Multiple ClearPath with Z axis break not working

[David-CNC]

I have V3 AcornCNC that I'm using 3 ClearPath SDSK servos on. In the past when using a C10 breackout or USB smoothstepper for control, I only used the HLFB for the Z motor to control an SSD to release an electric break that I use in-line on the Z axis to prevent a crash should the z axis servo error out.

I have tried to wire the SSD in series and parallel with the HLFB as drawn in ACORN_rev3 MULTIPLE CLEARPATH SDSK SERIES MOTORS date: 180420 ver:4 Filename: S14885.DWG but when I have the SSD connected, upon a small amount of any axis travel I get a drive fault on the AcornCNC.

The SSD doesn't pull an appreciable amount of voltage and I've even tried using an opto-isolator (Phoenix Contact PLC-BSP-24DC/1/SEN) in parallel that only pulls 100ma but I get the same results.

Any ideas are greatly appreciated.

Regards,
David

[Richards]

I've had the same problem when using just two ClearPath servos daisy-chained as per Acorn schematic 14885. When looking at the Teknic ClearPath manual, page 140, the maximum current of the HLFB line is 9mA. When looking at the Acorn CNC Controller Specification Manual, page 7 shows the desired Input Current to be 20mA. Connecting ONE ClearPath motor to the DriveOK signal worked for me. Connecting TWO ClearPath motors in series did not work. Measuring the voltage and the current showed why. The "at rest" state of the two ClearPath motors was above the maximum 5.9V "on" voltage and below the minimum 19.1V "off" voltage listed on page 7.

My solution was to add a PLC that works with 5-10mA. Each motor is connected to its own Input signal. The PLC handles the logic so that if either motor faults, the PLC signals the Acorn properly. It works for me. The bonus is that the PLC controls the servo enable lines to each of the ClearPath servos and to the DMM servo. No brake is needed.

Gary suggested having the HLFB signals control one relay per motor. The Input line to the Acorn could then be daisy-chained through NC contacts. That works too.

[David-CNC]

Thank you for the detailed information and ideas. I should be able to get it going now.

David

[Richards]

Here are two drawings that show what is happening when three ClearPath motors are daisy-chained to one Acorn Input. The drawing labeled "Desired" shows what the ideal input would have: three NC switches. In that ideal input, when all of the switches are closed (ClearPath motors = No Fault), the Acorn Input would sense a closed circuit AND the current limiting resistor inside the Acorn's Input circuit would be the only thing that restricts the amount of current flowing through the Acorn's opto-coupler.

The drawing labeled "Actual" shows what is really happening (with ClearPath transistors being shown as switches for simplicity). Each ClearPath servo HLFB circuit includes its own current limiting resistor. When daisy chaining ClearPath motors, the internal current limiting resistors are also daisy chained. The result is that the current flowing through the Acorn's opto-coupler is too weak to create a valid signal.

It's easy to test this by connecting one end of a 10K resistor to an Acorn Input and the other end of the 10K resistor to ground. Push Alt-I on the Acorn to see the status of the input signals. With a 10K resistor attached to an input, the input shows Green. If the input is connected directly to GROUND (the same as if three NC switches were attached), the input shows Red. Experimenting with different values of resistors will show how much resistance the Acorn can tolerate and still give a valid signal. Unfortunately, with even two ClearPath servos daisy chained, the amount of resistance creates an unreliable signal.

[Gary Campbell]

Makes one think that my little relay video (from almost a year ago) was an accident??

Shows that even an old blind squirrel will find a nut every now and then. In the video I was using the relay for gender (NO/NC) and voltage (5/24) swap, but the same principal works as a "4 into 1" for servo "alarm" or "ready". Kind of a "hamburger helper" for inputs.

https://youtu.be/kN-EzRbRa_k

By the way, if you speak to the good folks at Teknic, multiple HLFB circuits that are daisy chained will work just fine. Real world shows this not to be the case when used in conjunction with an Acorn board.

[Richards]

Gary,
There is a warning in the ClearPath manual, page 140, Appendix D that specifies non-inductive loads for the HLFB signal. Unfortunately, a relay coil is an inductive load. Placing a fly-black diode across the coil is essential. Some relays have that diode build in. Others don't.

(My ClearPath motors cost $299 each. The HLFB circuitry is inside the motor where it would be difficult to repair. I thought about using relays, but the Teknic warning helped me to decide to use a PLC instead. A bare bones AutomationDirect Click PLC is $80. Additional I/O modules start at $30. That's the reasoning behind deciding to use a PLC.)

[Gary Campbell]

Mike...
These relays have opto-isolated inputs, so all the HLFB is driving is an opto-coupler.

I buy them 10 at a time, both 5 and 24vdc for <$40

[Richards]

Gary,
I must have missed the part about the relays having opto-couplers. Opto-couplers work well, especially since the HLFB signal is only rated for 9mA maximum.

[Richards]

Gary,
It took some time to test the ClearPath servos HLFB outputs in series going to a PLC input. Teknic is correct in stating that the HLFB signal can be daisy chained. What they probably left unsaid was the requirements for the device receiving that signal. For the CLICK PLC input module C0-16ND3 that I use, the minimum current is listed as 3.5mA. It works at 2.5mA; however, if the specs say 3.5mA, then I would stay at 3.5mA or higher to ensure stable operation. That 3.5mA minimum is about what three daisy chained ClearPath motors would produce (based on the current that I get through two daisy chained ClearPath motors).

I don't think that it's possible to configure the HLFB signal to fault at a precise setting. For example, if I wanted a fault signal to be asserted as soon as the motor lagged 10-degrees, I don't think that I can do that. What I have found is that ClearPath motor will fault at some point when the mismatch between internal encoder count and pulses sent is great enough.

[Gary Campbell]

Mike...
100% agree. That's why I qualified my statement with "in conjunction with the Acorn Board". In any case even if they did, and I am using 4 motors in this case, it may not be reliable. Over the years what these CNC boards require has helped make them a bit more resistant to static. Either way, in the future there will be ample IO for my type of systems, and there is a low $ solution for the present.

As far as the ClearPath setting goes, you should be able to set it at what you desire with the MSP 2.0 software. Most cut their own throat with this setting. I would be hesitant to use a number under 100 counts (1/8 rev) on the RLN models unless I had carefully logged the encoder counts during accel and decel to ensure that number was well outside those parameters.