MooveInd

Delta DVP PLC to Inovance Servo Drive Pulse Count Error

17 posts in this topic

We currently have Delta DVP60ES2 and DVP32ES3 PLC's doing positioning control on Inovance IS 620PS2R8I 400w Servo Drives by Pulse/Direction.

Under both of these environments our pulse count read on the Inovance drives are reading higher than what the PLC is outputting.

We also have Mitsubishi & Inovance PLCs, which when connected to the same Inovance drives dont have any discrepancies in the pulse count.

We also have Delta ASDA servo drives, which when connected to the Delta DVP PLC's also don't have any discrepancies.

Given we have tested these combinations, it has led us to believe there is something unique about the pulse train output from the DVP PLC's that is causing the Inovance drives to read additional pulses.

Has anyone experinced a similar issue with Delta PLC pulse output? If so what was the solution? 

Share this post


Link to post
Share on other sites

I'm operating with Delta PLCs and Delta Servo drives only. I've never had experience (and troubles) with Inovance servo. All instructions bothering high speed pulse trains in DVP work perfect when combined with Delta servo. Why do you have to get your life harder with Inivance? Just use Delta on both ends and everything should be fine......

Share this post


Link to post
Share on other sites

Hi Ivan,

Thanks for your reply. Unfortunately we have 6 of the inovance drives, so its either I fork out to change all of those to delta drives, or we change the Delta PLC's to either Mitsubishi or Inovance (too much re-programming for my liking).

Agree with your comments, in the applications where we have Delta PLC to Delta Servo drive it is working perfectly. As is Inovance PLC to Inovance Drive 

We have confirmed that as the speed of the pulse output increases, the numer of additional pulses the inovance drive counts also increases. Given this is a somewhat linear relationship we concluded it can't be signal noise related. Perhaps something to do with the shape of the pulse output from the delta PLC.

 

Share this post


Link to post
Share on other sites

If you suspect problems with signal (shape or whatever else) then you have to check the signals characteristics of both PLCs by dual channel oscilloscope. The positioning instructions used by both brands are the same. 

Share this post


Link to post
Share on other sites

really good thought, I dont have an oscilloscope, but it might be the time to invest in one.

Do you have any other thoughts on how the Inovance drive could count aditional pulses from the delta output?

Share this post


Link to post
Share on other sites

Posted (edited)

No, I don't. My expirience with Inovance is very poor, but the only thing I'm sure about is that Inovance is uding the same positioning instructions as Delta. So if there's some incompatibilty between them there's nothing else I could offer you. I do have a scope although it's rarely used. By the way the shape of the pulses couldn't be any other than rectangular because the output transistors of the PLC could have only 2 states - open snd closed. Without any other circuitry is not possible to make another shape of the output signal (if there isn't any noise involved of cource!). Something else you can do is to check the settings of the servo. For example Delta DVP has special register for multiplication of the input pulses' frequency. Check if Inovance servo  has something as well. Check also if the problem occurs only at high frequency pulses transmission - you may check at very slow speed just to see if the problem is not because of high speed transmission phenomena.

Edited by IvanBratanov

Share this post


Link to post
Share on other sites

I share your thoughts on the inovance experience in general....poorly copied Delta products.. You are right on the transisters being binary state. We have ruled out noise as the issue for the following reasons (correct me if i'm wrong) we have tested with 2 different wiring setups, first was non shielded, then we went to shielded twisted pair, with no difference in the error. Secondly the error increases in a linear fashion as we increase the frequency of the pulse train. I wouldn't have thought noise would be so consistent and repeatable?

Thanks for your help.

Share this post


Link to post
Share on other sites

Posted (edited)

Quote

poorly copied Delta products..

Let me disagree with that. Inovance is poorly copied Mitsubishi, that's for sure. For example Inovance H2U and Mitsubishi FX2N, even the box is the same:

H5300b472cc614ba1b1e98452df8d4f77l.jpg

O1-CN01s-Y0-LWs2-CIbk-SVGM0-J-15888451.j

The same is valid about the servos and vfds.

 

Quote

I wouldn't have thought noise would be so consistent and repeatable?

It's only suppose from my side. I can't say anything more because I'm not there. Furthermore as I said I'm not familiar with Inovance so the problem could be in the Inivance servo somehow as well.  I have not ever seen such a servo, let alone worked with it.
 

Quote

 

We also have Mitsubishi & Inovance PLCs, which when connected to the same Inovance drives dont have any discrepancies in the pulse count.

 

 

 

 

It's time to look above of my post. They are the same. Don't be wondered that Inovance servo works fine with MEL and Inovance.
 

Quote

 

We also have Delta ASDA servo drives, which when connected to the Delta DVP PLC's also don't have any discrepancies.

 

 

Tested by me many times, confirmed.

Quote

it has led us to believe there is something unique about the pulse train output from the DVP PLC's that is causing the Inovance drives to read additional pulses.

Time to look both PLC's outputs thru a scope if you doubt that as a reason. You don't have to buy one but аsk sоmе friend to serve you with his one.

 

 

As you can see those are only lolgical conclusions by me. I couldn't suggest something else based on what has been said so far.

 

Edited by IvanBratanov

Share this post


Link to post
Share on other sites

Further speculation on my part....it's possible that the PLCs/drives that work fine have some sort of filtering on their outputs/inputs. In the real world. square waves really aren't; if you zoom in enough there will be spikes at the edges. Filtering can knock those down. A good scope will show that. You can look at the "good" ones and the "bad" ones to see. If they work fine as long as you match manufacturers but don't if you mix them, that would argue to me that there's something in the tuning of their input/output circuits that's optimized to work with their own products that makes them not work as well with the other. As has been suggested, a scope may be helpful here.

Share this post


Link to post
Share on other sites
Quote

it's possible that the PLCs/drives that work fine have some sort of filtering on their outputs/inputs.

No, it's not. Only a protection diode is available to save the output transistor in DVP:

images.png.ba6ab02a6c6bcac8b11d9d8e68d1c

Anyway I don't know how is it in Inovance.

Quote

In the real world. square waves really aren't;

In fact yes. The square wave is created via additive synthesis by increasing nimber of sine harmonics. So the nature of the square wave produced by different manufacturer's PLC outputs could have different characteristic. I'm not so familiar up to that level, but it could be possible.

Quote

As has been suggested, a scope may be helpful here.

I agree, that's why I have one. It's rarely used but when such moments occur it comes at first line.

Quote

If they work fine as long as you match manufacturers but don't if you mix them, that would argue to me that there's something in the tuning of their input/output circuits that's optimized to work with their own products that makes them not work as well with the other.

I have only been working with Delta for 15 years. In my practice there were few moments when other brands were necessary to be used in a single projets but they were  extremely rare. I've never had such situation where differemt brands were necessary to be used in same project and were not compatible. At least there is something that is unifying different brands' products and those are the standards, but anyway every manufacturer uses production tricks so the user would buy only his products.

 

 

There are other ways to control a servo amplifier - analog inputs, RS-485. If using of high speed pulse trains is not applicable here you may use other way to do it.

Share this post


Link to post
Share on other sites

Hi all, so the issue was ongoing and we ended up with a new Delta DVP ES3 PLC having the same issues with the Innovance drives. Changed the drive to a Delta ASDA A2 and alas, no better.

After digging deeper we noted the ES3 manual shows the need for a zenner diode + diode in parallel around the load. Something we were familiar with for inductive loads, but had no experience with on pulse train outputs. See attached.

Running positioning instructions at 150k with no issues now! Will implement on our ES2 PLC's and see if it resolves the problem there also.

 

Diode.pdf

Share this post


Link to post
Share on other sites

I've never used such extra diodes in my projects and they still run perfectly. My last servo-containing project was DVP-EH3+ASD-A2. No diodes of any kind were attached to the outputs and it still works like it should.

Not only ES3, but every Delta PLC manual recommends such additional diodes when it is about frequent on/off. Perhaps such problems appear when certain conditions are met at the same time, I'm not sure.

Share this post


Link to post
Share on other sites

The zener diode + diode combination on the output will absorb reflected pulses from the opposite end of the cable.  Figuring out the impedance of the cable and terminating the far end is a better solution to such problems.  Unless the transmitter's circuit is too wimpy to carry the load when properly terminated.

The following discussion for coax also applies to any kilohertz+ frequency in industrial systems:

https://sciencedemonstrations.fas.harvard.edu/presentations/pulse-reflections-coax-cable

Share this post


Link to post
Share on other sites
8 hours ago, IvanBratanov said:

I've never used such extra diodes in my projects and they still run perfectly. My last servo-containing project was DVP-EH3+ASD-A2. No diodes of any kind were attached to the outputs and it still works like it should.

Not only ES3, but every Delta PLC manual recommends such additional diodes when it is about frequent on/off. Perhaps such problems appear when certain conditions are met at the same time, I'm not sure.

How far away from the PLC was the servo drive? The post from Pturmel has me thinking the issue in my application has largely been caused by the length of cable from PLC to drive, which in my Delts DVP applications has all been quite long.

This was not the case where we have used either the Mitsubishi or Innovance PLC/servo combinations with no issues.

1 person likes this

Share this post


Link to post
Share on other sites

I can't remember the exact length of the line but it is no longer than 1m - PLC and servo driver are in the same cabinet, it just does not need more. Anyway I don't use such command method in my projects anymore - VFDs and servos are connected in the same Delta PLC link network and all the commands can be given by a simple twisted pair line. So high frequency side effects are in the past. I would recommend you to do the same.
 

"the issue in my application has largely been caused by the length of cable from PLC to drive, which in my Delts DVP applications has all been quite long."

 

I checked back the topic from the very beginning, but never noted you to say something about the line's length untill now. Seems that the problem here is similar to the one with the length of the line between a VFD and AC motor. 

Share this post


Link to post
Share on other sites

Hi Ivan, 

Completely agree with you comments on using comms protocols and twisted pair, will be using this method on future projects. We did try using preset positioning in the ASADA A2 drive, triggering each movement by RS485, the positioning was very repeatable, but we found it to be very laggy in both starting the move and the completion flag. The particular model of the drives we have ASD-A2-0421-L don't have Ethercat, CanOpen or DMCNET so I couldn't see any other option for variable positioning control?

The cable length was something I didnt mention in the beginning as we originally had the pulse/direction bundled in with the other input/outputs in a 20 or 30c cable (non shielded). Our first thought was to change to shielded cable for the pulse/com so we did that, which still didnt solve the problem. Based on that change I had asumed (seemingly wrongly) that we had eliminated the cause being the cable.

Hopefully the conversation in this thread gives someone experiencing a similar problem different areas to look!

Apreciate yours and everyone else's help.

Share this post


Link to post
Share on other sites

Posted (edited)

Like in the last post I checked back, this time whether you've mentioned what kind of machine do you use this servo in? You haven't do that. Giving more details about the problem in the early beginning of the conversation will clarify a lot of things and the people answering will spare themselves a lot of suggestions and speculations.

 

Of cource if the machine is a CNC router for example or something with frequent and precise postioning such protocols will not help a lot - if RS-485 and PLC link were good enough no one would ever try to develop such things as CANopen or EtherCAT.

My applications are far not so complicated as that, so it helps me in my cases. 
 

PLC link is a method of continuous sharing of registers of master and slaves in real time. Data to be read from the slaves is copied in real time to a group of special registers of the Master PLC and the data to be written in other group is copied to the slaves as well. Using of PLC link avoids using of single MODRD and MODWR instructions which require opening and completing data reading or writting thru the channel by flags.

Of cource as I mentioned above the complexity of the machine you're using the servos in should determine the interface and the protocol used to operate them. 

Edited by IvanBratanov

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now