G540 motor cable issues

[Mariss Freimanis]

Hi,

It's come to my attention some people aren't using the supplied DB9 mating connectors that come included with the G540 but choosing instead molded DB9 connector / cables much like what's on serial port (COM port) mouse. This presents 3 problems:

1) Undersized wires in the cable. A mouse cable is optimized for flexibility and has to carry next to zero current. The wires within are optimized for that purpose and are very flexible. This emphasis on flexibility renders the wires unable to carry more than 100mA of current before overheating.

2) A mouse DB9 connector is going to have a very low current rating per connector pin, likely to be in the 500mA range. This means the mating connector pins will overheat and may thermally damage the G540 DB9 connectors by annealing the pin sockets.

3) Using a mouse connector / cable means the current set resistor must be placed at the motor end of the cable instead of at the connector. The motor wires will inject noise into the current set resistor wires and this noise will then badly distort the motor winding waveforms. This distortion will cause the motor to run roughly and vibrate.

The 4 Geckodrive supplied mating DB9 connectors have a 5A per pin current rating, most inexpensive DB9 connectors are rated at 3A or less. The supplied connectors are also intended for mounting the current set resistor. This resistor must never be mounted on the motor end of the cable and the G540 was never designed for that being done. 22-gauge wire is recommended for cabling to the motor.

I am interested to know what people use in place of the supplied connectors. No G540 connector / cable vendor has ever submitted their product for our engineering review and I need to know what's being used out there. For instance, it may be necessary to generate a G540 circuit work-around if a lot of people are wiring the current set resistor at the motor end of the cable. I am aware of the convenience this provides if a suitable connector (5A rated) and cable (22-gauge wire or bigger) is used.

Mariss

[jsantos]

My G540 did not come with db9 connectors. How do I get these connectors?

[jsantos]

Oops. Hit submit too quickly.

I'm using cncrouterparts.com cables. I don't know if they comply with the db9 requirements of the G540. Also, my resistors are connected at the end of the motor since the db9 connector I'm using is molded.

[tulsaturbo]

I am running standard 20' db9 straight thru cables. The resistors are connected to the end of the motor leads. Having the db9 connectors at the end of the motor leads helps me swap the g540 between my two machines.

I did not know the resistor was not supposed to be put close to the motor...

Also aside from making leads, what are the options for purchasing db9 cables rated for the g540? I know cncrouterparts sells db9 cables but are there other source options that you can recommend that is rated for the g540 as well?

[strohkirchw]

Continuing this from the Taig forum, I moved my resistor on one axis to the g540 from the motor connector. I didn't notice any difference in how the motor sounded or acted but I noticed a big difference in how hot the motor got. After idling for 10min the motors with the resistor at the motor got warm to the touch while the modified cable was cool to the touch.

I'm wondering if this is why my motors get very hot after 20-30min of running a job.

Edit: 15min of idling generated a 20F difference in motor temp

[CarveOne]

Mariss,

I assumed that the wire resistance was not enough to matter if the resistor was at the motor end. Even for 26 awg wire. Is it the inductance of the wire in series on each side of the resistor that is the issue? Inductive current spikes causing arc welding at the pins?

I know that DB-9 connectors are not equal in plating quality and current handling ability, but have had my suspicions about using them for anything but low power signal transfer.

CarveOne

[Mariss Freimanis]

Originally Posted by jsantos My G540 did not come with db9 connectors. How do I get these connectors?

They are included with the G540 when you buy it from us.

Mariss

[Jeff-Birt]

I am running standard 20' db9 straight thru cables.

Wow, you're wasting a great deal of power as heat dissipation in the cables.

I assumed that the wire resistance was not enough to matter if the resistor was at the motor end. Even for 26 awg wire. Is it the inductance of the wire in series on each side of the resistor that is the issue? Inductive current spikes causing arc welding at the pins?

I'm not Mariss, nor do I play him on TV I suspect that the issue is that there can be an inductive coupling of the motor drive signal and the current programming lines causing the G540 to see a lot of 'noise'.

I know that DB-9 connectors are not equal in plating quality and current handling ability, but have had my suspicions about using them for anything but low power signal transfer.

It is important to choose a connector that fits the application at hand. The DB9 style connector has been used for many years for all sorts of applications and is available with many different ratings. What is important is to match the connector ratings to the job at hand, which Mariss has done.

[Mariss Freimanis]

Originally Posted by strohkirchw Continuing this from the Taig forum, I moved my resistor on one axis to the g540 from the motor connector. I didn't notice any difference in how the motor sounded or acted but I noticed a big difference in how hot the motor got. After idling for 10min the motors with the resistor at the motor got warm to the touch while the modified cable was cool to the touch.

If the original cable has them, the offending wires (pins 1, 5) are still there in the cable. They act as antennae inside the cable to pick up noise stong signals from the motor wires and mess-up the performance of the G540.

Mariss

[Mariss Freimanis]

Originally Posted by CarveOne Mariss, I assumed that the wire resistance was not enough to matter if the resistor was at the motor end. Even for 26 awg wire. Is it the inductance of the wire in series on each side of the resistor that is the issue? Inductive current spikes causing arc welding at the pins? I know that DB-9 connectors are not equal in plating quality and current handling ability, but have had my suspicions about using them for anything but low power signal transfer. CarveOne

These are two separate problems.

1) An inadequate wire gauge in the cable will cause it to get hot and potentially melt the cable.

2) The pin 1 and 5 wires act as antennas inside the cable and pick-up the noise from the motor wires to totally mess-up the correct current set voltage the drive needs.

I now understand the usefulness of having the resistor at the far end of the cable. Who in their right mind would ever want to solder 4 wires to a connector if offered a molded connector / cable.?

The problem is this option was never anticipated by me and the drives don't have the necessary filtering on the current set resistor inputs to cancel induced noise. In the drive design I always left open the option to quickly change the current set for reasons I may want to exercise in the future. Were the drives being used as I originally intended, this would be a non-issue.

Little bit of philosophy:

This is not a complaint. Rather it is the wonderful synergy I get from interacting with the CNC community. If I were to design stuff in a vacuum without any feedback, my designs would be what they are and no more. It is you people that use what I design in ways I had never thought of that brings new ideas I wouldn't have had before.

I'm in the middle of designing a revolutionary new drive whose "microstep" resolution automatically and transparently depends motor speed; the slower you go, the higher the resolution. The motor's motion is always monotonic (smooth) no matter how slowly you go even if it means 100,000 microsteps per step. One key is how current set is handled which is why it was left unfiltered.

Mariss

[NZCNC]

Is there aybody out there that can be skped into help sort out some issues with Mach 3,Granite devices servo drives and also Nema 43 Servo motors configuring this unit and getting it to work

Regards

Tim

[kanton]

Mariss,

The Deepgroove1 Gecko control box I got with my Taig mill came with 9 pin cables. The package was marked "P/N:5313118045F0 Made in China". Will these cables cause problems? If so what sort of problems? Should they only have 7 pins?

Thanks.