Constant Force Application, Getting Started

[atoledob]

The good news (if I actually have the correct Mitsubishi manual) is that the analog input can be configured to 16bit.

Unfortunately, the setting for 16 bit resolution only applies to the analog speed command. I’m not sure what the resolution for the analog torque command is. The resolution for the analog torque limit is 10 bits, so maybe that also applies to the analog torque command.

There is also the 4-20mA analog option which is current-driven and less sensitive to noise … as long as good wiring practice is followed, noise shouldn't be an issue.

That’s good to know! The load cell amplifier has an option for 4 - 20 mA output, as do the AI boards I am considering buying. So, I wouldn’t need to add a resistor in parallel to convert the signal back to voltage.

The description for the load cell says “29 Awg 4 Conductor Spiral Shielded Silicone Cable,” which includes a floating shield. So, that part (+/- Excitation, +/- Signal) shouldn’t be susceptible to noise, correct? As for the current output signal from the load cell amplifier, is it enough to use a shielded (or double-shielded) stranded wire?

If I’m understanding this correctly, the load cell amplifier connects to an AI of the PLC. The analog torque command makes a connection between the servo drive and an AO of the PLC. The most resolution I can get out of an AI is 12 bit + sign or 15 bit + sign. For the AO, the most resolution I can get is 12 bit or 14 bit. Is choosing the higher-bit option always better?

Further, how would I go about scaling down the +/- 10 V from the AO to the required +/- 8 V for the analog torque command? Is this something I do within the PLC programming?

Thanks again for taking the time to help me.

 

[Tinine]

Unfortunately, the setting for 16 bit resolution only applies to the analog speed command. I’m not sure what the resolution for the analog torque command is. The resolution for the analog torque limit is 10 bits, so maybe that also applies to the analog torque command.



That’s good to know! The load cell amplifier has an option for 4 - 20 mA output, as do the AI boards I am considering buying. So, I wouldn’t need to add a resistor in parallel to convert the signal back to voltage.

The description for the load cell says “29 Awg 4 Conductor Spiral Shielded Silicone Cable,” which includes a floating shield. So, that part (+/- Excitation, +/- Signal) shouldn’t be susceptible to noise, correct? As for the current output signal from the load cell amplifier, is it enough to use a shielded (or double-shielded) stranded wire?

If I’m understanding this correctly, the load cell amplifier connects to an AI of the PLC. The analog torque command makes a connection between the servo drive and an AO of the PLC. The most resolution I can get out of an AI is 12 bit + sign or 15 bit + sign. For the AO, the most resolution I can get is 12 bit or 14 bit. Is choosing the higher-bit option always better?

Further, how would I go about scaling down the +/- 10 V from the AO to the required +/- 8 V for the analog torque command? Is this something I do within the PLC programming?

Thanks again for taking the time to help me.

I think that there is an approach here that isn't being considered. We are looking at this as simply controlling the force of an actuator but this is no "dumb" actuator, it is smart and can be positioned with extreme precision.
When we increase the clamping force of a vise, it means that we are bringing the jaws closer together. Here, we have 22bit resolution and IIRC the ballscrew has a pitch of 5mm.
I would like to see what can be achieved simply with these elements alone....for force control. Now, what needs to happen dynamically as part of the complete process, I haven't been able to definitively absorb/ascertain due to long hours. Come Friday afternoon and I can review with a fine toothed comb.


Analog signal wiring, still one of life's mysteries for me.
Definitely twisted pair but I've had cases where unshielded was less noisy than shielded. Maybe the shield was an antenna(?). Dunno but I always start with shielded twisted-pair.
I wouldn't get too hung-up on the noise issue until we come across it, if ever. At least one of Pete's RMC products has 18bits of +/- 10v AO so the issue must be manageable.


Do I have the correct manual? The link that you provided, didn't like that I was outside the US and so I grabbed what appeared to be the correct document from elsewhere. Mine says nothing about 8v torque setting. However, if I'm not totally crazy, this could be a moot point because the position-loop P-gain will be all we need. We can make it stiff or spongy.


[prepares to be shot down in flames]


Craig

 

[atoledob]

Now, what needs to happen dynamically as part of the complete process, I haven't been able to definitively absorb/ascertain due to long hours. Come Friday afternoon and I can review with a fine toothed comb.

I appreciate you taking the time to think about this problem with me. Hopefully, the solution is as easy as just messing with the position of the actuator. I still have the option to use the Torque/Position mode instead of just Torque control.

Do I have the correct manual? ... Mine says nothing about 8v torque setting.

Yes, that is the correct manual. On page 101 of the pdf, you'll see section 3.2.3 Torque control mode, which shows how to wire the I/O. You will see where it says TC, the analog torque command, is between +/- 8 V. You can find more details on the Torque control mode on page 130 of the pdf.

 

[Tinine]

The positioning resolution is quite insane: 1/838860.6 mm


I have used this approach in the past (albeit much lower resolution) but I can't, for the life of me, remember where/when.


Ah, my manual is the "Positioning" edition. It appears that each mode has its own document. My 3.2.3 is "Diagnostics".






Craig

 

[Peter Nachtwey]

When you say RS485 is best, are you referring to using serial communication? The baud rate specs for an RS485 communications module/board are: 300 baud, 600 baud, 1.2 kbits, 2.4 kbits, 4.8 kbits, 9.6 kbits (default), 19.2 kbits, 38.4 kbits, 57.6 kbits, 76.8 kbits, 115.2 kbits.

I did some rough calculations, so I’m not sure how correct they are, but force changes about 0.16 [N/s] and about 0.37 [N/mm].

I wrote this wrong. I should have said use analog.
Modbus is TOO SLOW and timing dependent.
Serial isn't fast enough unless you are using KHz serial like what is available on SSI.

 

[atoledob]

I wrote this wrong. I should have said use analog.
Modbus is TOO SLOW and timing dependent.
Serial isn't fast enough unless you are using KHz serial like what is available on SSI.

Good to know. I will stick to analog signals, thanks!

 

[atoledob]

Ah, my manual is the "Positioning" edition. It appears that each mode has its own document. My 3.2.3 is "Diagnostics".

So we’re both on the same page, this is the manual I am using.

Also, I was mistaken before and my load cell amplifier only outputs 0-5 V and RS-485. So, unfortunately, I can’t go with the less noisy 4-20 mA signal transmission. If I keep the cables short and use a shielded twisted pair, maybe I will still be okay.

I think I will go ahead and order an AI signal module that offers 12 bit + sign resolution, and an AO signal board that offers 14 bit resolution (unless someone thinks I should spend the extra $188 to upgrade the AI resolution to 15 bit + sign resolution). Hopefully, this is enough resolution for my load cell to send values to my PLC, and for the PLC to send back torque values to the servo drive.

After some googling, it looks like I can convert the +/- 10 V from the AO to the +/- 8 V I need for the analog torque signal by using a voltage divider. Does this sound right?