Hi all,
S7-300, CPU 315-2 DP. PROFIBUS DP at 1.5 Mbit/s, eleven slaves, ET 200M stations and three MICROMASTER 440. Ran clean for twelve years.
Last month the electricians extended the segment to a new packing line, two more ET 200M and an MM440. Segment is now about 300 m by the drum count.
Since then the three new slaves and the last old one drop out a few times an hour. The diagnostics buffer fills with "distributed I/O: station failure" and BF flashes on the far stations, solid on the near ones. I've checked the termination at both ends and swapped the connector on station 14, no change.
Do I need to lower the speed, or is one of the new stations pulling the bus down? It ran twelve years and only started since the extension, so where would you start on a segment that long?
Same thing on the old press. We had a profibus run that started doing exactly this after an extension, and it turned out the electricians had used a different cable for the new bit, looked the same colour under the lid. Is the new bit the same part number as the old, off the same drum?
Same drum, same part number, I checked the print on the jacket. Old run and new run are both 6XV1830-0EH10, so it isn't the cable type. Baud rate is set in HW Config on the DP master system, 1.5 Mbit/s, and nobody's touched it.
300 m at 1.5 Mbit/s is the whole problem. Type A cable per segment: 1200 m up to 93.75 kbit/s, 1000 m at 187.5, 400 m at 500 kbit/s, 200 m at 1.5 Mbit/s, 100 m from 3 Mbit/s up. Your old 200 m sat right at the limit, the new 100 m is past it, and the far slaves see the worst of it.
Two fixes. A repeater at about 200 m, 6ES7972-0AA02-0XA0 if I remember right, two segments each inside the limit, still 1.5 Mbit removed link Or 500 kbit/s in HW Config on the DP master system properties, download, and every slave follows, they detect the rate. 500 kbit/s gives you 400 m and eleven slaves won't notice.
This is new to me, and I'm not sure I follow. Why does the allowed length go down when the speed goes up? Is it the cable getting worse, or the signal?
It's the signal, not the cable. Attenuation and edge rounding both grow with length, and the faster the rate the shorter each bit, so there's less time for a rounded edge before the receiver samples the thing wrong. Same cable, same driver, less time per bit. Roughly.
If you want the long version: https://plctr.com/profibus-segment-length-and-baud-rate/
Dropped it to 500 kbit removed link HW Config, DP master system properties, Network Settings, transmission rate 500 kbit/s, downloaded. Every slave came up, the MM440s too, and the ET 200M were happy before the download had finished. Eighteen hours, no station failure in the buffer, BF solid on every station.
So it was the extension pushing the segment to 300 m at 1.5 Mbit/s, where the limit is 200 m. The lower rate fixed it, 500 kbit/s gives 400 m. The repeater is on order anyway, I want the margin for the next extension.
One thing I can't explain, the old 200 m run sat at the limit for twelve years and never dropped once. Thanks dieter_k, sparkymike.
Twelve years at the limit. Somebody built that run right.