VSD-A Servo drive documentation

[Xerxes]

Purpose of this thread is to discuss about drive documentation and announce documentation changes. The more discussion, the better.

Similar Threads:

• Problem- old Mazak M2 Control on VQC Machining TRA-41 Servo Drive with HD101-12 DC axis drive
• Need Help!- TECO 20C Drive Breakout Board pin layout documentation missing...
• Need Help!- need documentation Sanyo AC Servo amplifier
• Newbie- CNC Lathe - Servo direct drive or pulley drive?
• online servo documentation

[Xerxes]

I just added some information about designing a power supply for drives in VSD-A documentation. Check it out at:
http://www.granitedevices.fi/index.php?id=8

It is best not to follow that guide too tightly because it simplifies things very much. Previous experience in PSU building may help preventing magic smoke from escaping the circuits :-)

[freezy]

Hi,

I added GDTool manual into Granite Devices webpage:
http://www.granitedevices.fi/pub_fil.../GDtool031.pdf
Next version of GDTool is under work so GDTool document will be updated after new software version is released.

[Xerxes]

VSD-A data sheet got just updated (version 0.40). I added chapters about known limitations and supported motor types.

[daedalus]

having read the datasheet for the drive and your tuning software, i noticed that you mention torque control as an option in the drive datasheet, but there doesnt seem to be a choice for it in the tuning software.

Also how does torque control via step / dir work? do you keep an internal desired torque, then step it up / down by a fixed amount? is that fixed amount user specified? For torque it would be much more useful to have pwm / dir control, as i dont know any motion control hardware that outputs step/dir, its usually just pwm or analog.

Also do you know if anyones working on spi interfacing your drives to EMC / mach?

[Xerxes]

You're right about torque control. It is supported but not shown in GDtool because it's only a temporary tuning software. Timo (freezy in CNCzone) is working on the next version of GDtool which should unleash all the features of the drive. However, I can write instructions about how to set torque mode on manually if needed. User desired step/dir command scaling can be done in all modes.

Currently step/dir or SPI is used for all modes. Hardware has built-in capability for PWM mode but it's not yet implemented in firmware (I'm working on it).

I haven't heard anything about direct SPI support. I have taken a look in EMC internals and I might try to write some a driver for it some day. But it would be faster if some EMC expert would be available for assistance

[daedalus]

Thanks for the info. One thing that might be helpful is if you had a list of planned updates / features (subject to change of course) for people to look at somewhere.

The step/dir scaling is pretty important, and probably should be on the datasheet, as people with P5's are going to need it if they have 10,000 step / rev (x4 decoding).

[buckie555]

Currently step/dir or SPI is used for all modes. Hardware has built-in capability for PWM mode but it's not yet implemented in firmware (I'm working on it).

I also plan on using EMC and the MESA 5i20 card so PWM mode for torque control would be a very welcome addition for me as well. Any ideas on when this is likely to be available?

Will existing customers be able to upgrade their firmware? What cabling/host application requirements are there?

[Xerxes]

Well, it seems that there is more need for PWM than I first thought so I'll try to implement it sooner. I'll try to experiment with it this week. If everything goes smoothly, it could be available in couple of weeks.

Firmware is upgradeable and requires nothing more than one jumper block Tuning software comes with flasher software which uploads firmware thru SPI tuning cable.

Planned features at the moment:
-Stepper support (open & closed loop)
-PWM input
-Controllled hard stop homing option (no home switch needed)
-32 bit absolute position commands via SPI in addition to 16 bit (good for point-to-point applications with long moves)

[Xerxes]

Documentation got updated again. It now includes full listing of features to reduce need of guessing

http://granitedevices.fi/index.php?id=15

[Eson]

Will there be stepper support for the drive? I read somewhere about a firmware upgrade. Is this true?

Maybe wrong thread...sorry

[Xerxes]

Eson,

Yes, stepper support will be in upcoming firmware version 110. VSD-A won't reach the optimal 2-phase stepper performance because it has 3-phase output but it seems to perform quite well anyway.

[JerryFlyGuy]

Jumping into an older thread here, but I'm working through implimenting my first drive as a test before doing the rest of my machine.

Some of the connections at both the command header and the encoder header aren't entirely clear from the documentation I have off the website.

Can someone tell me what the VCC on pin5 is [on the command header]?

I've never done a servo system before, and I understand a little about implimenting it. However, some general info on how to best tie the Fault pin's into a system as well as the Disable pin's would be helpful, as well as voltage level's and mA level's required on various pin's might save some grief later [rather than let that magic smoke out].

On the Encoder header, is it nesc to connect anything to the home-/+ pin's or can they be left un-connected?

Stumbling along...kinda lost since I fell through the rabbit hole..

Jerry

[JerryFlyGuy]

Ahhh.. well I'm still learning that addage.. RTFM, I've since found what I think is the list of various voltage and amperages needed for the CMD pins however I'm still at a loss as to what the VCC pin is.

Also, is the normal proceedure to put the fault output into a BOB-pin which would Estop the control or just stop [feed hold??] it? The clear fault pin, should it be just a momentary switch to clear the fault [located on the control panel for example]??

Where the manual states that we should connect the shielding to the ground pin, is it meaning ANY ground pin or the respective ground pin from which cable the shield is coming from [ie CMD cable to a CMD ground] or, is it wanting them all to go to the HV grnd or a logic grnd, or does it really matter?

Thanks guy's, I'll get there eventually.. I just don't follow [or enjoy] electronics too much... [I know.. why did I then tackle this?... because I plan to make stuff w/ the thing.. not wire it every 6 months.. ..well.. ok.. this will only be the second time I've wired it in 6 months.... I never get anything right on the first try..]

Jerry

[H.O]

Hi Jerry,
Tero will have to have to cofirm this but I guess that the Vcc in the CMD header is the drives internal +5V or +3.3V. Some PLC systems etc have isolated outputs and this voltage can then be used to drive the VSD-A inputs. If you're running from a PC and want to have complete isolation between the drive and the PC then you don't need to use it.

The voltage for a logic 1 on the optoisolated inputs is should be between 2 and 5V and whatever is driving that input (LPT-port?) needs to be able to source 10mA (max, worst case).

I would E-Stop the system on a servo drive fault. When that happens the motor is no longer under the drives control so you have basicly lost position and needs to re-home the machine.

Regarding the shielding, my guess is that you connect the encoder cable shileld to the GND available at the encoder connector. The motorcable shield at the HV-DC supply GND etc. Again, Tero will have to confirm.

HTH
/H.O

[JerryFlyGuy]

Thanks H.O.

So if I'm using a break out board.. do I use the VCC?? If I don't apply power there... where does it get it? Something needs to be connected there doesn't it...??

I think I'll be estoping the system as well, I'll probably run bunch of relays in series w/ the Estop button and then either or will cascade an estop to the other drives and the control. The unit should stay powered up though, shouldn't it? This would then allow it to maintain position [the HV power would be killed on my system in this conditions so the drive can't actually 'drive' anything] by keeping track of encoder counts??

I was planning on putting the shielding for each cable to a ground pin on it's respective connector [encoder to encoder etc as you'd stated] but just wanted to make sure I wouldn't fry anything..

Hopefully Tero is 'at the office' today and confirms this so I can work on it over the weekend

Jerry

[H.O]

Hi,
Well to be SURE you have to wait for Tero ;-)

No, I don't think you should use Vcc. The way I see it is that pin is not power TO the drive but power FROM the drive that can be used to power the drives isolated outputs. (Look at figure 7 on page 9 in VSD-A rev 2 manual). Or to power your BOB's S/D outputs if they were isolated. But since most drives already have optos on the S/D lines most BOB's doesn't. (I've never used any break-out-board but that's what I believe)

Hope that made sense....


E-Stop: Yes, cut the HV-supply, disable remaining drives, cut spindle power and alert the control but leave the logic supply to the drive ON.

I don't know IF the VSD-A keeps track of actual position when disabled but even if it does you can't (in an easy way) get that information out of it and into whatever CNC control program you are using. To do that the particular CNC control software needs to communicate with the drive(s) via the SPI interface. Doable, perhaps, but not available now as far as I know.

Again, I hope that made at least some sense, hopefully Tero will chime in too.

/H.O

[Xerxes]

Hi guys. Good you are having discussion here, so it is also helpful for others :-)

The CMD_VCC is supply voltage for optoisolator fault output. You can see how it is connected to optoisolator in schematics. However, you're right that the documentation should be improved about this.

So in practice, if you are using PC, connect PC ground to OUT- pin and +5V (from USB for example) to CMD_VCC pin. You can also leave these unconnected if you don't need fault signal from drive.

Since fault signal (OUT+) is quite weak (0.5mA), you need to be pay attention how to feed signal to PC. You can see VSD-A breakout board schematics for suitable circuit which amplifies the signal. Actually, why not use this VSD-A breakout design in the first place?

-----

About home switch input: you can leave it unconnected if you aren't planning to activate home switch search in drive. And if you use switches, then use separate switch for drives (don't share switch signal with PC since it would eliminate optoisolation benefits).

-----

About E-stopping: the best way is to cut HV supply voltage and empty capacitor bank by connecting bleeding resistors (by relay or mosfet) to disconnected supply. This will brake motors quickly since HV drops to zero rapidly. Without resistors, motors free-wheel until they stop by friction or collision (assuming that drives have become inactive via disable input or by fault condition, otherwise drive attempts to control motor until HV voltage drops below ~20Vdc).

Drive will keep track of encoder position as long as logic supply is on. So you can recover position after E-stop without homing.

[H.O]

See, I was totaly wrong about the Vcc pin even though I did look at that figure in the manual, thanks for stepping in and clearing that up Tero!

Are you saying that a low(ish) ohm resistor across the powersupply not only drains the capacitors but also breaks the motors? I was thinking about switching in (delayed) resistors across each motorwinding as well as the bleeder. Is that overkill or even risk of destroying the drive?

It's good to know that the drive keeps track of position even when disabled but if used in S/D mode with common CNC software (Mach3, EMC etc) you'd need a way to retrieve that actual position FROM the drive and update the position registers in the CNC software. A plugin for Mach3 to that would be NICE! Can you daisychain a set of drives and interrogate them one-by-one over the SPI interface?

Thanks!
/Henrik.

[Xerxes]

Yes, draining power supply to zero brakes motors almost the same way than putting resistor across motors. Relay switching in drive output might cause some sparking that may damage power stage.

CNC software doesn't need to know position as long as drive keeps motors at position where software expects them to be. VSD has soft recovery function, so after re-enabling drives, motors will softly return to the position where they should be.

I'm planning to document SPI some day throughly so software writers can support it if they wish to. Daisy chaining is not possible at the moment but parallel port could control at least 4 axes thru SPI.