Hello Guest it is January 22, 2020, 10:29:40 PM

Author Topic: EtherCat Pulgin...co-operative venture?  (Read 2519 times)

0 Members and 1 Guest are viewing this topic.

EtherCat Pulgin...co-operative venture?
« on: June 04, 2017, 10:40:06 PM »
Hi All,
a wee while ago I came across my first 'distributed motion control' set up, a proprietary communication protocol by Panasonic,
called 'Smart Servo', if memory serves, based on RS485 comms. Really clever.

As a consequence I did a little reading about distributed motion controls and various field bus strategies. All really fascinating.
The one that appealed on the basis of performance, flexibility and reasonably widespread adoption by industry is EtherCat.
If my understanding is correct if there were a plugin Mach4 could communicate to EtherCat capable servo drives via a humble
RJ45 cable. No hardware motion controller required!

While EtherCat is not so common amongst the cheapest servo and drives common on hobbyist gear most of the better servos and drives
are EtherCat capable. I am of the opinion that EtherCat may well come to dominate the distributed motion control field on the basis
of capability and  apparently flexible and favourable licence terms. It already appears to have more market penetration than any
of the other protocols.

If Mach4 is ever going to hit the industrial mainstream it will require some distributed motion control capability, at this stage EtherCat
being the most significant. Is there any interest amongst the community to write a plugin?

My own programming skills are not upto it but I could probably help in some sort of joint effort.

I need to find out a little more about plugins and Machs SDK.

Also need to find out a bit more about potential costs of licencing EtherCat. Don't want to spend time and effort on a plugin only to find
that you can't publish or sell the thing because of a licence snafu.

Craig
My wife left with my best friend...
     and I miss him!
Re: EtherCat Pulgin...co-operative venture?
« Reply #1 on: June 04, 2017, 11:38:24 PM »
Hi All,
just doing a bit more reading and looks like no cost need be involved licencing EtherCat.

If you wish to make an EtherCat capable slave device, a servo drive for instance, the EtherCat chip you buy to implement it is already
licence paid. To implement an EtherCat Master a licence is required but no cost!  Gotta luv no cost options! There are patented
technologies which may have to be paid for and whatever you build must comply with the specification. It might be that you need a certificate
of conformance which you would pay to have someone test and certify.

In the early days of field bus control of machine tools safety signals were not permitted to be communicated over the same data link as the
machine code, ie if the link failed or went cranky the Estop at the machine operators station would still work. In more recent years safety signals
are permitted to share the data link, they do however have to comply with some extra stuff...just what and how I don't know but could not rule
out having to pay for it.

In general the licencing stuff I have read is meant to encourage widespread use of EtherCat and is free or nearly so to do it. I think this policy is
likely to ensure that EtherCat is a dominant player in years to come.

Craig
My wife left with my best friend...
     and I miss him!
Re: EtherCat Pulgin...co-operative venture?
« Reply #2 on: June 07, 2017, 02:51:59 PM »
Hi All,
have been talking to Todd at NF and quite supportive of the idea but its not as simple as it sounds. EtherCat is a realtime system
and while there are technologies to make a realtime system out of Windows they are expensive.

In this case its not the cost of the licence of EtherCat that's the stumbling block but the proprietary technology required to implement it.
Todd also assures me that NF are keen to produce a plugin, a relatively simple task, if an economically sensible solution can be found
to the problem of realtime determinism.

As such I will shelve this idea, RTX, a realtime solution is beyond my means.

Craig
My wife left with my best friend...
     and I miss him!

Offline smurph

*
  • *
  •  1,187 1,187
  • "That there... that's an RV."
    • View Profile
Re: EtherCat Pulgin...co-operative venture?
« Reply #3 on: June 09, 2017, 01:27:42 AM »
Craig,

EtherCat can be done with the Galil.  Their 5000 series controllers are basically 4000 series controllers with EtherCat.  Also, there are several EtherCat controller companies that are working on Mach 4 plugins.  The Galil can go right now, but others are coming. 

BTW, Mach 4 was demonstrated running a Galil EtherCat controller running Yaskawa drives and motors at IMTS 2014.  :)

Steve
Re: EtherCat Pulgin...co-operative venture?
« Reply #4 on: June 09, 2017, 03:02:29 AM »
Hi smurph,
that's kool. I'm unfamiliar with Galil...am I correct in the idea that the Galil device actually runs proprietary motion code which is derived
from Gcode? The PC really only translates Gcode instructions which actually run on a realtime computing device...

The major reason for me to attempt to write this plugin is to learn and also contribute to the CNC community. It would appear however that in
absence of realtime determinism any 'contribution' would be sort of meaningless as it could not really be used with guaranteed results.

That Galil has already and others are working to the same end supports my contention that EtherCat may well become a dominant technology in
years to come.

My experience is that even a modest Atom base PC can quite happily run Mach and PP. This suggests to me that a PC, even a modest one, actually
has the 'horsepower' to do realtime computing job but Windows all but precludes realtime determinism. The question I'm asking myself is 'does a
Windows PC do near enough to realtime to be useful as an EtherCat Master?' You can see I haven't given up on the idea totally.

Craig



My wife left with my best friend...
     and I miss him!
Re: EtherCat Pulgin...co-operative venture?
« Reply #5 on: June 09, 2017, 10:07:28 AM »
Craig,

Vital Systems is also offering a EtherCat option on some of their motion controllers.

Andrew
Andrew
MachMotion

Offline smurph

*
  • *
  •  1,187 1,187
  • "That there... that's an RV."
    • View Profile
Re: EtherCat Pulgin...co-operative venture?
« Reply #6 on: June 09, 2017, 01:42:33 PM »
Craig,

Yeah, well...  time will tell.  The road is paved with protocols and ideas that never got out of the way of the traffic.  Some of these where indeed better ideas.  There were lots of "better" Modbus type protocols but yet Modbus remains.  Yaskawa had Mechatrolink that they are now abandoning.  Then there was the BetaMax vs. VHS battle.  Technically, BetaMax was better but yet VHS endured.  At this point, it does look like EtherCat is the "survivor" of these type of technologies.  But will the rest of the world move to it?  It is hard to say.  I know people that want to move to something like EtherCat and then I know others that say "No way!  Why?  We already move motors with a proven way.  Why is it needed?"  It does offer a better way of handling a distributed system.  But does every system need to be a distributed system?  

One of the things that EtherCat suffers from is that it gives a set of standards but fails to define how those very standards should be implemented.  For example (and it is a loose example), they may define that an EtherCat slave servo motor drive should have a velocity register.  But the EtherCat specification doesn't specify what the velocity is.  Is it counts per second squared?  Or just counts per second?  Or is it mm/s?  Or some dreamed up unit that the manufacturer came up with?  So all of the EtherCat hardware manufactures seem to do it all a bit differently.  They all seem to WANT to be on the EtherCat bandwagon, but they all want to do it their own way.  All of these manufacturers want you to buy EVERYTHING from them.  The controller, the servo drive, and the motor.  So the tendency to make something "special" is undeniable.  Until they just can't anymore.  

At this point, all of the EtherCat controllers either only support a few servo drive manufacturers or their systems need to be flexible enough to "define" what they are talking to.  Even though the basic technology (EtherCat) is the same!  This makes configuring an EtherCat system a lot more difficult than it really needs to be.  And in no way is it to the point where you can just hook anything EtherCat from manufacturer X to anything else EtherCat from manufacturer Y and have it all work.  So, in my opinion, the jury is still out because it just hasn't matured enough.  Others will disagree.  

Windows does NOT have what it takes to ever be a real-time system.  It will never be "real-time enough" to get the job done.  It has a hard time even calculating what a millisecond of time is.  Forget microsecond and nanoseconds.  All of the systems that use Windows for real-time are not really using Windows at all but are rather using another operating system hidden behind the scenes.  It is very much the case where two machines running, one real-time and the other Windows, on the same hardware.  There are multiple ways to do this, but the common idea is that Windows doesn't do the real-time processing, ever.  It just looks like it does.  

Steve
Re: EtherCat Pulgin...co-operative venture?
« Reply #7 on: June 09, 2017, 06:08:30 PM »
Hi smurph,
thanks for that clarification, its great to get a seasoned viewpoint from someone who has no vested interest.

My first exposure to a distributed motion control system was Panasonics RS485 system. They publish enuf hard facts about
the communication protocol that I could in theory make my own. This was on a large plasma table belonging to a customer thru
work. While it uses Panasonic servos it uses Logosol drives and these particular drives support RS485 comms exclusively.
I was trying to make the case to the owner to change to Mach4 but would need to fit genuine Panasonic drives at the same
time because they can accept step/dir type input. That unfortunately pushed the price up that he declined to change.

Thats when I became interested in distributed motion control in general. EtherCat appealed on the basis of flexibility and very
favourable licencing arrangements. What I did not twig to at the time was that EtherCat Master must be realtime effective.
I have looked at a couple of software solutions that use one or more CPU cores to run a genuine realtime system while reserving
one core to run Windows. Unfortunately they are expensive.

As to Why? I can see that EtherCat or some other bus system would be great in a factory with large numbers of servos, valves,
sensors etc. For a CNC machine even with co-ordinated slaved multiaxes more than 8-10 servos would be rare. Ehtercat would be
convenient to reduce wiring and so on but not necessary on a single machine. I imagine years ago when analogue servos were the
norm and pulse (step/dir) input was the new kid on the block lots of people asked Why? As you say time will tell whether distributed
control will take over and what protocols and standards evolve if it does.

My motivation is more personal, I would like a software project that would stretch my skills and make some useful contribution to
the community. Even were I to achieve a useful plugin would there be any uptake, even if free, if it required a multi-thousand dollar
investment in a realtime OS to run it?

If memory serves the PC running the plasma table was not running a realtime OS. The servodrives had a PVT table of 64, I think, entries
which allowed a buffer of moves. Depending on the granularity of the moves co-ordinated motion is possible. The difference between
that and realtime passed over my head at the time!

Craig
My wife left with my best friend...
     and I miss him!
Re: EtherCat Pulgin...co-operative venture?
« Reply #8 on: June 14, 2017, 02:38:34 AM »
As far as I understood the slave documentation of EtherCAT, you need to buy a Conformance Test Tool if you want to sell EtherCAT slave devices.

Complexity of EtherCAT master seems to be less than with slave system, but still there's special plugin hardware for PC. This could be the point to start with.
Re: EtherCat Pulgin...co-operative venture?
« Reply #9 on: June 14, 2017, 03:34:43 AM »
Hi,
I read it somewhat differently, if you wish to sell devices they are supposed to comply, you can either test it yourself or pay
someone else to certify it. Unless your customer demands compliance you do not have to do so.

EtherCat Master is a realtime system, PCs are unsuitable as is and the solutions to achieve realtime determinism are just to expensive
for me to get into as a hobby project.

Craig
My wife left with my best friend...
     and I miss him!