RS-232 Trouble! Ancient Dyna Mechtronics Mill.

[AndrewZ]

I am trying to get RS-232 communication working on an ancient Dyna Mechtronics EM-3116 mill. The headache is excruciating, help would be lovely!

The controller's RS-232 port is a 9 pin male port, and according to the manual the pinout is the "standard" wiring where the data pins 2, 3 are crossed and 5 is ground. For hardware handshake ports 7, 8 are also crossed. In theory with software handshake connecting only three pins should be enough to connect.

A male USB-RS232 connector and some cables are used to connect the pins.

After hours of tinkering and making sure all the COM settings and the cabling is correct, no results. No data is received on the PC running CNCLink software. I checked the controller's RS-232 port with volt meter, and between the ground and data pins there is nothing. Between some other ports there is 14V. There is some voltage between the data pins of the connectors with ground connected, but below 5V. I wonder if it is possible that the pinout is not correct on the manual.

Are there any troubleshooting steps to confirm whether the port is dead or could the cabling or something else still be set incorrectly?

 

[tc227]

I am trying to get RS-232 communication working on an ancient Dyna Mechtronics EM-3116 mill. The headache is excruciating, help would be lovely!

The controller's RS-232 port is a 9 pin male port, and according to the manual the pinout is the "standard" wiring where the data pins 2, 3 are crossed and 5 is ground. For hardware handshake ports 7, 8 are also crossed. In theory with software handshake connecting only three pins should be enough to connect.

A male USB-RS232 connector and some cables are used to connect the pins.

After hours of tinkering and making sure all the COM settings and the cabling is correct, no results. No data is received on the PC running CNCLink software. I checked the controller's RS-232 port with volt meter, and between the ground and data pins there is nothing. Between some other ports there is 14V. There is some voltage between the data pins of the connectors with ground connected, but below 5V. I wonder if it is possible that the pinout is not correct on the manual.

Are there any troubleshooting steps to confirm whether the port is dead or could the cabling or something else still be set incorrectly?

Your probably going to have to go on ebay and buy a rs-232 jump box with red and green lights on it to see if your voltage changes + or -
You might have to ditch the USB and go with a PC. USB only has 5 volts.
When you do get it working then you can try the USB connection.

 

[AndrewZ]

Your probably going to have to go on ebay and buy a rs-232 jump box with red and green lights on it to see if your voltage changes + or -
You might have to ditch the USB and go with a PC. USB only has 5 volts.
When you do get it working then you can try the USB connection.

Thank you for input.

I will try connecting with PC port, but should the voltage mismatch matter?

Volt meter is inadequate to see the voltage changes? What I find suspicious is that the data lines are at zero voltage, while on the USB-RS232 connector one of the data line does have signal voltage. The voltages changes by very small amount on the controller when an upload is performed.

 

[gustafson]

so you are using a usb to serial converter, not just a cable, correct?

 

[AndrewZ]

so you are using a usb to serial converter, not just a cable, correct?

Yes. I am using a USB to serial (male) on PC side, and small connector cables to connect each pin on the male - male setup.

 

[JBTR]

It's not clear exactly what testing you've done so far, and there are a number of places that things could go wrong.
For starters, have you verified the USB-RS232 converter itself as functional?
To do that, you can jumper its RS232 pins 2 - 3 then run a terminal program like Putty. In this configuration, when you type in a character, you should see it the character echoed on your local terminal. If you disconnect that jumper, then you'll no longer see the local echo.
That can verify that your adapter and local driver is working.
But then there can still be other issues with the interface to the controller, regarding voltage levels and the various required signals.
Let me know the results, and then let's continue...

(Edit for clarity: for the above check, you should be completely disconnected from the CNC controller.)

John

 

[AndrewZ]

It's not clear exactly what testing you've done so far, and there are a number of places that things could go wrong.
For starters, have you verified the USB-RS232 converter itself as functional?
To do that, you can jumper its RS232 pins 2 - 3 then run a terminal program like Putty. In this configuration, when you type in a character, you should see it the character echoed on your local terminal. If you disconnect that jumper, then you'll no longer see the local echo.
That can verify that your adapter and local driver is working.
But then there can still be other issues with the interface to the controller, regarding voltage levels and the various required signals.
Let me know the results, and then let's continue...

(Edit for clarity: for the above check, you should be completely disconnected from the CNC controller.)

John

I have confirmed that the USB-RS232 converter is functional with a other controller (same PC). Both receive and transmit works and are on pins 2-3 as intended.

This said I bridged the "usual" pins, at least 4 and 6.

 

[JBTR]

OK great, so you have another controller? Is it the same make/model, or something else?

RS232 uses both positive and negative voltage levels to indicate logic states. You mentioned that both data lines are zero, and that doesn't seem good. The Tx data line should be some + or - voltage level, somewhere from 3 to 15 V, but not at zero during operation. A multimeter or voltmeter isn't very good at examining the lines when data is flowing because the changes are too fast to make any sense to the meter. So these measurements are best made at idle state.

Do you have any other indication that the controller is properly powered up and ready to run?

 

[gustafson]

baud rate?
error messages?
Tried sending from the control?

 

[AndrewZ]

OK great, so you have another controller? Is it the same make/model, or something else?

RS232 uses both positive and negative voltage levels to indicate logic states. You mentioned that both data lines are zero, and that doesn't seem good. The Tx data line should be some + or - voltage level, somewhere from 3 to 15 V, but not at zero during operation. A multimeter or voltmeter isn't very good at examining the lines when data is flowing because the changes are too fast to make any sense to the meter. So these measurements are best made at idle state.

Do you have any other indication that the controller is properly powered up and ready to run?

A different CNC machine. Anyway, the loop back works on the USB. I also have an old PC and a lengthy cable with 25 pin connector, with passing loop back.

The controller is for sure powered up, and the pins are at slightly above 11V plus or minus or at zero relative to each other. When a program send is initiated at least some relative pin voltages change on the controller by a tiny amount.

Below is a complete voltage profile when idle (controller). Minus after number indicates around above -11V, otherwise 11V, except when it's 22V. If pin not listed then zero. Repeats not listed, the polarity may not be correct and may have changed between pins.

Pin 1:
less than 10 mv to all.

Pin 2:
4, 7, 8-

Pin 3:
4, 7, 8-

Pin 4:
5-, 6-, 8- (22.45 volts!), 9-

Pin 5:
7-, 8

Pin 6:
7, 8

Pin 7:
8 (22.45 volts!), 9

Pin 8:
9-

I also did the same test on USB-RS232, and indeed it has voltage between ground and pin 3. The relative voltages between the pins are in fact quite different in comparison to the controller.

 

[JBTR]

I'm not sure exactly what is meant by the above. Are you measuring between the indicated pin to other pins? Actually you should just measure each pin to pin 5, ground.

If this:

Pin 4:
5-, 8- (22.45 volts!), 9-

means that you measured between pins 4 and 8 to get the -22 volts, then that's not a needed measurement, as every pin should be measured with respect to pin 5. But it also could mean that you were measuring an input to an output, so was the controller connected to the PC for those measurements?

I suggest starting with the controller not connected but powered up and ready to go. Then measure voltage on each pin referenced to pin 5. Compare this to the same measurements on the good controller, with both in the same state.

 

[AndrewZ]

I'm not sure exactly what is meant by the above. Are you measuring between the indicated pin to other pins? Actually you should just measure each pin to pin 5, ground.

If this:


means that you measured between pins 4 and 8 to get the -22 volts, then that's not a needed measurement, as every pin should be measured with respect to pin 5. But it also could mean that you were measuring an input to an output, so was the controller connected to the PC for those measurements?

I suggest starting with the controller not connected but powered up and ready to go. Then measure voltage on each pin referenced to pin 5. Compare this to the same measurements on the good controller, with both in the same state.

You understood correctly, all voltages are between controller and itself. I measured each pin because if for some reason the pins are not connected correctly, then pin 5 may not be ground.

So the requested measurement gives around 11V to 4,7,8, from the ground pin 5, otherwise around 0.

 

[AndrewZ]

baud rate?
error messages?
Tried sending from the control?

The baud rate is set to 2400 in both the controller and the PC. No messages of any kind on send!

Below are all the settings chosen on the controller, the same settings were used on the PC except for the port and time setting.

Port: COM1 (The value be set to 2, but no second port is visible on the controller. Setting the parameter to 2 has been tried)
Baud rate: 2400 (interestingly, no 4800 available! 9600 has also been tried.)
Data bits: 7 (only option)
Stop bits: 2 (only option)
Parity: even
Handshake: Xon/Xoff (can also be set to hardware)
Time: 400 (IIRC)

 

[JBTR]

It seems like your pin 5 is probably ground, but you can check this by measuring resistance (ohms) between pin 5 and equipment ground. Resistance measurements are best made when power is off.

I'm not sure that it's right that you're getting a logic level on both pins 7 and 8. What do you get on pins 2 and 3 when powered but not connected?

The male DE-9 connector plus crossing pins 2 and 3 indicates that the controller is wired as DTE so I expect 3 and 7 to be outputs with some logic level. 2 and 8 should be inputs.

For reference, can you try making the same measurements on your other CNC controller when powered but not connected?

 

[AndrewZ]

It seems like your pin 5 is probably ground, but you can check this by measuring resistance (ohms) between pin 5 and equipment ground. Resistance measurements are best made when power is off.

I'm not sure that it's right that you're getting a logic level on both pins 7 and 8. What do you get on pins 2 and 3 when powered but not connected?

The male DE-9 connector plus crossing pins 2 and 3 indicates that the controller is wired as DTE so I expect 3 and 7 to be outputs with some logic level. 2 and 8 should be inputs.

For reference, can you try making the same measurements on your other CNC controller when powered but not connected?

It will probably be difficult to locate equipment ground, but I will see what can be done.

All the measurements are taken when not connected (if by connected it is meant the adapters are in any way connected). The other CNC is a Female 25 pin, I can check it too. Indeed, the setup is different as in this case the connection is from DTE to DTE which is non-standard I believe. The manual says the pinout is the standard pinout with 2-3 crossed, but I wonder how trust-worthy it is.

The USB-RS232 is also a male 9 pin, so when idle from pin 5 I get voltage between 3, 4, 7 and no voltage otherwise. The controller for sure has voltage to both 7, 8 (and 4) from 5.

 

[JBTR]

Well if you keep the configuration as Xon-Xoff, that is software handshake, so if both ends agree with that, you won't need 7 & 8, shouldn't need 4 & 6.

But measuring a logic level on a non-connected (what should be an) output pin 7, I'm concerned that the controller might be damaged or wired incorrectly. Do you have any wiring diagrams, schematic, or other documentation?

Equipment ground can probably be any metal part of the controller chassis, but it seems like 5 is acting like it is ground. It would be unusual for that to be wired differently. Is the controller's 9-pin connector right on the chassis/enclosure, or is it on a cable?

Also I'll add that everything about these connections are non-standard, so don't be too concerned about needing to use the Null (2 - 3 crossed) connection. RS-232 was originally used to connect modems to terminals, so other uses are adaptations and non-standard, but generally can work well.

 

[AndrewZ]

Well if you keep the configuration as Xon-Xoff, that is software handshake, so if both ends agree with that, you won't need 7 & 8, shouldn't need 4 & 6.

But measuring a logic level on a non-connected (what should be an) output pin 7, I'm concerned that the controller might be damaged or wired incorrectly. Do you have any wiring diagrams, schematic, or other documentation?

Equipment ground can probably be any metal part of the controller chassis, but it seems like 5 is acting like it is ground. It would be unusual for that to be wired differently. Is the controller's 9-pin connector right on the chassis/enclosure, or is it on a cable?

I will try the chassis, but does this always work? The controller uses a Pentium 4 processor, so the insides may be close to a regular computer if opened.

The connector is directly on the enclosure.

 

[JBTR]

No need to open it. Usually you can find a screw or something connected to the chassis. Even the shell of the DE-9 connector.

 

[AndrewZ]

No need to open it. Usually you can find a screw or something connected to the chassis. Even the shell of the DE-9 connector.

Seems to have worked. The ground is probably pin 6 with around 0.4 ohm resistance. The rest were kilos if not mega ohms.

Pin 6 connections were: 7, 8. At least one of them should be a data pin if correct.

 

[JBTR]

I will try the chassis, but does this always work? The controller uses a Pentium 4 processor, so the insides may be close to a regular computer if opened.

The connector is directly on the enclosure.

So if the controller is actually a PC, this suggests some other places to check.

Often those ports are controlled/enabled/disabled by BIOS settings that you would need to get into at boot/power-up time. There might be settings to disable/enable. Also these settings might be retained by a small coin cell battery.

The RS232 port could be on a plug-in card, or it might be connected via a short cable that plugs into a pin header on the motherboard. You did mention that it gave you choice of COM1 or 2 with no second port available. That second port could possibly be a separate section of header pins on the motherboard or card. If you're lucky, there will be some identifying labels screened on the PC board.

I'm thinking that if this port is suspected of being electrically damaged, then maybe you can switch to the other port internally to give that a try.