Determining Unknown CT polarity

If you have access to a 2 channel oscilloscope and refer to the mains - be careful to make a resistive attenuator that you are in range of the scope- trigger on the same alternance and compare with the others. This can be done online but beware the high voltage!!! Do not disconnect the CT's.

This begs the question, is this a new installation? If not, has it worked in the past. If it has worked before, something has been changed or failed. CT's don't change polarity on there own!

The motor did work in the past and the problem ocurred after the motor was removed for repairs. If I could access the CT's without major disassembly, I could physically check the polarities (dots). Thanks.

As Tony says, CT's don't change polarity. If all the main wires run the same way through the CT's, you should be back to the old setup. However, if the CT wires are connected differently, differences will occur. When the motor has new windings, there can have been made a mistake on re-connecting these too. Does the engine perform as before? You could measure the currents with a ampere clamp. A schematic on how the CT's are connected can help us think with you.

I agree with you. The wires weren't reconnected properly. I can't use the battery test because of the insulation on the bus bars and not knowing how the CT primaries were reoriented in the motor. The supervisors don't want to run the motor without protection so I'm stuck to determine the polarities by trial and error (6 x 6) with only 3 starts every 24 hrs. I will send you a schematic as soon as I can. Meanwhile the CT's, all 6 of them are connected for differential protection (one set, 3 phases, looking toward the power input and the other 3 phase set looking at power leaving the motor windings before it gets to the neutral point). Polarity is critical in this case because both sets are tied to the 87 (SEL 710) relay's terminals, and each phase value must equal zero (forward = reverse).

I think you have 2 options:

1.Shut down and check your connections physically, which is the most inconvenient way to do and also shuts the system down for a now not to determine time.

In principle, the dots on CT's make it easy to work.

Read dot for dummy. (sorry) The color code of the wires goes also with it. 2.Work around it: When you know the pairs of wires per CT, you can reverse the wrongly connected ones. To determine which ones, I refer to a oscilloscope - maybe somewhat unusual to work with for electricians -

How to work:

With only one channel scope: make a connection with all 3 wires that you would suppose to be the same phase.

This connection is 3 same value capacitors (e.g.polyester- type less than a microfarad) each connected to one different CT tap.

The other ends to your probe of the oscilloscope. You will see a sinusoid that corresponds with the measurement of each capacitor end. The commons need connected to the other pin of the probe.

When all connected, 3 sinus should appear on the screen, 120 degrees shifted in phase. If one sinus is upside down, that is the problem one. Reverse these wires and you should be good to go. In principle, reversing the wires in the CT should give you the same result: a 180 degrees phase shift.

Work safe: do not disconnect the CT's when the motor runs. As for the other 3 CT's, work the same way in a second setup.

Sorry to keep harping back to this but nothing has been changed on the switchgear, so, forget it!

The motor has been changed, look to that!

For goodness sake it's not rocket science!

TonyS:

Thank you my friend (I guess you didn't read my posts?). I know that nothing has changed on the switchgear, besides I can verify that because those CT's are accessible. All I said was I can't run the motor with a differential indicated because of mgt. And without major disassembly I can't physically check the motor CT's, OK? I'll take care of it from here, thanks again, I only wondered if someone out there might have an idea of how to check CT polarity with just the secondary windings on a deengergized CT (voltage/current injection w/oscilloscope).

A CT is a current transformer when the burden is connected.[p] If you have access to the circuit when disengaged, and you are able to send a wire through all the CT's in groups of 3, you can disconnect the CT's and use it as a voltage transformer.

When you monitor the signal on the wire that you have installed and send e.g a DC voltage pulse through the CT's, one by one, you can monitor the polarity and compare all 3.

When you trigger e.g. on a positive flank, the wrongly phased CT will cause a totally different reaction.

Make sure nobody can start up the engine while testing.

As for the internal motor CT's, you might be able to check on each of them when inducing a pulse on one. Since we get nothing in concreto - like CT characteristics, you'll need to experiment with the scope settings. Sensitivity and time base.

You could also steal some current (when working) with supplementary CT's, that you connect well (check the dot) and monitor those signals. I hope nobody gets ideas to do something else with this.LOL

When monitoring voltage, put a 10k ohms resistor as load of the wire through the CT's.

If you need more info, let us know.

Success

I am not 100% sure of this on the SEL710, but all the SEL's and especially the 7 series relays have the capability to extract waveform and samples of the waveform. Knowing that you do not have an actual motor diff, why not run the motor and manually trigger the built-in SEL waveform recorder or set the trigger criteria on an input or even the current level. Then you can analyze the waveforms and know which signals to swap.

Better yet, if you look at the event report of the (nuisance) diff trip, the report will give you all the current magnitudes and angles at the time of the trip. This way also, you will be able to tell which signals to swap.

Hope this helps. You'll have to read the SEL manual about how to extract the data. Its there.

A21M: Thanks!! You're right, the SEL 710 has an oscillograph output that shows all six currents and their relationships. I was able to get the waveforms prior to the 87 shutdown, and make the appropriate wiring changes. The 2500 hp 4.16 kv York chiller at the University of Texas (Austin) is back in service thanks to you!

hughes838