Devicenet Configuration for S.E.W. Eurodrive SNI Gateway

[raven]

Hello


I am trying to get the device net to communicate with an S.E.W. Eurodrive
Gateway Model UFF41B . According to the technician the Gateway needs to
have 35- (8 bit) words produced and consumed. However I am only seeing 6 (8
bit) words when I use the Telnet command and do the exec "dnShow"
I have set the macid on the gateway to = 8
In the KRC I have the following settings
....
.....
.....
[1]
Macid=8 ; SEW Gateway macid


The Devnet=2,dnInt,dn2drv.0 has been uncommented in the iosys.ini
My i/o mapping is commented out in the same file
I have no errors when I enable the driver devicenet comes online
I perform the Telnet command and get

I: 00 00 02 00 00 00
O: 00 00 00 00 00 00

I then mapped the I/O as follows

INB0=8,0,x6 ;SEW gateway using IN$[1-48]
OUTB0=8,0,x6 ;SEW gateway using OUT$[1-48]

Save and close the file
Then I perform a Reset, then a Reconfigure i/o driver
No errors when finished

I can now see that on word 3 bit 2 is high on the controller (this is coming
from the Gateway)
I cannot however set any outputs and see them on the gateway
SEW Technician says that there is something SEW has that is called an EDS file
they need to install when using RSlogix 5000 in order to communicate with
the gateway via devicenet. I'm not sure if this EDS file can be installed on
the KRC or should be.

Do you have and thoughts as to how I should proceed?


I'm also assuming I'm going to have the same problem on my FESTO Pneumatic valve banks since festo also has an EDS file

[SkyeFire]

The EDS file information is generally not relevant if you are using the DeviceNet port built into the MFC card.  This is the port that uses the dn2drv.o driver.  EDS information like Product Code, Vendor ID, etc is usually only necessary if you are using the add-on two-port LPDN DeviceNet cards, which use the DNSCx drivers.

This is puzzling.  Your DeviceNet network is obviously working correctly, because you are able to interrogate the devices using dnShow.  And if you can do that, the dn2drv driver can access the Produce/Consume byte sizes and self-configure for handling the slave properly.

Your assignment

Code:

[1]
Macid=8

is in the file DEVNET.INI, correct? 

In IOSYS.INI, your assignments

Code:

INB0=8,0,x6 ;SEW gateway using IN$[1-48]
OUTB0=8,0,x6 ;SEW gateway using OUT$[1-48]

are just below the header [DEVNET], correct?

Given your I/O assignments, and the fact that the second bit of the fourth input byte is active, you should see that $IN[26] is True in the I/O Monitor.  Is it?  Are any of the first 48 $IN variables true? 

Just in case, check your network physical quality.  Especially check that you have the proper termination:  on each of the two furthest points of the network, you should have one 120-ohm resistor between the CAN+ and CAN- wires, making for a parallel bus impedance of 60 ohms.  This is unlikely to be the problem, since your Telnet diagnostics are working properly, but it's always worth checking, as it seems to be the one thing that always gets done wrong on DeviceNet networks at least once.

[raven]

Thank you for responding Skyfire

To answer your questions

Code:
[1]
Macid=8
is in the file DEVNET.INI 

In IOSYS.INI,
Code:
INB0=8,0,x6 ;SEW gateway using IN$[1-48]
OUTB0=8,0,x6 ;SEW gateway using OUT$[1-48]

is just below the header [DEVNET]

when I open the I/O monitor I see IN$[18] is set true not IN$[26] no other inputs are set true.

There is an end of line resistor on the X801 connector in the  KRC across pins 2 and 5 and another at the end of the line .

[SkyeFire]

when I open the I/O monitor I see IN$[18] is set true not IN$[26] no other inputs are set true.

There is an end of line resistor on the X801 connector in the  KRC across pins 2 and 5 and another at the end of the line .

Hm!  Now that's interesting.  You said that the Telnet monitor showed an Input pattern of:
I: 00 00 02 00 00 00
O: 00 00 00 00 00 00

That would match if the left-most byte was the 0 byte -- I had thought that the 0 byte should be the right-most byte.  But I could be misremembering.

Everything looks good.  The only idea I can offer at this point is to either find a way to force on particular signals from the SNI and look at what inputs show up on the robot, or try setting each $OUT from 1 through 48 one at a time and see which inputs the SNI receives.