Strange Noise Problem 3

[simonsk]

I am witnessing a strange situation on a data acquisition system.

I have one data acquisition comms module (RS232) and two RTD temperature modules. The RTDs monitor the temperature in 4 mixing tanks.


The data acquisition modules are housed in a separate metal, (earthed) enclosure along with a DC power supply. All of the RTD cables are shielded and connected to earth at one end.

The main control panel houses three inverter drives (all with manufacturers input filters fitted), and some other control gear. Motor cables are all screened and screens are terminated correctly.

The system has been running perfectly for several months, until last week. I had a call to go to site because the temperature readings were 'all over the place'.

Whenever any of the four mixer motors are now run, the temperature readings float from -50 to +200. Stopping the motor returns the temperatures to their normal level.

The only event that has occurred recently is that some sub-contractors dropped something on the armoured supply cable (which is outside), and fractured the insulation on the case of the cable and some of the conductor insulation. Due to heavy rain, water has leaked into the cable and it has eventually tripped out the supply. Is it possible that this may have damaged the EMC filters?

Thanks in advance

Simon

 

[itsmoked]

Yes or a loose wire with a non open but high resistance connection is now causing havoc.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[itsmoked]

More thoughts.

You replaced the cable correct?

I'm not sure what kind of [red]all with manufacturers input filters fitted[/red] you have but a typical type has the hot and the neutral run thru it and a ground hooked to it. If the ground is open, look out! The case of the filter and the chassis it's hooked to becomes exactly half the line voltage. More than enough to give you and the RTDs a thrill you don't want.

Find a separate ground and measure all the different pieces of metal that the system is comprised of with different conditions - motor on/off and see if you find anything unexpected.

Also measure from the RTD leads to system ground leaving everything connected, see if you see any voltage over 3 or 4 volts.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[Skogsgurra]

I second what Keith says. 100 %

It may also be interesting to use a scope to look for higher frequency components that pour like water through a sieve from live to ground connection of the filters. If filter ground is interrupted or is going hi impedance anywhere, you will have lots of interference.


Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[simonsk]

Thank you for your comments

There are two types of drive, a Moeller one and two Mitsubishi drives. The Mitsubishi ones are fitted with the manufaturers foot filter FR-E540 range of drives (Moeller one similarly has the recommended filter for that device).

The filters are installed as per manufacturers recommendations and correctly grounded (and all has been running fine for months).

Do you think that if the supply ground had been damaged (open circuit) on the trailing cable, it could have caused damage to the filters in the panel?

One odd thing I did notice last night (it was getting late and dark so could not pursue further), was that between Neutral and Live I was getting 240V but Live to Earth I was getting 247V.

Thanks

Simon

 

[Skogsgurra]

How much were you getting between Neutral and Earth?

Try and connect the two (if your country's code allows) and see what happens to the interference.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[simonsk]

About 4V

 

[Skogsgurra]

Where in the World are you?

Is this a TN, IT or TT system?

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[simonsk]

In the UK - TN.

I have just spoken to a friend of mine who is involved with EMC design and he suggested that there is a possiblilty that the filters may have been damaged when the supply shorted - any comments welcome.

Simon

 

[Skogsgurra]

Not at all unlikely. Continue talking to your friend.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[ozmosis]

As a test, try taking your EMC filter out of circuit and connecting your mains power directly into the inverter. If an EMC filter is damaged, they can cause problems worse than if they are not actually there.

 

[sreid]

My experience is that when there is suddenly lots of noise vs. a little noise is that the problem is conducted noise. This happens to me in new machines that have a motor phase shorted to safety ground usually through the motor cable shield. Easy to check, Ohm motor leads to ground with the power off.

 

[simonsk]

I have been back to site this evening and there does appear to be a noise problem.

Unfortunately I did not have an oscilloscope with me (only a DMM).

I measured from a separate ground to the panel ground and it was around 3 ohms. When I ran one of the drives the resistance jumped to 39 ohms.

In the system there are two Mitsubishi FR-E540 drives and one Moeller DF51 drive. On my last site visit I thought they were all problematic but it is only the two Mitsubishi drives. I think that the filters fitted to the Mitsubishi drives must have been slightly more sensitive to the incoming supply failure than the Moeller.

Further evidence that the problem is noise related was confirmed by dropping the carrier frequency to the minimum setting on the Mitsubishi drives. This almost eliminated the noise and the data acquisition system ran ok.

The Moeller drive can still be run with a reasonable carrier frequency (to make the motor sound sweet).

I think the next step is to replace the two Mitsubishi filters – any comments?

I am away for a few days now.

 

[itsmoked]

Depending on the cost, and your time, I think I'd replace the filters or one of them if they're real pricey. If that solves the problem on that drive, change the other too.

I'd also check/redo the grounds with those abnormalities.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[ozmosis]

If dropping the carrier frequency has reduced the interference then this is usually an indication that radiated interference from the drive is getting into the complete circuit and the bonding of the installation is not quite up to scratch, as Keith mentions. If it had been like this from day one I would say your problem is larger but the fact that this only happened after the cable was damaged, then look here first. Also, is it possible that somebody has increased the carrier frequency since the accident? If yes, then keep it low if you can.

 

[waross]

I would do some more checking on the damaged cable. Damaged insulation and tripping when wet are important. this is not a perfect world and the total installation may be less than perfect.
I believe there is a possibility that your grounding system may have been compromised by the original incident. It is possible that a poor connection in your grounding system was adequate before the initial incident but was further degraded by the fault current of the original incident.
In regards to your resistance apparently jumping from 3 ohms to 39 ohms:
This is not reasonable.
What is reasonable is that the operation of your drive causes a current flow over the ground connection. This current flow causes a voltage drop. Your meter interprets the voltage drop as a change in resistance. The difference between ground potential and neutral potential supports this suggestion.
I believe that the first step is to repair the original damage with particular attention to hidden damage of the grounding system. It is possible that you will find that some conductors and/or connections have been damaged by the original fault and until you verify that the supply system is properly grounded you may be wasting your time checking filters.
respectfully

 

[simonsk]

Thanks for the comments.

The carrier frequency has not been changed by another operator.

I am very particular about bonding and all supply/motor cables are screened (with screens connected at the panel's main earth point). All the drives and filters have dedicated earths running to the main earth stud.

I have checked the repaired cable and all seems to be ok.

The supply to the equipment is about 30 metres away in an old barn. The supply has two outlets and there are two cables that run to the equipment (one cable is for the main panel and the second cable is for a separate piece of equipment). The noise problem is present no matter which cable is used. The distribution point is quite old and the barn is not 100% watertight. Could the recent heavy rain have compromised the supply bonding at the distribution panel?

I have not been to site yet so have not ordered/fitted any new filters.

Any further comments?

Simon

 

[simonsk]

Anyone got anything else to add on this, suggestions etc? I am planning a site visit soon.

The customer is quite happy to leave it running 'as is'. With the reduced carrier frequency on the two Mistusbishi drives, everything is running fine (apart from the audible noise, but they don't mind that).

Ideally I would like to replace the filters and double check everything else.

Has anyone else had experience of old agricultural power distributions in wet weather?

Thanks