Upgrading to a motion control board

Hello everyone,

I'd like to upgrade one of our machines from using the parallel breakout board to a Mesa driver board. Currently the breakout board is optically isolated and has a separate power supply that supplies the voltage for the output. So, in this case, since the servo drives are 3 phase Panasonic Minas unit and the control circuit voltage for them is 24v with the open of open collector or line driver (differential), I just connect a 24v source to the output side of the breadboard, the positive outputs on the breakout board go the corresponding pin on the driver and the opposite pin on the driver just goes to ground on the breakout board since I am not using a line driver or differential voltage.

My question is: How would I interface the drivers with say a Mesa 6I25? I'm guessing I would use a 6I25 with a 7i76 card, but I am a little confused as to whether this setup can drive the pulse and dir lines at 25 volts. The inputs I did see can handle up to 36 volts so thats no problem.

Any input would be greatly appreciated.

Thanks!
Frank

The 7I76 cannot drive 24V step.dir inputs but
can drive differential (line driver type) inputs directly

Ok, that makes sense.

So then I guess I could make a new breakout card the same as I made this one to act as a level shifter like this one is currently working. I would eliminate the inputs from this break-out-turn level shifter card and use it only for the level shifting and/or maybe add some maxim line driver IC for a full solution to go from the 7I76 card 5V TTL output to 24v lvds step dir output. The maxim lvds line drivers operate at approx. 1350MHz so I would guess the increase in signal latency shouldn't be that big of a deal. The encoder outputs (differential) from the drivers and from the glass scales would then go directly to the 7I76. Does this sound ok?

BTW,
Is that an Anglo-Nubian on your profile picture?
Frank

No, it is an American Blackbelly sheep (that particular one is Blackie)

The 7I76 only has one encoder input (as it is designed for open loop step/dir systems)
if you need a combination of step/dir and encoder inputs I would use the 7I85

Blackie is beautiful! I asked because I have been looking at goats and sheep...maybe one day I'll get one or two for my little hobby farm.

Ok, I think I follow you.

Sorry to be a pain, but I am a little confused about this though: Do the line driver lvds also use 24v on the driver or is it voltage agnostic since as I understand it in lvds the logic level is determined by the voltage difference and not signal line voltage level as in TTL? So... if I wanted to use the line driver inputs on the driver then I could use the 6i25 with two 7i85S's connect the encoder differential outputs from the drivers and glass scales directly to the 7i85S (the driver encoder outputs I believe are also 24v lvds while the glass scales outputs are 5V TTL) and then the 10 step and dir lvds outputs on the 7i85S's would go directly to the drivers without the need to use level shifting card in between?

Frank

The line drivers are 5V nominal (and differential)
I dont think I have ever seen a 24V differential system (24V single ended is fairly common)

Ahhh, ok I understand now. The thing is the documentation did state the 24v was for single end but didn't say anything about the differential voltage, so me being pretty ignoramous I assumed that the voltage stayed the same, it just added a line compliment. That really does clear it up for me.

Thanks!

Hey Andy,
On the Minas drivers, and most of the other AC drivers I've come across, you can usually specify one of three types of control schema: position, torque and velocity. I'm not really quite sure what torque is useful for, but velocity control is what you are referring to where it wants +-10v DC. This is akin to using it as a spindle drive. This is used allot in conveyor systems and dosing/mixing were the accuracy in relative speed of two components is more important that the point-in-time position of the components. In position control you use pulse and direction precisely the same way as you control a stepper motor. In the newer controls like the Minas A5 you can even connect your glass scales to the driver and then from there to your controller the same way that the servo encoder goes to the driver and then to the controller. The driver performs the same full-closed loop error checking as the controller does only it does so to continually update the amplification-power table in almost real time. This way the amplifier continually updates the amount of power required to accelerate or decelerate according to the changes in input with the least amount of error and lag. It used to be that you would run a power calibration at the time of installation say on your VMC and it would create the amplification parameters from that. However, once you load the table the loads become greater than what the amplification parameters were set for so you then have to live with a larger permissible lag before the servo spits a follow lag error. Now, if set up correctly, the servo is continuously updating its amplification curves based on the real-time loading/performance information provided by the encoder and the scales, so for instance if you load your tabel with 6000 pounds or with 100 pounds, the performance should be the same. Couple this with a good look-ahead schema and external position verification both to the controller and you can really dial things in neatly. This is precisely what I plan to do on my next two retrofit projects, the little lathe and the big VMC.
Frank