causes the machine to hang with no connection and when you reset and add the connection it will always hang..

We had to power the machine off then restart to get the thing to trans or recieve.

since transmitting with no wires connected is of little usefullness this bug hasn't been a big issue.

[gar]

050527-0645 EST USA

Al and Michel:

Thanks.

Some more background. We have 5 differnet HAAS machines in our shop ranging from 1993 thru 2000. The machine in my post of yesterday is in a customer's location and is circa 1998.

All HAASs that I have looked at have pins 6-8-20 jumpered on the printed circuit board just behind their 25 pin connector. Thus, no need to externally jumper these. However, for Fanuc these must be externally jumpered. Therefore we make our cables with 6-8-20 jumpered. The reason for running the test without a cable was to isolate the problem.

HAAS uses a 100 ohm resistor in series with each input/output line except 1, 6, 8, and 20. Pin 1 goes directly to chassis ground. Note pin 7 is fed thru a 100 ohm resistor just like the other I/O pins. On the machine side of each 100 ohm resistor is a bi-directional transient limiter to machine ground ( same as pin 1 ). This provides good protection as you will probably burn up the 100 ohm resistor before failure of the transient limiter.

I have tested all 5 of our machines for operation with nothing connected to the RS232 connector to verify that in XON/XOFF and DC modes that no signal is required at pin 5 ( CTS ). The reason for testing our machines was that when I was at our customer site that a HAAS factory tech told me that CTS was active in XON/XOFF and DC mode and thus required a high signal. He was wrong for two reasons. First, at our customer site +10 v on pin 5 had no effect on the problem. Second, on our 5 machines XON/XOFF or DC modes work fine with any combination of pin 5 open or +/- 10 v applied.

Michel - on one of our machines I probably have the problem you describe that RS232 is totally hung until powered down. This occurs following power up. At the customer site it was possible to recover by cooling, and resetting while not powering down.

Al - your suggestion was my original thought. However, we should never get the WAITING signal in XON/XOFF in any state of pin 5. Thus, I think it is further into the system.

My contact with this all resulted because the customer thought they had a communication noise problem and wanted to install our I232 system to break a ground loop.

But the problem was not noise it was a HAAS problem. They bought the I232 system anyway because it allowed them send files at 115.2 kbaud and greatly reduce their load time.

.