- Posts: 25
- Thank you received: 0
do I need Opto isolators for General Mechatronics?
- toolmaker69
- Topic Author
- Offline
- Junior Member
-
I am putting together a micromachining center using Mitsubishi DC permanent magnet servo motors and meldar (mitsubishi) drives in a linuxcnc system, starting with four axes. To avoid loading the server up with lots of pictures, the link to info is at forum.linuxcnc.org/forum/38-general-linu...-hardware-components
I am about to order General Mechatronics GM6 card www.generalmechatronics.com/en/productdi...-motion-control-card and it looks like I would need optoisolators www.generalmechatronics.com/en/productdi...xis-optical-isolator in order to hook up the resolvers.
Is this correct? I have them all in my shopping cart, but would not order the isolators if I don't need them.
Thanks for any help.
I'm learning as I go, so please excuse my ignorance.
Please Log in to join the conversation.
toolmaker69 wrote: I am about to order General Mechatronics GM6 card www.generalmechatronics.com/en/productdi...-motion-control-card and it looks like I would need optoisolators www.generalmechatronics.com/en/productdi...xis-optical-isolator in order to hook up the resolvers.
If you have Resolvers, then I don't think that opto-isolators are appropriate.
Tamagawa are best known for Resolvers, but the devices on your servos seem to specify 1250 counts/turn so probably are not resolvers but encoders.
I would suggest that you ask GM for advice.
Please Log in to join the conversation.
I didn't see any resolvers, but there are clearly labeled encoders on the motors.toolmaker69 wrote: Hello,
I am putting together a micromachining center using Mitsubishi DC permanent magnet servo motors and meldar (mitsubishi) drives in a linuxcnc system, starting with four axes. To avoid loading the server up with lots of pictures, the link to info is at forum.linuxcnc.org/forum/38-general-linu...-hardware-components
I am about to order General Mechatronics GM6 card www.generalmechatronics.com/en/productdi...-motion-control-card and it looks like I would need optoisolators www.generalmechatronics.com/en/productdi...xis-optical-isolator in order to hook up the resolvers.
You generally would not use optoisolators with encoders.
Jon
Please Log in to join the conversation.
- toolmaker69
- Topic Author
- Offline
- Junior Member
-
- Posts: 25
- Thank you received: 0
I will contact them tonight after work.
I didn't realize there is a difference between encoders and resolvers. I just looked it up.
Please Log in to join the conversation.
- toolmaker69
- Topic Author
- Offline
- Junior Member
-
- Posts: 25
- Thank you received: 0
I checked GM's site and they mention differential and TTL encoders, though I don't understand what they are saying.
www.generalmechatronics.com/en/productdi...xis-optical-isolator
"This module can be used for isolating all the signals referred to the PC ground, and interfacing with TTL or differential output encoders."
I don't know what signals they refer to. Do the encoders need a PC ground? Or is the isolator for a different application when it comes to encoders?
Again, please understand that I'm new to this.
Please Log in to join the conversation.
I am not familiar enough with GM products to be able to advise.
Please Log in to join the conversation.
- toolmaker69
- Topic Author
- Offline
- Junior Member
-
- Posts: 25
- Thank you received: 0
"In general, it is not necessary if your servo driver electronics are already isolated. So one isolation is important between the control electronics and the power electronics."
I now have something else to research, if the answer is not obvious. I had assumed that I would be using PWM, since that is how I have programmed robots in the past, and I saw the letters PWM in another (Japanese) manual that seemed to use the same drive. GM doesn't support PWM, only voltage or current reference.
I probably just "misspoke" when I mentioned PWM to them. Do you know if CNC's used PWM around 1987? Can you tell from the pics of the drive what kind of input it might use? Some parts of my main machine installation manual mentions voltage input, but is not too technical.
EDIT:
I found this info in the Meldas II manual.
"The transistor amplifier amplifies the power by high frequency switching according to the pulse
width modulation method for the error voltage (command voltage) by comparing the position
detector output and the calculation result output from the logic card of the control unit in regard
to the movement command."
This seems to be if one is using the built-in closed loop of the amplifier.
Then I found this page discussing linuxcnc with mitsubishi motors and meldas drives: emc-users.narkive.com/13JUHu1G/running-m...motors-with-linuxcnc
It has a reverse engineered pinout, showing it to be analog velocity input emc-users.narkive.com/13JUHu1G/running-m...-linuxcnc:i.4.1.full
I'm not sure what any of this means yet, but I guess I'll hold off on ordering the GM board for now.
Please Log in to join the conversation.
What is it that made you decide on the GM drives rather than the alternatives?
Please Log in to join the conversation.
- toolmaker69
- Topic Author
- Offline
- Junior Member
-
- Posts: 25
- Thank you received: 0
Sorry Andy, what you said is over my head for now. I'll study it over the course of the day.
I chose the GM board because when I first started on this project, I ordered the Gigabyte GA-E350N motherboard due to its low jitter. It only has one slot, and it's PCI. The multiaxis Mesa cards seem to be for a different type of motherboard.
I edited my post above, adding some links. In the link www.cnczone.com/forums/mazak-mitsubishi-...e-help-please-2.html you are actually the last poster.
The link emc-users.narkive.com/13JUHu1G/running-m...motors-with-linuxcnc has a lot of information.
My head is swimming, but it does that sometimes.
Please Log in to join the conversation.
- toolmaker69
- Topic Author
- Offline
- Junior Member
-
- Posts: 25
- Thank you received: 0
The velocity loop goes back to the drive, bypassing linuxcnc. As in the Mazak link above, I would try to fit encoders between the motors and the tachometers, and feed the encoders back to linuxcnc.
The drive uses PWM internally to adjust output to the motor when comparing commanded signal to internal feedback signal. I hope that's correct. The Z axis already has a linear encoder, so it's only the X, Y and C axes that would need encoders.
So now I'm back looking at Mesa 5I20, with I don't know which daughter cards. Or perhaps still the GM card, using voltage output.
I need to stop and try to figure this out.
What would you recommend?
Please Log in to join the conversation.
