Square D reduced voltage starter problem - stutters

Status
Not open for further replies.
G

Garmark

Member
Location
Garland, TX USA
We have eight of these in the plant running vertical turbine water pumps and centrifugal blowers - Square D Class 8606 reduced voltage starters. These have been installed for about 25 years.

We have one vertical turbine pump that has trouble starting. With the load disconnected, the starter sequences to the RUN condition perfectly, time after time. Connect the pump and it usually starts stuttering about 2 seconds into a 6 second start sequence (timer relay shifts to full-voltage after 6 seconds). Same problem in HAND or AUTO.

We have replaced the contacts in 1S and 2S, the ON-delay timer, the control relay (Type XO) that activates the timer, ALL auxillary contacts on ALL contactors. We replaced the 1S and 2S coils with coils from another pump starter of the same type that started perfectly every time (and still does with the swapped coils). We have monitored coil voltages during start-up sequence on every coil in the cabinet. We have compared every voltage and resistance, including autotransformer resistances, with the pump starter that works perfectly. Everything lines up. 65% voltage during start-up is 318V, right where it should be for 490V line. We have run the 120V part of the system from a wall outlet... no change.

The motor is just back from a complete refurb - new heaters and Klixons, inspected and tested, etc. but not rewound. When the motor does start and run, it sounds normal and amps are what they should be.

After we replaced contacts in 1S and 2S, the starter was much quieter, less hum. After about six perfect starts, it started acting up again and is noisy again during the start sequence before RUN contactor pulls in. The other starter is smooth and quiet. We have opened the RUN contactor and inspected it. Nothing to see there. During the stuttering, we get small arc flashes out of the bottom of 1S (the two pole contactor).

So, according to everything we've looked at, there is nothing wrong! What are we missing???

Looking for some help! I have not found any troubleshooting info on the web.

Thanks,
Mark Elam
Rowlett Creek Wastewater Treatment Plant
Garland, TX
 
A Square D Class 8606 is an RVAT (Reduced Voltage Auto Transformer) type starter. Sq. D is a mfr that puts a thermal cutout switch (Klixon) embedded into the transformer windings. If that Klixon is defective and opening too soon, it will drop out, then close again, then drop out repeatedly, aka "stutter". Could also be a bad connection to that Klixon, it could also be a defective transformer and is actually heating up that fast, but that would typically reveal itself in the voltage measurements, which you say are correct. I would start with looking at that Klixon by finding the wires and jumping across them to see if it stays on. If so, it might mean changing the transformer, because in some designs they buried the temperature switch below the windings.

If that's what it is, Klixon are very simplistic devices and can "wear out" by being repeatedly cycled. The simplest of them are just a concave disc that inverts when heated, so after repeated cycles it retains less and less of its original shape, meaning it takes less and less heat to "trip" it. The point of this is that there could be something in the way you are using this starter that caused the Klixon to repeatedly trip and eventually deform. Something to look at is the duty cycle. RVATs have very very low duty cycles, something like no more than 4 starts per hour, because of the transformer heating up. A bad control device that keeps starting and stopping this starter could have cause the Klixon to keep cycling to shut this down in order to protect the transformer, but it took a toll on the Klixon.
 
Garmark said:
We have eight of these in the plant running vertical turbine water pumps and centrifugal blowers - Square D Class 8606 reduced voltage starters. These have been installed for about 25 years.

We have one vertical turbine pump that has trouble starting. With the load disconnected, the starter sequences to the RUN condition perfectly, time after time. Connect the pump and it usually starts stuttering about 2 seconds into a 6 second start sequence (timer relay shifts to full-voltage after 6 seconds). Same problem in HAND or AUTO.

We have replaced the contacts in 1S and 2S, the ON-delay timer, the control relay (Type XO) that activates the timer, ALL auxillary contacts on ALL contactors. We replaced the 1S and 2S coils with coils from another pump starter of the same type that started perfectly every time (and still does with the swapped coils). We have monitored coil voltages during start-up sequence on every coil in the cabinet. We have compared every voltage and resistance, including autotransformer resistances, with the pump starter that works perfectly. Everything lines up. 65% voltage during start-up is 318V, right where it should be for 490V line. We have run the 120V part of the system from a wall outlet... no change.

The motor is just back from a complete refurb - new heaters and Klixons, inspected and tested, etc. but not rewound. When the motor does start and run, it sounds normal and amps are what they should be.

After we replaced contacts in 1S and 2S, the starter was much quieter, less hum. After about six perfect starts, it started acting up again and is noisy again during the start sequence before RUN contactor pulls in. The other starter is smooth and quiet. We have opened the RUN contactor and inspected it. Nothing to see there. During the stuttering, we get small arc flashes out of the bottom of 1S (the two pole contactor).

So, according to everything we've looked at, there is nothing wrong! What are we missing???

Looking for some help! I have not found any troubleshooting info on the web.

Thanks,
Mark Elam
Rowlett Creek Wastewater Treatment Plant
Garland, TX
Click to expand...

You've changed the contacts you say. Have you tried changing the actual contactors?
 
Jraef said:
A Square D Class 8606 is an RVAT (Reduced Voltage Auto Transformer) type starter. Sq. D is a mfr that puts a thermal cutout switch (Klixon) embedded into the transformer windings. If that Klixon is defective and opening too soon, it will drop out, then close again, then drop out repeatedly, aka "stutter". Could also be a bad connection to that Klixon, it could also be a defective transformer and is actually heating up that fast, but that would typically reveal itself in the voltage measurements, which you say are correct. I would start with looking at that Klixon by finding the wires and jumping across them to see if it stays on. If so, it might mean changing the transformer, because in some designs they buried the temperature switch below the windings.

If that's what it is, Klixon are very simplistic devices and can "wear out" by being repeatedly cycled. The simplest of them are just a concave disc that inverts when heated, so after repeated cycles it retains less and less of its original shape, meaning it takes less and less heat to "trip" it. The point of this is that there could be something in the way you are using this starter that caused the Klixon to repeatedly trip and eventually deform. Something to look at is the duty cycle. RVATs have very very low duty cycles, something like no more than 4 starts per hour, because of the transformer heating up. A bad control device that keeps starting and stopping this starter could have cause the Klixon to keep cycling to shut this down in order to protect the transformer, but it took a toll on the Klixon.
Click to expand...

I'm thinking this also.
The lack of stuttering with no motor load could mean that the magnetic field from the motor current in the AT affects the internal klixon.
 
retirede said:
I'm thinking this also.
The lack of stuttering with no motor load could mean that the magnetic field from the motor current in the AT affects the internal klixon.
Click to expand...

Much more likely, IMHO, that the acceleration is much greater with no pump load on the motor, and so the current decreases to the point that it does not trip the klixon or other cutout during the starting process.

One way of getting effectively no load from a centrifugal or axial pump is simply to close a valve in the fluid path while the motor starts.
 
FIXED!

FIXED!

It was the phase monitor!

We monitored various voltages during the start sequence with a digital meter. Very slow responding display compared to the immediate response of a needle on an analog meter, so... we used an analog meter (Simpson 260). We started at the source of our 480V control circuit and kept moving the sampling point until we observed a rhythmic dip that coincided with the start-up spasms. Going into the Sq D XO control relay, the 480 was stable but coming out, it would dip and return in rhythm with the starter spasms. That XO relay has a 120V coil, so we checked there and observed the same movements of our meter needle. The 120V was repeatedly dropping out and coming right back. It happened too fast to be observable on the digital meter but the needle gave it away!

We bypassed the phase monitor and VOILA! Success! Replaced the phase monitor and have had zero problems since.

Hindsight - This starter worked perfectly with no load. Only mis-behaved when the motor was connected. So... what component of the starter is most sensitive to voltage dips? Uh... the phase monitor! Can't believe I didn't check that sooner. We also fixed a similar persistent problem in another such starter by replacing the phase monitor. This has been a good week!

By the way, no klixons in the autotransformer windings, but thanks for the idea. I'll file it for future use. Also, this problem occurred from dead cold to almost-too-hot-to-touch. So, even if we had klixons, it was not a winding temp issue.

Thanks for responses. Hope this helps somebody!

Mark
 
Thanks!

Thanks!

Besoeker said:
You've changed the contacts you say. Have you tried changing the actual contactors?
Click to expand...

Yes, we changed the contacts AND the coils with known good coils. Didn't help.

We DID take the RUN contactor out of the cabinet and take it apart to remove a zip tie head that we saw down inside the unit. It wasn't interfering with anything, but we thought we should get it out of there. But, of course, our problem involved the reduced-voltage part of the sequence, so the RUN contactor was never a suspect.

Thanks for the response!
 
Garmark said:
Yes, we changed the contacts AND the coils with known good coils. Didn't help.

We DID take the RUN contactor out of the cabinet and take it apart to remove a zip tie head that we saw down inside the unit. It wasn't interfering with anything, but we thought we should get it out of there. But, of course, our problem involved the reduced-voltage part of the sequence, so the RUN contactor was never a suspect.

Thanks for the response!
Click to expand...
Welcome! Good that you fixed it..........:thumbsup:

As a matter of interest, what rating is it?
 
Status
Not open for further replies.
Top