Back to top

What are the limitations of AKD Ethernet IP?

Often we are asked what the maximum axis count is for AKD Ethernet IP or what the limitations are.

There are always limitations to an Ethernet based fieldbus.

In the case of the AKD:


1. Update rate. The AKD's Ethernet port that Ethernet IP uses as the communications interface can only process messages from the PLC every 20mS without workbench and 40ms with Workbench. This is generally not a problem unless you're trying to do high cycle rate motion.. If your application cycle time is faster than 20msec then this architecture  is not the correct selection. Also, Ethernet IP is a non-deterministic fieldbus so if you want or need determinism you will need to consider our KAS/PCMM/PDMM platform with Ethercat slaves. The PCMM/PDMM can serve as the motion controller and interface to the AB PLC via Ethernet IP if the PLC is a must. Keep in mind some functions take longer to execute than others ( i.e. the enable may take upwards of 300-350msec to complete ).


2. The AKD only supports torque, jogging, and motion task (point to point moves) over EIP. This means the AKD doesn't use or need CIP Motion (AB attempt at real time motion control). This can result in a cost savings. Example: 1769-L30ERM has 4 axis of motion and the 1769-L30ER has only EIP communications. Both PLCs will work with the AKD EIP drive but the motion ready processor costs more. However the AKD Ethernet IP drive works as follows: You load a block with typical motion task parameters ( indexing ) and trigger the enable ( typically using the Add On Instructions for AB PLC users). This means that you can start two axes at the same time but you cannot guarantee synchronism over Ethernet IP ( remember motion tasking is point to point motion ). In that case you will need to either use Electronic Gearing between a master and slave drives with traditional hardwiring of the EEO output to the Encoder input of the slave drives and communicate Ethernet IP to each drive for diagnostics, etc.  or use the KAS/PDMM/PCMM platform where synchronism is required between axes or where events must be on a clock and deterministic.There are no S-curve acceleration capabilities and  no changing the motion profile on-the-fly over Ethernet IP with the AKD-Ethernet IP drive. The AKD_Move AOI is not designed for blended moves ( start one move and blend into another ). If blended moves are absolutely required for the application, the block method will be required. The block method is where the user creates a chain of predefined motion tasks using Workbench and sets one or more motion tasks to blend into the other. Using pre-defined motion tasks also allows the use of S-Curve acceleration using the Motion Profile Table since this is achieved in the drive and not over Ethernet IP. The Ethernet IP landing page has a link where additional AOIs (AKD_Start_MotionTask & AKD_Set_Motion_Task Block ) can be downloaded and added to your project for using the block method. In that zip file there are 2 AOIs: One is for setting the motion task parameters ( if the predefined moves need to be created from the PLC or HMI ( i.e. recipes ) and the other is for pointing to the motion task number ( the first move to start the chain of blended moves ) and executing it. For non-AB controllers that support Ethernet IP cyclic messaging, the programmer will have to set the correct byte in the command assembly for setting the block number and use the control word to start the block.AOIs can be downloaded and added to your project for using the block method. In that zip file there are 2 AOIs: One is for setting the motion task parameters ( if the predefined moves need to be created from the PLC or HMI ( i.e. recipes ) and the other is for pointing to the motion task number ( the first move to start the chain of blended moves ) and executing it.


3. We supply AOI blocks (Add On Instruction). The issue is these blocks have to be executed one at a time and the block has to finish before the next block is triggered. May seem obvious but usually the PLC is running a much faster update rate than 40ms. It’s hard writing a state machine in ladder logic to keep one command from over writing another. As stated before, the intent of AKD Ethernet IP is to perform independent point-to-point moves in a non-high-speed application.


4. Most people familiar with AB drives, look to do the drive setup in the PLC GUI and also want the ability to have the PLC download the parameters/setup to a drive in the case a drive is replaced due to failure or other reason. There is no practical way to do this with the AKD over Ethernet IP. The user will have to commission the AKD drive, archive the parameter file, and in the case of drive replacement download the parameters ALL using Workbench.


5. As far as an maximum axis count, the limitation is primarily the bandwidth of your network. Use a managed switch to limit crashes and IP address conflicts.  
    If the axis count is high, you may want to consider the KAS/PCMM/PDMM platform as a different architecture where the KAS/PCMM/PDMM can handle the motion control  and  the PLC handles the system/machine supervision, HMI, recipe data, other non-motion devices and operations, etc.