MM4 causes Profibus faults. What to check ?

Hi all.

I am supporting a collegue of mine who is onsite and I am assisting him remotely.

This is the case:
Main panel. Has 319-3PN/DP CPU. 1st DP port X1 is connected to many ET200S and 1 MM4 in the same panel. Works without problems.
Distributed panel, 100 m away. 2nd DP port X2 in main panel is connected to many ET200S and 2 MM4 in this panel.
Bus speed is 187.5k.
All MM4 have output chokes.
When we start the bigger of the two MM4 in the distributed panel, profibus net X2 goes on BF.

What we have checked:
Tightened all connections everywhere.
Checked that screen on motor cable is clamped to earth bus bar.
Checked that screen on profibus cable is clamped to earth bus bar.
Checked all Profibus connectors.
Checked earth to both main and distributed panel. Supply cable earth to both panels are connected to same earth star point.
Big MM4 has been replaced (!).
As far as it is possible, Profibus cables have been laid seperately from power cables.
Nothing of this helped or revealed anything.

Then we increased the Retry Limit in the Bus profile from 1 to 3. This made the BF disappear.

OK, it works so far, but I think that this is not an extreme case, and since everything is done properly, it should not be necessary to tweak the bus profile.

Any other ideas as to what we can check or do ?

I had face to face same problem 5 years ago,I arranged screening of profibus cable again and changed all of profibus connectors and problem fixed.Probably your problem is same,check again all of these again..

Have seen similar problems before. With machine at standstill all's well, when machine is started profibus communication failures like crazy. Sometimes machines would fail every start-up or while running and communications would fail frequently, sometimes commuinications would fail very intermittently possibly no problems for > 3 days. For us problem usually exist in profibus connector with termination options. Typically isolated by pulling profibus connectors one at a time and checking resistance at each connector. If resistance wasn't < 1 ohm we would replace it. Sometimes could correct resistance by sliding termination switch a couple of times or tapping on connector. Some series of connectors really gave us a difficult time.

I think we will try with a new profibus cable and a new connector, just to the troublesome MM4.
If that cures it, we know where to start looking.

Do you have a BIG ground cable between the two cabinets? That sounds like drive noise is trying to find a path to ground and the easiest path is through the Profibus cable shield.

Keith

There is only the ground in the supply cables to either cabinet, tied to the same star point at the main supply.
That is also one thing we will investigate, but not easy to find 100 m ground cable in a hurry.

MM4 drives are notoriously bad for generating noise.

You mentioned that the motor cable shielding is connected to the earth bar; this in not correct. The motor earth and shielding should be connected directly to the drive. This is because the drive is the source of the noise which must get back to the drive to be cancelled out; if the route back to the drive is through your panel then all the noise will pass through your panel. Another common mistake is to use a metal gland in the bottom of the panel and earth the shield at that point.

Whilst on the subject of the shielding wire, ensure that you use the full 360 degrees of the braiding and not just a tail. This is important because the noise generated is high frequency noise and subject to skin effect so surface area is more important than cross sectional area. Whilst the switching frequency of the drive is only quite low, maybe 4Khz, the rise time of the voltage when the transistors switch on is extremely rapid giving rise to an instantaneous high frequency.

Chech the MM4 manual carefully, I think that the recommended connections also include a braided connection to other related components.

Good Luck,

Nick

Check Motor...yes motor. I had a customer with similar problem and it was the motor ground terminal. Try this. Disconnect the output of the drive and see whether problem goes away (with old retry rate).

Manglemender, we have followed the guidelines for connecting the screen to the book. It says several places in the manual to connect the screen as we have done it. I know that VFDs generate uneven currents in the fases, so there is often a higher earth current that has to go back to the VFD to be cancelled out. But I dont think that applies to the screen. I might be wrong, I am willing to learn.
And yes, of course we use proper clamps over an deisolated bit of the cable, so there is a very good contact from screen to earth bar.

Nick. I will try to check that. But I think the drive will immediately fault on missing output current. Is there are way to override this ? Is that what you mean by "old retry rate".

Hi,

Are there in the two MM4 converters the encoder feedback motor card add to the Profibus fieldbus com card also ??

regards

No, no encoders.

Jesper,

Attached is the recommended connection diagram for a MM4 drive. Please note that the motor cable screen is shown connected to the motor at one end and at drive PE teminal; not anywhere else.

Cabling between the filter and chokes is also shown as screened.

Nick

Hi Nick.

I think this picture from the manual shows better how the screens should be grounded, and that is also how we have done it (the version on the left in the picture).
How else do you get the screen and the VFD earth together ? There is no place on the VFD's earth to put a clamp over the screen of any of the cables.

In the meantime we have found a few things that have improved the situation marginally.
Some earth connections between panel sections were loose. Have been tightened.
Earth straps between panel mounting plates and earth bar were too small, bigger have been installed.
Screens on Profibus cables have been clamped to earth bar, before they were only therminated in the connector.
The problem didnt entirely disappear though.

The big difference is if the very last node on the bus is connected or not. It is a small TP177-4 operator panel.
We checked everything on this panel. Yes, there is an earth connection directly on the panel to the earth bus bar.
We will probably move this panel from Profibus to Ethernet to rid the problem once and for all. Means laying another 100m ethernet cable.