Kinextix 300 Not Always Starting on First Try

Archie

Member
Join Date
May 2002
Location
Orangeburg, SC
Posts
1,944
I have a machine with 14 Kinetix 300 drives communicating via Ethernet/IP. When the program sets the Start Index it does not always start. I had to add a retry timer that drops the signal and re-applies. It repeats this until motion starts. Has anyone experienced this before?

This machine originally had Ultra 100 servo drives and the original programmer also had a piece of reset code that would do the same thing when the drive stopped responding to start signals.
 
I don't have Tech Connect to view the link.

There are no drive faults.

Turning the Start signal on/off several times always makes it work and no fault reset is ever needed. It's like it just ignores the signal.
 
Now that I think about it, it got much worst with a faster processor. Maybe the program only turns off the signal for a few scans before re-applying. This would not be enough time to update the IO in the drive which is set to an RPI of 30ms. The new processor scans at 8ms
 
If your logic doesn't account for asynchronous I/O updates (especially if it removes the Start signal or asserts a Stop signal) then this is the classic way it manifests.

Can you post a screenshot of the index start command logic ?
 
The logic maps the IO to account for asynchronous updates. But it does not use a feedback to ensure the start signal was released as seen by the drive before re-applying. I looked at the manual and do not see a corresponding input for the StartMotion output.

StartLogic.png
 
Last edited:
I feel like I need to add logic to ensure the StartMotion remains off for at least the RPI. I do not consider adding a timer a very good solution because if the RPI is increased, this will no longer work. I really wish there was an input that would verify the StartMotion signal change made it to the drive before changing it again.

Has anyone else run into this? If so, how did you handle it?

StartLogicMod.png
 
We have used a few of these drives on smaller projects, looking through past logic, I have put a 500ms delay from the Safety Relay IP to the hardwired enable and the Ethernet enable signals and then another 500ms delay from the Kinetix Enabled IP before any motion move is commanded.

It's been 18 months since our last one, but your problem does ring some bells. A 35ms delay looks a bit keen IMO.
 
It's been 18 months since our last one, but your problem does ring some bells. A 35ms delay looks a bit keen IMO.
I chose 35 because the RPI is set to 30ms. That was based on my suspicion being the program can change the state of the outputs faster than the RPI. This machine has a time sensitive cycle so I need minimum delay as possible.

Today I will do some more testing to see if I can find a pattern.
 
Once it's enabled though, it's enabled so the timer isn't used after that, it just stays on.

Ours is only not enabled when a safety circuit is open, unless you are disabling the drive between moves it won't need to be as fast as 35ms
 
The last 300 I setup, I had more issues with starting the motion than the other 300's I'd used, probably because this project had more motions occurring more often than the others (RPI = 25ms, by the way). Here's what I learned through tech notes, tech support conversations & lots of testing:

The Kinetix 300 only looks for the .StartMotion 0 ->1 transition

.DriveEn MUST =1, .MotionComplete MUST =1 and .AbortIndex MUST = 0 for the drive to respond to this transition.

Through my own observations I needed to take it a step further and make sure .AbortIndex was off for one scan (One Shot Falling bit) before turning .StartMotion on. I believe this was due to the asynchronous update that Ken mentioned above.

Hope this helps.
 
Last edited:
I added in the timer to ensure the StartMotion signal was off for at least the RPI. Now everything is working fine.

The RPI was actually 20ms so I used a 25ms timer as a minimum off time for the StartMotion.

So my lesson learned here is that if the program scan is faster than the RPI, be sure the signal does not change faster than the RPI.
 

Similar Topics

Dear sir, I am using SIMATIC 300, CPU 315-2DP , (6ES7 315-2AF03-0AB0) VIPA 603-1CC21 A1.0 RAM 32KB, Firmware=V4.0.8 The problem Im using MPI...
Replies
1
Views
57
Hello. I have a few machines that use Kinetix 300 (each machine has two drives). Both drives on one of the machine keep losing IP address. They...
Replies
2
Views
75
Folks, I have a client with an old ABB Advant / MOD300 system (v14.4). Around y2k I installed the ABB Industrial IT MOD300 OPC Server 1.1/2...
Replies
0
Views
64
Hi, I received this SIMATIC S7-300 training kit for maintenance. When I power it up, the PLC doesn't go to RUN mode and the STOP mode led is...
Replies
7
Views
249
We have a AB Kinet 300 that went bad on us. I bought a new one and put the original memory chip in it. It’s all powered up and showing the ip...
Replies
1
Views
59
Back
Top Bottom