Machsupport Forum
Mach Discussion => FAQs => Topic started by: atvbarry on March 20, 2012, 03:56:43 PM
-
Will the Delta ASDA-B2 750w ac servo drive work with Mach step/direction output from the breakout board? 2nd if Delta ac servo will interface, which pins on the ac drive do you connect the step/direction from the breakout to? I was wanting to use the internal power supply from the Delta drive but if I couldn't no big deal.
I have hooked up all kinds of stepper and dc servo configurations but just starting to get my feet wet on the ac servo drives, any help would be appreciated.
Thanks
-
A link to the drives and their manual would help ;)
Hood
-
I guess it would be a BIG help post more information!
I have a C11 breakout board from cnc4pc.com, I was told that board would work, I wasn't for sure how to connect the step/direction terminals to the pin#'s on the ASDA-B2 Delta AC drive to interface.
Like I say the drive is a Delta ASDA-B2, (750W).
Here is the link to Delta manual http://www.delta.com.tw/product/em/download/download_main.asp?act=3&pid=2&cid=1&tpid=1
Thanks
Barry
-
Ideally you want to make up a line driver so that the step/dir signals are differential. That means for every positive step pulse there is a negative pulse signal and same for Dir.
You can buy line drivers ready made or you can easily make up your own with a line driver chip and some breadboard.
Hood
-
I wondered about the line driver thing, which line driver do you recommend and where do you buy it.
Thanks
Barry
-
I made up my own circuit boards, initial ones were just made with routed out copper clad then I progressed to getting PCBs made and next will be a full breakout board with 24v I/O on two ports worth of pins and differential (or single ended if user chooses) on one port of I/O. Sadly due to money being tight I have had to postpone the BOB as getting boards fabricated isnt cheap in small volumes.
The actual line driver I used on the first two boards was SN75172
I know US Digital sell line driver units, also CNCDrives in Hungary do them for their drives.
Pics below are of my three stages of progression ;D
Hood
-
Looks like you moved right along in development, your right on the low numbers manufacturing, the cost can get scary!
You get this new BOB out and I will be one of your first customers.
I have worked in prototyping new products (mostly mechanical and fluid power with some electronics) for years and one of the most important aspects is manufacturing cost.
Thanks for your help, I will research this and see where it takes me, I want promise you that I will not be SCREAMING HELP later...
-
I dont really have any plans to sell them, was thinking about it but dont fancy soldering up too many of these tiny surface mount components :D
The Bob will actualy be split into 3 as I will be using it on various machines and will have different requirements for each and thus I can mix and match. One big requirement for me is the majority of the I/O be 24v as its much more noise immune and likewise the 5v I/O boards should be differential for such things as Step/Dir out and MPG and encoder inputs. I had originally designed the board with the Ethernet SmoothStepper in mind hence the three ports worth of I/O but I plan to use it with pokeys as well, especially if the encoder feedback pans out on it.
Hood
-
I like the idea of mixing and matching, being more noise immune would be a plus for sure, I have ran to some noise problems on a cnc plasma tube coper running on Mach that I designed and was using one of the SS usb connection on a laptop, the SS can't be close to the plasma at ALL that thing would work perfect one time and the next it would go crazy, might do anything, then I changed and ran Mach on a tower with the 25 pin parallel port and I actually believe you could sit the pc tower on top of the plasma and not have any problems, oh yeah forgot to mention that I had the SS hooked to the tower usb and it acted the same as with the laptop, I have not did anything with ethernet SS yet but I like the performance of the SS, the only down fall i saw was the loss of the backlash comp.
Barry
-
The Ethernet SS seems to be very robust regards noise, I have heard of it being tested sitting on top of a plasma torch and it not falling over.
I never really suffered any noise issues with the USB version anyway, did have a slight issue on one machine where if I hit the E-Stop (dropped out all contactors) it would occasionally crash the SS but that was solved by removing the cap and resistor on the SS.
I have the ESS on the big lathe and it is nice, still one niggle for me with the SS and that is Step/Dir spindle is not 100%, it seems very occasionally it will skip a beat, its for a very short duration and doesnt trip the drive but you can hear it. Also if varying the spindle speed via the Override Pot it clunks a bit. Greg knows about this and has done for at least 2 years but so far nothing has been done, probably because there are very few that use servo spindles via step/dir so it wont be high priority.
Regards backlash comp, that has been dealt with, probably getting on for a year or so, and it works well by all accounts..
Hood
-
Hi Guys,
I came across this thread as I'm busy trying to set up the exact same drives for a client using Mach3 as they are unhappy with their PLC control solution. I've been looking at the datasheets and it looks like there is an option to use single-ended step and direction signals for a breakout board. Please see the attached application note. My only question is...which signal is step and which is direction? PULSE or SIGN?? Lets try get this interface solved as I'm sure it would benefit quite a few other users.
Thanks!
-
And I note that we would need to add a PNP transistor for each signal :) But that is straight forward and cheap.
Please correct me if my assumptions are wrong. I'm referring to the first diagram in the attached file.
Thanks again!
-
STEP = PULSE
DIRECTION = SIGN
-
Appreciated,
But I have seen a reference in some of their documentation that PULSE is direction. I know, makes no sense to me either, but it does say it. So I'm hoping that you've set up a drive with STEP = PULSE and DIR = SIGN? That would make my life much easier :)
Thanks again.
-
Yup ... Automation Direct, SureServo.
(seen a couple others also)
Russ
-
Ha, superb.
Thanks very much. I'll post the connection diagrams here for reference once the setup is done and tested.
CIAO
-
Just an update on all of the above. Step/Direction issue resolved. Trying today to connect the step and direction from a 5V BoB with NO Line Driver. So just single-ended 0-5V. I appreciate that this is not too noise immune and it may not work at all. There are two 51ohm resistors either side of the optocoupler input diode inside the Delta ASDA-B2 drive. If you do as they say with the line driver, that is a 10V differential across these resistors in series. Gives about 0.04901A, or 49mA of current by their (Delta's) design to switch the optocoupler. HOWEVER... we do not have access to the tech. specs. for the optocoupler which they use and it is my suspicion that using single-sided input (0-5V) may just work. It is obviously not ideal, but it will still give 24.5mA across the optocoupler diode which may still be enough to switch it on. The only place I see a problem with this is that at high speeds there won't be good step clarity and it may result in lost steps to the drives. (To the drives, not to the motors, which are obviously closed-loop). So, in summary, I will test this now and my guess is that provided we stay below 100kHz pulse trains (200kHz is the spec.) that it should be ok. Will post updates shortly. (I note again that you will need EXCELLENT noise shielding to do this and not lose steps.)
-
Right, so as promised, I did the setup and testing. Delta ASDA-B2 400W drive works perfectly well with single-ended inputs. NO LINE DRIVER. And it's reliable too. No lost steps between Mach 3 and the drive control loop. I was pleasantly surprised as to how well it worked actually. I applied a constant 5V supply to pins 39 and 43 of the CN1 connector of the drive and connected STEP and DIRECTION signals from my BoB (Cheap Chinese thingy... the cheapest ones on E-Bay) to pins 37(DIR) and 41(STEP) of the CN1 connector on the servo drive. In Mach, I simply assigned the step and direction pins as per the BoB and made sure that the pulse width was 5us and that the gap between pulse and direction change was 5us under 'motor tuning'. The drive is set up as 17bit from factory (132000 and something steps per rev) so I had to go into the drive configuration software and electrically gear this down a bit so that the Mach kernel speeds could actually get the motor spinning at 3000rpm. All done though and working great. Hope this helps some others struggling with the config.