S7 Profibus DP Scan Rate

Hi All
I need to find out how fast (worst case) an input located on a profibus network can be scanned (updated) in an S7-319F. On a Controllogix you can set the scan rate for remote racks. Is this possible in S7?

I have an input that is sending pulses and I need to work out the highest frequency that I will be able to recieve in the PLC. The expected maximum input frequency is less than 15Hz. Input card is a standard input not a high speed counter.

Let me know what other info I need to supply to help get an answer.

We need to know the I/O you are scanning

I have a Profibus DP diagnostic boards that shows packet send and received and a time at which they are sent. Profibus DP goes and 12 MBs. Due to over head the peak seems to be around half that. I believe SST/Woodhead claims 6000 bits per millisecond.

So what is the I/O you are scanning? A few I/O devices with a large I/O count scan faster that many devices with a small I/O counts.

There are about 8 nodes on the network. 2 are et200m (with safety IO) IO racks, the rest are field devices. Racks each have about 6 IO cards. Can get more acurate numbers later.

I use the ET200S counter module for applications like this. If the count rate goes up or more I/O gets added you don't have to worry about the bus update rate.

I have finally had a chance to get actual numbers for the network.

There are 9 nodes (including master), 2 are ET200M racks with F-IO.

IO count is as follows

Node 2
64 input bytes
32 Output bytes

Node 3
22 input bytes
24 output bytes

Node 4
12 input bytes
12 output bytes

Node 5
12 input bytes
12 output bytes

Node 6
200 input bytes

Node 7
4 input bytes
2 input bytes

Node 8
4 input bytes
4 output bytes


Transmission rate is currently 1.5Mbps, what are some guidelines for transmission rate?

Profibus bus parameters in the HW config list the watchdog as 176.3ms, Ttr as 46.0ms and Ttr typical as 12.6ms.

Does this mean the normal scan rate is every 12.6ms and the worst case is 176.3ms?

How fast is an input "scanned" ?
With "scanned" one must mean all the delays that it takes from the physical change of the inputs electrical state, to it can be observed as a change of state on an output.
It is not enough to consider the transfer rate on Profibus, you have to consider:
Propagation delay in the input module.
Cyclical scanning of the backplane of the profibus node.
Transfer of the data via Profibus to the master (*).
Cyclical scanning of the i/o image in the Profibus master (**).
Cyclical scanning of the user program (**).
If the output is also on Profibus, then you have to add the delays all the way back.
*: You can find this time by investigating the Target Token Rotation time (TTR) in the Profibus Bus parameters. This is the worst-case transfer time. The typical time may be less (aka TTRtyp), but you must always consider worst case.
**: With hardware interrupts, you can bypass these delay times.

This link gives you an estimate on the update time:
http://support.automation.siemens.com/WW/view/en/21178141

There should also be a better link to a more precise method of eaccounting exactly for all the delays, but I cannot find it now. I think there is one for ET200S and one for ET200M.

Profibus bus parameters in the HW config list the watchdog as 176.3ms, Ttr as 46.0ms and Ttr typical as 12.6ms.

Does this mean the normal scan rate is every 12.6ms and the worst case is 176.3ms?

Click to expand...

'normal worst case is' TTR= 46 ms.
If the token does not make a complete roundtrip within 176 ms, it triggers a busfault.
Notice, like I said above, the Profibus TTR time is just part of what makes the complete scan time delay.

Hilscher CIF60-PB PBus master scan-rate.

Has anyone input for Scan-rate setting with Hilscher CIF60-PB master with sycon PBus configurator/simulator? How scan-rate is related with Bus Cycle time & Slot-time/Target Token Rotation Time/Minimum Slave Interval Time?

Thanks in Advance.

Regards,
Bidyut

JesperMP said:

If the token does not make a complete roundtrip within 176 ms, it triggers a busfault.

Click to expand...

I would question this statement. The TOKEN is a mechanism used for multi-master Profibus networks to control the communications on the Bus. The 175ms watchdog time is set for as similar use as a control panel watchdog. Everytime the master recieves an error free response telegram from a slave it resets its watchdog timer for that slave, however if the watchdog timer expires (i.e. the slave is disconected or powered down) the slave will enter communications failure due to watchdog timeout. This function also happens visa-versa, the slave is always monitoring the master for its request telegrams and resetting its watchdog timer, if it expires the slave fails to its safe condition.

Hope this clarifies.

Cheers

Scott

Tycon Automation Ltd