Inverter Noise

Hi,

Has anybody had problems with inverter noise on a plc or dcs system?
We have a CT 75Kw inverter running 11 6Kw fan motors on a bus system,the dcs system is communicating on a rs485 network,the problem is that very often the network going to 1 of the dcs systems freezes up and shows comms failure or error.If we disconnect the inverter and start the fans by forward/reverse contactors the problem disappears.
I was wondering whether anybody else had this problem or could maybe give me some solution apart from the obvious trail and error fix's

We had a similar problem to this.The problem was solved by screening the cables something that should have been done in the first place.

We use a 'Rasmi' choke on all inverter wiring. It's a donut-like device and passing the wiring through this works a treat in 'absorbing' emi.

With an ASI system I tried metal cases and screen etc. But only cure was to move the inverter next to motor. Never again will I put an inverter and PLC in same cabinet. Still ran the inverter from ASI with slave in same box without a problem.

An adicional mesure for less noise you can slow down the switching frequency from the IGBTs. The motor will do small audible noise, but must not afect the task.The drive as an parameter to do it.

Inverters inducing electrical noise into other comm/instrumentation wiring is a long-standing, little-understood, and worsening problem.

Deciding to never put PLC's and inverters in the same cabinet is not the solution. Inverters in adjacent cabinets can be just as much trouble.

Sources for the problem are really three-fold: Radiated noise from the motor leads due to the high frequency content in the fast-rise-time PWM pulses going to the motor, Power supply noise caused by the input rectifiers in the inverter using input AC non-linearly, and common mode (ground) currents in common grounds between other equipment and the inverter.

I have had plenty of bitter trial-and-error experience with these problems and have learned the hard way that this is NOT the way to proceed.

After you have checked to be sure you are using best-possible wiring, shielding, and grounding practice on your other equipment, it is necessary to determine which of the above-mentioned three sources is making trouble. This takes expert analysis and you should engage a person that has a solid reputation in the field. Often, but not always, these are people who work for companies that make reactors, wave traps, and similar equipment. MTE and TCI are the largest in that field and both are based in Milwaukee.

Once the source has been determined, a solution is usually an off-the-shelf product but may not be cheap. Particularly in the case of power supply and common mode noise, it is often cheaper to install a separate protected super-clean AC supply for the instrumentation and let the inverter just go on doing its evil thing! This is especially true if the inverter is high hp.

Don't be seduced into just trying something! Years ago, I put a poor customer thru a very expensive exercise running all motor leads in shielded cable imported from Europe and only made the problem worse! It ends up the problem was common mode noise which wasn't improved a bit with the shielded cable. Very awkward and the hard way to learn, I assure you!!

Just like Dick, I have had problems with PLCs and VFDs in separate enclosures, and I have PLCs in mounted right the VFD enclosure work just fine.

A trend I see too much of lately is specifiying 18-pulse or IEEE-519 compliant VFDs to try to avoid noise and trasients. While these often work, I don't think they are cost effectove in most cases. I can buy a lot of filtering and isolation equipment for my computers and PLCs for the cost premium of a IEEE-519 drive!

I have found the best way is to buy good quality inverters designed to work with PLC's I have used Movimot on SEW motors with built in ASI without any problems. Since the ASi is built in these must be designed to work with PLC's and they do. Expensive but trouble free.

Tom Jenkins, you are sure right about the 12, 18, and even 36 (mercy!) pulse "snake oil" drives being foisted on the buying public! Robicon is especially notorious for this hoax.

Good grief, instead of an expensive, complex, less efficient, and somewhat less reliable drive, you can buy a standard 6 pulse drive and a simple MTE matrix filter and get better results in every respect.

However, more complex input rectifiers or matrix filters only address one of the three sources of RFI and EMI mentioned in my earlier post--that of power supply distortion due to non-linear use of AC power. RAdiated motor lead noise and common-mode noise are separate issues and need separate solutions.

Of course, as I mentioned before, it does no good to solve a problem you don't have! Better get an expert to analyse specifically why you are having trouble and fix only that.

Incidently, IEEE 519 is a spec that includes the whole power supply network, not just the drive. It follows then that, any supplier claiming their drive is IEEE 519 compliant without defining the network is actually supplying, well, just more snake oil. Sorry!

Pity, but there sure is a lot of that stuff around!