[General] Communication failure with MAX232

[Mithun_K_Das]

I'll measure the voltages and check if there are any abnormalities. Other issues can be checked.

MAX232 ICs are dumb, but there is no easy alternation in RS232 communication IC rather than MAX232. If there any, maybe it would perform better than MAX232.

 

[betwixt]

I'm staring into my crystal ball and seeing only the usual fog but my guess is the problem isn't the RS232 interface at all. If it stops working but isn't apparently faulty, by elimination the problem has to be somewhere else.

I have two guesses:
1. the data to/from the PC is powered up before the MAX232 has had chance to bring it's charge pumps up to full voltage and some kind of reverse bias or latch up is occurring, this would be made worse by increasing the capacitor values.

2. The data lines are connecting before the ground pin and causing a similar problem. The PCB shows the mounting points of the socket are connected to the ground plane so the shell of the plug should ground before the actual ground pin if the socket has a metal body!

I would check there is a ground connection all the way between the PCB ground and the shell of the plug at the PC end of the cable, before plugging it into the PC (that might create an alternative path). Many RS232 cables leave the shells floating and as the ground pin on an RS232 'D' is the same length as the other pins, there is no guarantee it will connect before the data pins.

Many times I have taken the precautionary measure of opening the 9-pin 'D' connectors on the cable and soldering a short link between pin 5 and the shell so I can be sure it connects before any other pins touch.

Brian.

 

[Mithun_K_Das]

I'm staring into my crystal ball and seeing only the usual fog but my guess is the problem isn't the RS232 interface at all. If it stops working but isn't apparently faulty, by elimination the problem has to be somewhere else.

I have two guesses:
1. the data to/from the PC is powered up before the MAX232 has had chance to bring it's charge pumps up to full voltage and some kind of reverse bias or latch up is occurring, this would be made worse by increasing the capacitor values.

2. The data lines are connecting before the ground pin and causing a similar problem. The PCB shows the mounting points of the socket are connected to the ground plane so the shell of the plug should ground before the actual ground pin if the socket has a metal body!

I would check there is a ground connection all the way between the PCB ground and the shell of the plug at the PC end of the cable, before plugging it into the PC (that might create an alternative path). Many RS232 cables leave the shells floating and as the ground pin on an RS232 'D' is the same length as the other pins, there is no guarantee it will connect before the data pins.

Many times I have taken the precautionary measure of opening the 9-pin 'D' connectors on the cable and soldering a short link between pin 5 and the shell so I can be sure it connects before any other pins touch.

Brian.

When the power is on, the PC need time to start up. But the control circuit needs only seconds. Is this an issue?

The metal cases are connected to the GND panel. The PC body is metal and screws with the body of the CDM. But there is a separate GND connection with its body.

 

[betwixt]

When the power is on, the PC need time to start up. But the control circuit needs only seconds. Is this an issue?

It shouldn't be. The power to the PC RS232 almost certainly comes directly from PSU +/-12V lines whereas the MAX232 end needs a moment for it's charge pump to reach full voltage. The data from the PC can therefore have a voltage on it before the MAX232 is ready but to the best of my knowledge, the charge pump is only used to transmit data FROM the MAX232 so it shouldn't be an issue.

Check there is a ground connection between the metal shields on the connecting cable, just a short circuit check from the outside to the plug/socket from one end to the other. As all the pins are the same length in a 'D' connector it is possible from one or both of the data pins to connect before the ground pin and that opens a risk of reverse voltages or even static discharge. Most RS232 cables do not connect the shields together but they inevitably make contact to the other end before the data or ground pins can connect so it is wise in most circumstances to link pin 5 (for 9 pin connectors) to the shield at both ends if it isn't already done.

Brian.

 

[KlausST]

Hi,

MAX232 ICs are dumb, but there is no easy alternation in RS232 communication IC rather than MAX232. If there any, maybe it would perform better than MAX232.

There is the ESD protected version: MAX232E
If I´m not mistaken, then there are pin compatible ICs from other manufacturers. I´m not sure.

Klaus

 

[BradtheRad]

There may be security in PC software. But we still not sure what type of security is used here as the person who created this PC software is no longer in our company now.

Many computer programs create a Preferences file. Is there a chance the programmer included a security routine, which stores a running history of all encounters, and flags the devices which were unsuccessful or suspect? So it rejects future communications with any device which once gave it a sour taste?

Does the program have a command so you can delete the history of past encounters?

When programs misbehave a useful strategy is to delete the Preferences file. It is not necessarily easy to locate. You may have to search for it.
* It might be in the same directory as the program or it might be in another directory.
* It might end with 'ini' or 'txt' or something else.
* It might have a recent creation or modification date.

If all else fails you might get success by reinstalling the program. Of course that's a typical recommendation heard from help lines everywhere.

 

[Mithun_K_Das]

I checked the SMPS, ripple voltage is only 80~90mV which is pretty noise-free. Then I modified the circuit diagram of the MAX232 associated part as well as the relay and motor part to reduce EMI.

MAX232 part:

And for the motor part: added series resistor to limit current.


The result is better than the previous one. The fault is reduced significantly. Sent 25 units to test in the field. Let's check the feedback.