[gcwwtpbuilder]

Sporadic VFD overcurrent faults


Application: “Grit Separator”: VFD running a 1.5hp motor which turns a gear drive. The gear drive rotates a torque tube with paddles that rotates in a conical shaped pit in the headworks of a wastewater treatment plant to separate grit from the influent flow passing through the chamber and keep the grit fluidized. The Grit Separator rotates continuously, however every two hours a grit pump pumps the collected fluidized grit to a grit classifier which separates the liquid and grit, then dumps the grit onto a conveyor belt. The grit pump and classifier run cycle last for 20 minutes. The system collects about one 5gal bucket worth of grit every two hours.

Loading of the motor due to the collected grit is minimal. (5 gal worth every two hours)

The VFD overcurrent shutdowns are random (have occurred both day or night) and sporadic (3-8 days between).

Feeders between the VFD and Motor are 12 awg, 3 phase with ground, approx. 65’ run, local disconnect at the motor with interlock to prevent from VFD running if disconnect is opened.
The feeders, motor and disconnect have been meggered and check out.

The electrician ran the VFD feeds in the same conduit and the grit pump and grit classifier. (Three, 3 phase feeders in the same conduit).

The motor typically runs at 1.2-1.3 amps and the VFD is set to trip at 2.3 amps. The VFD runs continuous 24/7 at 17.5Hz, so there are no accelerations or decelerations. The amperage remains steady while running (1.2-1.3 amps observed, see below). The motor also has overload protection, however the thermal switch has never tripped out.

The equipment vendor has replaced the VFD and motor, as opposed to legitimately troubleshooting it, and we are still getting the same overcurrent fault.

There is an existing system adjacent to ours;. same manufacture and set up, which has never experienced a VFD overcurrent fault.

The VFD is housed inside a control panel, so unless you have the control panel door open you cannot observe the VFD. I have set up my office inside the control room this past week, with the control panel door open and set the VFD display to show amperage and have been observing the system trying catch any abnormalities. So far I have not been able to catch it fault while the door has been open. The amperage remains steady even when the grit pump and classifier cycle their power on and off, as well as other equipment being fed out of the same MCC. I thought I might see instances where maybe the amperage would rise up on occasion, even if it did not reach the 2.3amp VFD overcurrent fault set point. Instead it has been steady as can be, so I am not seeing a condition developing even though it might not have been enough to trip the VFD. I have to believe condition must be limited to the actual trips and must occur quickly. With the control panel door closed (not being able to observe the amperage) I have seen the trip occur and was able to cycle power to the VFD, clearing the fault and restart the system back up quickly, expecting that it may again trip out, but it did not and continued to run for the next three day before tripping again.


The electrician insists even though he ran all three feeds in the same conduit this is not causing any issues because we are not observing any amperage spikes when the system cycles. I am not sure if combination of interaction between the three feeds and whether harmonics from the VFD feed and/or EMI might produce conditions for a perfect storm to cause a random current spike every 3-8 days or not.

I just not sure where to go from here. Any thoughts?

 

[Big John]

...The equipment vendor has replaced the VFD and motor, as opposed to legitimately troubleshooting it, and we are still getting the same overcurrent fault....

So you have good reason to believe the cables are sound, the motor doesn't have a fault, the VFD doesn't have an internal failure.

My next step would be to assume a mechanical bind (which honestly would've been a step I'd chosen before replacing VFD and motor). Sounds like there are a lot of potential points of failure between the motor shaft, the gear-train, and the paddle assembly. Has anyone looked at all that for possible wear?

 

[ponyboy]

One of these VFD "random tripping" threads should be made sticky.

 

[gcwwtpbuilder]

Sorry,

Yes, the vendor did check the system mechanically. When I commented on their troubleshooting, it wasn't clear from the way I wrote it but I was referring to their trouble shooting on the electrical end.

The unit that is faulting actually runs very smooth and steady.

 

[JRaef]

Most VFDs have at least two "Over Current" faults; Over Current as in the VFD's rating, and Over Current regarding the motor over load current (although most will say Over Load instead of Over Current). Some have as many as 4 or 5 different versions of "OC" for different conditions. Is it a text display or is it a code? Either way, post the make of the VFD and the indication it gave.

Here's why. If for example you are experiencing a periodic phase loss, the VFD may not be rated for single phase input. VFDs take in AC, convert it to DC, then re-create the AC going to the motor. Because of that, a VFD does not know, or really care, if the incoming voltage was on all 3 phases or only two. The output is still always 3 phase, because it is manufactured 3 phase anyway. However the INPUT current will increase if you drop one phase, because ALL of the power has to come from the 2 remaining phases instead of 3, so the current increases by the Sq. Rt. of three, 1.732X. If the VFD rectifier is not designed to handle that higher input current, the VFD will trip to protect itself even though the MOTOR current remains normal.

On some VFDs, another issue might be that the capacitors are failing or having a problem, which means there is too much ripple on the DC bus portion. Expensive drives tell you something like "High DC Bus Current Ripple" on a nice big LCD display, cheap little drives with 4 digit 7 segment LED displays can't say all that, so they say something like "OC-3" and leave it up to you to decypher what that means.

So knowing the EXACT fault display and make and model of the VFD will help in trying to understand it.

Final word: running multiple OUTPUT cables in the same conduit (if that's what you meant) is just asking for trouble, all kinds of trouble. Induced currents and capacitive coupling between conductors can cause all kinds of squirrely things to happen in the VFD output section, the OC trips may just be the only thing the VFD can do about it. For example it might be that there is a high speed "spike" in current, called a dI/dt event (d = delta or "change", I = Current, t = time, so dI/dt = change in current over change in time, meaning too rapid of a change). If that wiring mistake cannot change, I would at the very least install a load reactor after each VFD. The reactor slows down the rate of change in current. The problem wont really go away, but the EFFECT it has on the VFD might.

 

[gcwwtpbuilder]

The VFD is a Baldor VS1ST42-OF (rated for 460Vac, 3PH, 2Hp, with an IP20 enclosure and optional EMC filter.

The fault code that comes up is O-1. This is the only fault code that has shown up ever. This is the second VFD (the vendor replaced the original that was faulting on the same code).

There is a note in the O & M regarding the EMC filter that says EMC filters inherently have a high leakage current and that removing the EMC screw on the side of the VFD can reduce trips caused by this condition, however that condition or corrective action is not described in the troubleshooting section of the O & M.

On the cabling, yes, the Output cables are all in one conduit. There is an empty spare conduit available to run the VFD Output cable separately if necessary. I am going to have another discussion with the electrician about this.

 

[JRaef]

Forget the EMC filter thing, irrelevant here.

Looking at the manual, they indicate that the O-I fault code is for output (load) current in excess of the VFD rating, but also it can be due to heat sink over temperature or a cable fault as well. Given the sporadic nature of this, I'd lean toward heat sink over temperature first because it may be that it only happens under a set of "perfect storm" conditions inside of your control panel with regard to other loads running at the same time. So when you open the door to try to watch it happen on the display, it never happens because you are letting the heat out. Can you put a temperature probe inside and feed the lead for it through the door gasket?

 

[Big John]

...So when you open the door to try to watch it happen on the display, it never happens because you are letting the heat out....

Seen that type of fault more than once on electronic controls.

How common is it for a drive to display a temperature fault as an overcurrent? Are they actually looking at temperature rise as a measure of current flow?

 

[gcwwtpbuilder]

I appreciate your input gentlemen.:thumbsup:

It stays relatively cool inside the control room so I assumed temp was not an issue but I should run a check on the temp inside the panel with the door closed.

I had never really thought about how the VFD determines current flow. Since they do not use CTs, temperature is probably very likely means.

Right now my plan is to:

1) verify temp inside control panel and either confirm temp is the issue or strike it from the list of possiblaties.
2) re-visit the VFD Output cable routing with the electician.
3) look into possiblily of installing a line reactor.

 

[gcwwtpbuilder]

3) look into the possibly installing a "load" reactor.

 

[Nukie Poo]

Also, try relocating VFD as close to the motor as possible.

 

[JRaef]

I appreciate your input gentlemen.:thumbsup:

It stays relatively cool inside the control room so I assumed temp was not an issue but I should run a check on the temp inside the panel with the door closed.

If the room is cool, but the box is sealed, it may not be cool enough to get enough radiant heat dissipation through the enclosure walls. Seals keep out dust and moisture, but keep in heat.

I had never really thought about how the VFD determines current flow. Since they do not use CTs, temperature is probably very likely means.

Hall effect transducers. In cheap VFDs, it is one hall effect on the DC bus, then the motor current is calculated based on that. On more expensive drives, it will be separate hall effect transducers on each of the output transistor circuits.

Right now my plan is to:

1) verify temp inside control panel and either confirm temp is the issue or strike it from the list of possiblaties.
2) re-visit the VFD Output cable routing with the electician.
3) look into possiblily of installing a line reactor.

Good plan. A line reactor may not help with this issue, but they are cheap insurance.

 

[lefleuron]

I had a baldor once that did the exact same kind of random tripping, this was a fresh air intake fan.

When the wind blew in just the right direction, it would trip during ramp on OC.

I ended up extending the ramp time by 15 seconds on that particular drive and it never happened again.

 

[oliquir]

change the pwm frequency of the vfd (lower it a little) , it may do nothing but it can also solve the problem
sometimes electrical noise give all kind of trouble
also check the overload time setting. give it a 1-2 seconds delay before it gives error. it may be set to zero so anything that gives a very short overload may trip the vfd

 

[gcwwtpbuilder]

Right now the VFD is set to run at a constant speed which is 17.5Hz. You are saying i should adjust it to say 17.0Hz?

On the overload time delay. Unfortunately, the Baldor VFD does not have one. If it sees the current hit the overcurrent set point it goes to fault.

 

[oliquir]

no the pwm frequency is inside the paramters (not the output frequency) it is set between 1 to 15 khz depending on vfd model

 

[gcwwtpbuilder]

10-4.

I will take a look in the O&M and figure out which parameter pmw frequency is and try a small adjustment.

 

[Big John]

Be careful adjusting your carrier frequency. Increasing it will increase you switching losses and actually make a drive heating problem worse (if that's what's causing the problem).

 

[gcwwtpbuilder]

Thanks for the heads up. I will make sure it is a small adjustment down.

 

[JRaef]

10-4.

I will take a look in the O&M and figure out which parameter pmw frequency is and try a small adjustment.

You can read it in P00-14, you can adjust it in P-21.

Looks like the factory default is 16kHz, that's actually pretty high. I would turn that down as low as you can live with, 1khz is the best for the motor and drive, but it makes the motor whine fairly loud. I'm actually surprised that the factory default is so high, they wanted to make people think their drive was quieter than others maybe? It's not a good idea though, and it may be a contributing factor for sure.Typical factory defaults for that are in the 2.4-6khz range.