Kinetix 5500 Motor Feedback Data Loss

Hello all, new to the community!

I have been experiencing an intermittent servo fault on a Kinetix 5500 connected to a VPL motor. Fault code S45: Feedback Comm Fault (Motor Feedback Data Loss Factory Limit)

The system is comprised of 20 devices including distributed I/O, 5 servos, EIP driven solenoid valve manifolds, etc. in a DLR setup. The factory cable has been uncut, connected, and shielded properly. Oddly enough, I have experienced the same axis fault on 2 identical machines and no faults on any other controllers.

Both the motor and controller have been replaced with no success. Any suggestions?

Thanks in advance!

The most likely culprit is the cable between the VPL motor and the Kinetix 5500.

Exposition: VP series are the successors to the popular MP series servomotors from Rockwell Automation. A major feature of these servos is that the feedback is done digitally via Hiperface Digital Servo Link (DSL) protocol and combined with the power in a single cable that connects to a single connector on the motor.

The cables pigtail at the Kinetix 5500 with the power and brake connected to their plugs, and the two DSL wires go into a hefty right-angle connector.

The problem just has to be in the feedback signal, whether it's a physical loose pin or damaged cable, or if it's electrical noise and poor shielding.

Thanks for the feedback Ken,

We thought the same about the cable and was suggested by RA tech support to check the shielding both inside the feedback connector and on the main cable shielding clamp. The problem persisted, so we elected to route a new cable with the same issue. We have been discussing the possibility of noise, but am unsure how to determine if noise is the cause. Since the motor, cable, and 5500 has been replaced, maybe try replacing the feedback connector with a spare??? That is about the only piece of hardware we haven't changed. Thoughts?

I would say double check your wiring. I have been working with those drives couple of times and all the issues I had were man faults either incorrectly wired or bended pins or loose connection. Never had issue with the noise though.

I am guessing you are using Rockwell recommended cables.

Try to find the fault and subcode. These codes can help you search for technotes and can help Rockwell support to guide you. It should look something like:

“Feedback Data Loss Factory Limit”
Fault Entry:
Type: 3
Code: 47
Subcode: 5
Stop Action: 4
State Change: 2


What firmware are the drives? Some early firmware can periodically download bad data to the encoder. Version 2.008 has all the bug fixes and works in all versions of Studio5000 V21 though V28. There are newer versions than 2.008, but they could cause some heartburn if you are using RSLogix versions lower than 24. I had similar issues -thought I was losing motors, but upgrading to 2.008 seems to have been the magic pill.

Other stuff (all of my info is copied from an old email from a now retired automation specialist at our Rockwell Distributor):

*Excess motor cable coil up next to a ground source like the cabinet. The cable can be cut, just a royal pain to get it perfect. Don’t care if the excess cable is in a nice loop in the wire-way, just not on top of each other on the ground plate.

*Shield clamps for both feedback and motor power not tight with exposed shield on both sides of the clamp OR loose encoder wire (they are 26AWG tiny).

*REALLY excess radial loading on the shafting especially when trying to pull through a maximum speed move. More common on an eccentric move with a max velocity limit axis setting OR coupling wasn’t quite plumb.
Reworded a bit, shock loads big enough to vibrate though the motor. Becomes apparent pretty quick with the next statement below.

*Motors with date codes of manufacture between August 2012 to September 25, 2013 had a known product service advisory. The PSA PN-2103-10-003 is somewhat vague on the root cause and does state the problem shows up within the first three hours of operation. From experience I have seen these run for a quite a bit longer before it can show up. More demanding the application sooner it appears. Bottom line, there was some issue with the physical aspect of the internal encoder mounting. 95% of applications would never see it. Problem has since been completely resolved.

*General grounding and bonding (yes I doubt it) with the typical not to painted surfaces statement.

Doesn't the inclusion of the DSL wires with the motor power conductors in the same cable, go against best practices? I understand that there is shielding on the DSL conductors in that cable. Does Rockwell still recommending wiring motor power and encoder cables separately?

We too, are experiencing our second set of Fault codes S45: Feedback Comm Fault (Motor Feedback Data Loss Factory Limit) in just two months with different axes on our first machine with Kinetix 5500s. In both cases examination of the shielding and connectors showed no problems.

So far we have found no solution. The operator resets the fault as it occurs every couple of hours, and continues running.

Rockwell recommends it because that's the only way to get it. The VPL motors have a single connector that includes the motor and encoder. An option would be to install a non VPL motor with separate feedback cable, but you will need to also add a converter card to the feedback to make it serial for the drive.

I'll say that once I flashed up my firmware, my S45 faults went away.

Earthing, shielding, and H-DSL plug termination aside (Check all are ok), I have had something else happen once:


Check the tune on the axis. I've had a K5500/VPL axis that was put into use with the factory tune on it (Auto or Manual tune never done when commissioned). As the mechanism bedded, the motor developed some harmonics, which would interfere with it's own H-DSL feedback and give your exact fault. There are a couple of phone apps that do quick FFT and will be able to show you if your motor has any detrimental harmonics. I use iAnalyzer Lite on iPhone. Android has a different one. A quick 2 minute tune with load solved an issue that had other people replacing motors and cables to no avail.

Thanks Jeev and kekrahulik!

The FW of my 5500 is 7.001, so that does not appear to be an issue. I will try retuning the axis and will check out the apps you mentioned.

Kindly, how was this problem "Feedback Data Loss Factory Limit" solved ?
Thank you ..

Check the motor coupling

I had a S45 Feedback Fault on a 5500 drive and a VPL motor. The motor was turning an harmonic coupling to trim the speed of a roll. Found that the key fell out of the coupling. Every once in a while the coupling would catch and abruptly change the speed of the servo motor. The result was a S45. It doesn't sound right to me but that's what happened.

Bob T. Jones said:

Doesn't the inclusion of the DSL wires with the motor power conductors in the same cable, go against best practices? I understand that there is shielding on the DSL conductors in that cable. Does Rockwell still recommending wiring motor power and encoder cables separately?

We too, are experiencing our second set of Fault codes S45: Feedback Comm Fault (Motor Feedback Data Loss Factory Limit) in just two months with different axes on our first machine with Kinetix 5500s. In both cases examination of the shielding and connectors showed no problems.

So far we have found no solution. The operator resets the fault as it occurs every couple of hours, and continues running.

Click to expand...

How did you solve the S45 Fault?

Kataeb said:

Kindly, how was this problem "Feedback Data Loss Factory Limit" solved ?
Thank you ..

Click to expand...

solved by putting a pulley with belt to eliminate the vibrations on the ...
...

How many packets are lost to occur S45 fault?

Hi all,
Do you know that how many packets are lost to occur S45 fault?
or do you have any idea to figure it out?

I think if 8 consecutive messages are wrongly received, the drive will activate this alarm.
But why this is important ?