Page 1 of 1

PROFINET Basics

Posted: 22 Mar 2024, 13:18
by Webinar_Question
In case of having more than 1024 byte to transfer is there any possiblity to fragment image in real time processing?

Re: PROFINET Basics

Posted: 22 Mar 2024, 13:20
by Alex.W
Webinar_Question wrote: 22 Mar 2024, 13:18 In case of having more than 1024 byte to transfer is there any possiblity to fragment image in real time processing?
PROFINET distinguishes two diffent mehtods to exchange data:
  • 1. Cyclical data exchange: This is the fast and deterministic data exchange between a PROFINET Controller and a PROFINET Device. Data is exchanded via Ethernet frames and have a limitation of up to 1,440 bytes per frame (= Application Releation (AR) between Controller and Device)
  • 2. Acyclic data exchange: Also known as record service, every PROFINET devkit supports this service. In addition to cyclic data exchange, data records can also be offered at each subslot. Each data record can have a length of up to 64kB, although it should be noted that not all controllers or devkits also offer the 64kB. Common are 8..32kB. However, several data sets can be offered. The data exchange is initiated by the user, i.e. a job is created in the PLC user programme, for example, which is to read or write a specific data record. Of course, you can also use this service from your own
    tool to parameterise your own device.

Is there a function like the rack optimisation in CIP EthernetIP for head devices in PROFINET?

Posted: 22 Mar 2024, 13:21
by Webinar_Question
Is there a function like the rack optimisation in CIP EthernetIP for head devices in PROFINET?

Re: Is there a function like the rack optimisation in CIP EthernetIP for head devices in PROFINET?

Posted: 22 Mar 2024, 13:23
by Alex.W
Webinar_Question wrote: 22 Mar 2024, 13:21 Is there a function like the rack optimisation in CIP EthernetIP for head devices in PROFINET?
Basically, we do not need this function with PROFINET. The performance of the controllers and the network is so good, even at 100 MBit/s transmission speed, that 64 devices, e.g. servo drives or modular field devices with 20 analogue channels can be updated per millisecond. In addition, another 50% of the 100 MBit/s are free for TCP data exchange, e.g. an HMI or a connection to the ERP system or camera data or OPC UA or ...

Performance can be further increased by a function that was provided with the PROFINET Perfromance Upgrade (since ~2012). The PROFINET Perfromace Upgrade makes it possible to transfer data optimised via the bus using the DFP (Dynamic Frame Packing) function. For example, 256 drive axes can be transmitted in one millisecond, which are distributed across 256 individual axis drives. There are therefore 256 real PROFINET participants.

It should be noted that this function is unfortunately not offered by all devkits.

Trace

Posted: 22 Mar 2024, 13:24
by Webinar_Question
Trace doesnt has any relation to trace function (Data acquisition) or it is just for diagnostic function?

Re: Trace

Posted: 22 Mar 2024, 13:25
by Alex.W
Webinar_Question wrote: 22 Mar 2024, 13:24 Trace doesnt has any relation to trace function (Data acquisition) or it is just for diagnostic function?
The Trace feature is something required to develop the PN device. With the trace function you can get infromation from the stack with no influence on the real-time behaviour of your device.

The diagnostic function of your device is something that you offer yourself. How a diagnosis is created and which services are used is part of every devkit. You will find examples that you can customise to your needs.

Controller configuration vs PC parameterization

Posted: 22 Mar 2024, 13:25
by Webinar_Question
What is the difference between the controller configuration vs pc parameterization?

Re: Controller configuration vs PC parameterization

Posted: 22 Mar 2024, 13:27
by Alex.W
Webinar_Question wrote: 22 Mar 2024, 13:25 What is the difference between the controller configuration vs pc parameterization?
The controller configures the device during start-up. The settings made during project planning are transferred to the device and checked to see whether the device meets the expectations, e.g. whether the order number is correct or whether the firmware version has a certain minimum version.

The device can also be parameterized using external tools. This can be a web server on the device or a separate tool that parameterizes my device via RPOFINET data records. In the classic sense, this is done for complex devices such as drives or laser scanners. In this way, not everything has to be configured in the controller engineering tool, but the know-how can be "hidden" in your own tool, e.g. because you don't want to show all device dependencies publicly.

PROFINET also offers a standard, called TCI (= Tool Calling Interface), to call up your own parameterisation tool from the engineering tool of the controller and already transfer the data, such as what is contained in the project, which configuration was selected, which (IP) address settings were made. You can also save the data from your own tool together with the controller project. This makes the work of the system builder or operator much easier.

Is there a recommendation for a switch-chip with 2 ports to integrate into a device?

Posted: 22 Mar 2024, 13:40
by Webinar_Question
Is there a recommendation for a switch-chip with 2 ports to integrate into a device?

Re: Is there a recommendation for a switch-chip with 2 ports to integrate into a device?

Posted: 22 Mar 2024, 13:42
by Alex.W
Webinar_Question wrote: 22 Mar 2024, 13:40 Is there a recommendation for a switch-chip with 2 ports to integrate into a device?
The Microchip KSZ8863 is an option. Please keep in mind that PROFINET sees this device as a PROFINET device with 2 PROFINET ports. The Switch controller needs to be a managed Switch Controller (Layer 2+) and your PROFINET solution needs to bring the management functionality to manage this switch. You are not only embedding a Plug-and-Play Ethernet Switch – you are embedding a Managed Ethernet Switch.

Only then the PROFINET Stack can provide LLDP data, SNMP data, VLAN settings and miscellaneous other settings. Practically you would connect the switch with the MII or RMII interface to the Ethernet port of your MCU and in parallel with SPI to access the management registers. The frames coming to your MCU and leaving from the MCU will have a so called tail-tag, this tag defines on which Ethernet port the switch has received the incoming frame resp. which Ethernet port the outgoing frame is supposed to go out. This tail-tag is stripped off by standard Layer 2 Ethernet drivers, latest the TCP/IP Stack is discarding it. You will need to retain this tag by your own measures in your used Switch Management Software.

There are PROFINET stacks available with integrated Ethernet Switch Management.

Where to define if a device is modular or compact?

Posted: 22 Mar 2024, 13:42
by Webinar_Question
Where to define if a device is modular or compact?

Re: Where to define if a device is modular or compact?

Posted: 22 Mar 2024, 13:43
by Alex.W
Webinar_Question wrote: 22 Mar 2024, 13:42 Where to define if a device is modular or compact?
In the GSDML of your own developed device.

Is there standard PROFINET support for remote device firmware update?

Posted: 22 Mar 2024, 13:44
by Webinar_Question
Is there standard PROFINET support for remote device firmware update?

Re: Is there standard PROFINET support for remote device firmware update?

Posted: 22 Mar 2024, 13:45
by Alex.W
Webinar_Question wrote: 22 Mar 2024, 13:44 Is there standard PROFINET support for remote device firmware update?
No, no mechanism for updating the firmware has yet been defined.

Is it possible to have a compact and a modular head unit in the same rack?

Posted: 22 Mar 2024, 13:45
by Webinar_Question
Is it possible to have a compact and a modular head unit in the same rack?

Re: Is it possible to have a compact and a modular head unit in the same rack?

Posted: 22 Mar 2024, 13:46
by Alex.W
Webinar_Question wrote: 22 Mar 2024, 13:45 Is it possible to have a compact and a modular head unit in the same rack?
The device is either defined as a compact device, in which case the user has no option to insert (or remove) modules.

However, you can also design a modular device in such a way that certain slots are already filled, which the customer can then change, or that these are permanently configured. In any case, it is a modular field device. The devkit will allow all possibilities. Ultimately, the GSDML is decisive here.

Re: PROFINET Basics

Posted: 05 Apr 2024, 23:15
by Robbie_B
In IO-Link there is a standardized way to send "commands" the a field device to perform one time actions within the device. For example, resetting of a counter. It is done with the SystemCommand index. The IODD file is often used to graphically show these commands as "buttons" which the user can push to execute the command.

Does PROFINET have a similar mechanism to SystemCommand? If nothing is standardized, is there a common way that vendors have implemented this type of feature in the past?