Messages

271

Modbus / Re: Discrete input/Brains

« on: November 15, 2011, 02:28:48 PM »

Greetings Jan:

I did find a screen capture copy of my modbus config, so can offer some further info if needed.
Always nice when things work.

272

Modbus / Re: Discrete input/Brains

« on: November 14, 2011, 11:12:00 PM »

Hello Jan:

It looks like some of the parameters in your brain input box may be incorrect.

I don't have access to my machine at this time. Look at the brain videos Art has done, and write down some notes.

The different boxes relating to the correct format in the brain are not really clear or intuitive.

The wording in the "Device" column is for information only, the numbers appearing in that label have no bearing.

Also, in the modbus test box, you have "Format" Hex , and a decimal value in the test box area. Just a guess here. However, you did get an input showing in the test box, so the modbus info is getting read into Mach. It is the brain that outputs the data into the program, so you are halfway there.

I am no expert in brains, but remember when setting up my modbus, it took a lot of fiddling and trying different numbers in the brain entry boxes before I got it working correctly.

It is not an issue with a bug in Mach3. It is just one of those tricky things that need working out One little wrong number in the wrong box, and nothing works!!

Sorry to not have more helpful information. I am away (in another state this next two weeks) from my machine, otherwise I could do a point by point set-up for a brain input from my own working system. I have (3) potentiometers, 13 pushbuttons, and a keypad all going through my modbus. Once it is working, one tends to forget about it. It just grinds away in the background, like a good liitle slave.

If you work with just one modbus input, and one brain until it works, then you can move to the next. You may have modbus interacting if the numbers are not right. They are using the same address. Just turn the enable off on one.

I do remember, the potentiometers all had increasing address's.

 
 

273

dspMC/IP Motion Controller / Re: Lagun knee mill with scales build using DSPMC

« on: November 13, 2011, 01:48:46 PM »

Welcome to the Mach3 forum j_boyss:

The dspmc controller is an excellent unit, and provides very solid control to servo motors.

I have one working for about a year and a half, and it has never given me a problem.

My servos are DC brush, .75 KW, with tach feedback, and 2000 line differential encoders. The motor tuning software has a visual graphic indication of the accel/decel curve, and motor response. The settings can be easily altered, and tested in real time, to achieve the best response for each motor/axis.

Why you wish to remove the tach feedback is unclear. A servo amp designed with a tach feedback will be inoperable (I believe) if the tach signal is deleted from the system. However, I do not claim to be an expert on all the servo/amplifier systems out there.

Linear scales or differential encoders are the digital "position" feedback to the system, not a part of the analog "velocity/acceleration" feedback, which the tach provides.

Mach3 does not "output" an E-Stop signal. Mach reads whether the physical E-Stop output signal is open/closed, hi/lo. The E-Stop signal is wired into the 7535 board, and correspondingly indicated in the "Inputs" section of the Ports and Pins configuration, i.e., pin#, active hi or lo. The E-Stop is a mechanical "Emergency stop."

You might read up on the "Reference" switches for setting the machine "Zero" coordinates. These are activated at the machine home limits. After these machine coordinates are set, then the "Work" zero position may be set, which can be saved in the Work table offsets.

The dspmc can utilize the encoder "Z" index channel for the machine coordinate reference position, which gives a very repeatable machine Zero if the encoder is driven direct to the leadscrew. I do not know if the glass scales incorporate a settable "Reference" point.

There are quite a few required settings in total, but just take them one at a time, and pretty soon it will all make sense. Get one axis working, and then proceed to the next.

274

dspMC/IP Motion Controller / Re: DSP and ME :)

« on: October 10, 2011, 01:10:40 PM »

Hello Aeroshade:

I agree 100%, i.e., the Vital Systems controller.  You just don’t see the myriad of crazy problems like so many other posts on this forum, regarding the dspmc.  It just plain WORKS, and very well at that! 

Regarding analog, take a look at the Cubloc PLC, and using modbus with Mach3. They have a large selection of units, from a single chip unit, to a large industrial modular system.

I am very pleased with the functionality of the Cubloc (www.cubloc.com), as well as the reasonable cost; my setup uses the CB280 module and a 32-M base board, plus a 20 key keypad.  Like all things electronic/programmable, it takes a little study, and fiddling. Very stable, and a nice addition to the retro-fit system.

As an aside, when I began looking at systems to retro-fit my machine, I did look into the Camsoft system, and Galil. The Camsoft seemed outrageous cost wise, so that never happened. Have read many negative forum posts regarding that system. I guess I just lucked out
 

275

dspMC/IP Motion Controller / Re: Error download encoder polartiy. Download aborted.

« on: October 08, 2011, 01:53:42 PM »

Need more info to be of any help.  


Some things to check:

An Ethernet "crossover" cable must be used between the computer and the dspmc controller.

The voltages at the encoders must be 5v, and a good ground to the correct terminal. A lower voltage will result in an error.

The "Ethernet connected" symbol shows in the system tray if there is communication to the dspmc, otherwise nothing will work.

The controller must have the power led's indicating correctly, per the manual, and the ground connection must be complete.

All encoder cables must be shielded, and grounded at one end.

After communication is established, then the correct "firmware" can be downloaded. There are more than one version, for different controllers.

The correct motor settings must be checked in "ports and pins", i.e., the correct port, and active state.

The correct encoder settings, i.e., counts per rev. must be set. A 2000 pulse per rev. encoder will have 40,000/rev. due to the quadrature.

The P.I.D. settings must be "in the ballpark" for initial motor tuning. Starting values are in the manual.

The E-Stop must be wired correctly, and operational.

The "reference" switches must be active, and to the correct pins in the (#7535) I/O board.

The I/O board (#7535) must have 24v, 5v, and ground.

The encoder A+, B+, Z+, 5v., and GND must be to the correct terminals at the encoder and at the board pinouts per the manual.
Channel 0 is X; channel 1 is Y; channel 2 is Z; channel 3 is A.

The motor signal (+/- 10v) must be connected correctly from I/O board to the motor amplifiers, and within a shielded cable.

After the motors are active, the settings must be arrived at by trial and error, using the dspmc plug-in and tuned for best response. The tuning functions are built into the plug-in.

The machine must be referenced, setting the machine coordinate "Zero", and then the work offsets can be established.

The "homing limits" must be set in the Mach configuration to reflect the machine work envelope, with respect to the machine coordinates. Direction of axis can also be changed/reversed if incorrect.

The encoder "Z" channel should be used for the homing function, as it is the most accurate. Info is in the manual.



Your mileage may vary! 

276

General Mach Discussion / Re: Oscillating bit in Z axis to equalize tool wear

« on: September 22, 2011, 12:32:35 PM »

Just thinking outloud;
Set-up a second computer running Mach3, driving just the Z axis. Wire up a multi-pin plug for the Z motor power/encoder/tack feedback to make a quick changeover. The Z progam could be incremental steps rising and falling, with shorter steps at beginning and end of each end of travel to mimic a s-curve for smoother performance.  Use a pot for feedoverride to regulate the speed of osscilation. Lots of cheap used computers out there, maybe even an old laptop would work. A second breakout board needed also.