Problems you have or had with RS232?

[gar]

051214-1057 EST USA

What kind of RS232 problems have you previously had, and/or currently have? RS232 here can be read to mean many things. For example in the following list I mention dropping a laptop.

I can not do this with checkoff boxes so a written response is necessary. Define your machine, control, and application.

Some example areas of problems:

1. Cable length
2. Transmission data errors
3. CNC's maximum baud rate is too low
4. Have to operate at a baud rate lower than CNC machine's maximum capability
5. Ground loop noise causing errors
6. Noise generated within the CNC causing transmission errors
7. CNC is starved for data in drip mode (DNC) on short strokes
8. Circuit damage from elcectrical storms, or electrical faults, or ground problems
9. Pinouts for wiring computer to CNC
10. Lack of understanding how RS232 works, and related problems
11. Sending multiple files from computer to CNC
12. Noise from welders or plasma cutters
13. Difficulty of putting CNC in send or receive mode
14. Time out problems
15. CNC Program file management
16. Having to go back and forth between computer and CNC
17. Need to run CNC programs larger than available memory in the CNC
18. Need to do simultaneous drip feeds
19. Damaging your laptop because you dropped it
20. Format of file being sent to CNC is not accepted by CNC
21. Handshake problems
22. Need for delays between CNC program blocks

.

[gar]

051222-1037 EST USA

Here is a real world problem one of our customers had:

There are two VF-3 machines, side by side 5 ft apart, one has DC servos and the other has brushless servos. Separate RS232 cables about 100 ft long to each machine connected from a central computer. 25 pin connectors at both ends with pin 1 connected to pin 1 and the same for pin 7. Each machine was grounded to the service entrance thru and overhead bus bar system. Also each machine additionally had its chassis connected to a ground rod at the CNC machine.

If the servos were off, then one could communicate in either direction with either or both machines from the central computer.

The machine with brush type DC servos presented no problem when the servos were on, and thus you could use DNC mode (drip feed). Cable length did limit them to 9600 baud. Also this machine had no problem when the adjacent machine with brushless servos had servos turned on.

The machine with brushless servos worked fine in every respect except when its servos where on it was not possible to communicate in either direction to the central computer. The obvious cause was the noise from the brushless drives. This noise was sufficiently high to introduce errors into the RS232 connection.

We installed our I232 Isolator System on both machines and increased the customers baud rate to 115.2 kbaud from 9600, a 12 x increase in speed, with its advantages, and eliminated the noise problem. No other changes were made to their setup. We did not remove the supplementary ground even though I do not think it is a good idea to use it. Details on ground and noise are discussed under Noise and Grounding on our website. Thus, it was possible to make an unuseful machine into a useful one.

Note that high baud rates, like 115.2 kbaud, generally eliminate starvation in drip mode.

.