×
INTELLIGENT WORK FORUMS
FOR ENGINEERING PROFESSIONALS

Log In

Come Join Us!

Are you an
Engineering professional?
Join Eng-Tips Forums!
  • Talk With Other Members
  • Be Notified Of Responses
    To Your Posts
  • Keyword Search
  • One-Click Access To Your
    Favorite Forums
  • Automated Signatures
    On Your Posts
  • Best Of All, It's Free!
  • Students Click Here

*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.

Students Click Here

Jobs

questions about recorded motor starting current waveforms
3

questions about recorded motor starting current waveforms

questions about recorded motor starting current waveforms

(OP)
Below is some background provided for info, and at the end some very targeted questions related to interpretation of attached current traces.  You can skip the background info if you'd like.  I would really like to focus on the 3 questions.

=============BACKGROUND:=============
This is a 60hp 460 vac motor.  KVA code G.  Started with combo starter including HMCP 100R3 instantaneous only molded case breaker with setting G (nominal 900A)

We have had two trips of the instantaneous breaker among multiple starts as indicated below (trips occurred 5/13/08 09:16 and 8/30/2008    20:19 where column "Success?" has an entry of "No"

Date        Time    Success?    Highest Peak
5/13/2008    09:16    No    Not recorded
5/13/2008    18:00    Yes    1111
5/15/2008    16:17    Yes    1049.5
5/15//2008    16:35    Yes    1013.5
5/15/2008    16:52    Yes    937.88
5/15/2008    17:10    Yes    1026.6
5/15/2008    17:28    Yes    602.25
5/15/2008    20:02    Yes    Not recorded
6/12/2008    11:49    Yes    Not recorded
6/12/2008    13:22    Yes    Not recorded
7/7/2008    23:06    Yes    Not recorded
7/7/2008    23:23    Yes    Not recorded
7/8/2008    17:11    Yes    Not recorded
8/7/2008    9:04    Yes    Not recorded
8/30/2008    20:19    No    Not recorded
9/1/2008    16:36    Yes    732.67
9/1/2008    16:51    Yes    998.75
9/2/2008    10:46    Yes    991.9
9/4/2008    9:18    Yes    1032
9/4/2008    9:36    Yes    959.75
9/4/2008    9:55    Yes    1030.7
9/4/2008    10:12    Yes    1056.5
9/4/2008    10:30    Yes    1038.4
9/4/2008    10:58    Yes    1053
9/4/2008    11:28    Yes    1043.5
9/4/2008    11:59    Yes    998.5
9/4/2008    14:09    Yes    1027.9
9/4/2008    14:39    Yes    1039

The peak recorded is the highest peak among any of the three phases.  

We did replace both the breaker and the starter after the first trip... (2nd trip occurred with a different breaker and starter.).   Both breakers were tested many times (pulse method) and never tripped under 900A.  Considering that this corresponds to 1272 peak and the highest recorded current peak was 1111A, the reason for the trips is not obvious.  (Yes, I know we can go to setpoint 1000A = code H but the questions are focused towards understanding these previous trips).

Review of above peak values does show one thing of minor interest - the peaks on 5/15/2008 17:28 and 9/1/2008    16:36 seem to be outliers in the low direction.

=======  DISCUSSION OF ATTACHED CURRENT TRACES ==========
The motor that tripped will be label U2 and an "identical" sister motr labeled U1.

There are three runs displayed in the powerpoint:
U2 run #1 9/1/2008    16:36     on slide 1
U2 run #2 9/1/2008    16:51       on slide 2-4
U1 1 run on 9/1/2008    (sister)  on  slides 5-7
Slides 8-10 contain various comparisons btween the second item item (U2 run#2) and the 3rd item (U1 = sister)

General Observations from all runs:
GO1 -  phase to phase voltages appear close to expected peak value ~ 460*sqrt2) = 650vac.

GO2 - The decayign DC components don't seem to decay directly to 0.  They seem to overshoot and oscillate for longer than expected.

GO3 - When the LRC reaches quasi-steady state, the current magnitude in A phase is only ~ 450A  PEAK.   This is a surprise since rms LRC should be around 450 rms.  The other 2 phases (B and C) have peak up around expected 600 - 650A peak.   Also surprising that similar behavior shows up on the sister motor (will check in to test instrument error, but for the time being I'm assuming this is real).

Observations from Slide 1:

Slide 1 is run #1 of U2.   This happened to be a "bump" test to check rotation after motor was determinated for troubleshooting, so the duration was 0.7 seconds.   This is one of the two total tests hat showed extremely low peak.

S1O1 - When the contactor opens, the phase-to-phase voltages recorded on the load side of the contactor seem to transition fairly smoothly to a lower level.

S1O2 - I'll bet the worst case angle for dc offset would have occurred on phase A.  This explains why the recorded peak was so much lower for this start than the others (if we can understand GO3 above).

============   QUESTIONS: ==============
Queston 1 (Observation GO2) - What do you think is the explanation for oscillating dc component (seems to overshoot instead of decaying to 0)?  It occurred on both motors.  I have seen it on other tests before.  I have a theory about it, but I'd like to hear yours. (by the way, these waveforms were recorded with AC clamp-on probes, not hall effect transducers)

QUESTION 2 (Observation GO3)- What do you think about the current so much lower on A phase than the other phases B and C? (assuming it is real).    I have often seen that the unbalance in current can be very high at no-load and disappears as load increases.   I wonder if for some reason the starting condition might act like the no-load condition in this respect.  Or does this indicate some anomaly in ths system.

QUESTION 3 (Observation S1O1) - What do you think about the voltage behavior in slide 1 where voltage on the line side of the contactor seems to smoothly decrease after contactor opening. Is this the residual voltage generated by the motor, or could it indicate some malfunction of the contactor?

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

RE: questions about recorded motor starting current waveforms

(OP)

Quote:

where voltage on the line side of the contactor
Sorry. The only voltage measured was on the load side of the contactor (not line side, not at motor).

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

RE: questions about recorded motor starting current waveforms

(OP)
No response?  Disappointing.  Has anyone seen behavior similar to this?  i.e. the "oscillating" dc component in question 1,  the apparent unbalance in question 2, measurement of load side voltage after disconnection as mentioned in question 3?

Haven't measured starting waveformst? Haven't seen this behavior? Or don't care?  (maybe the question is so long-winded that people don't pay attention?).

As I mentioned, I have a theory for question 1 that I am fairly satisfied with.  It starts with a discussion of the characteristics of ac clamp-on probe as discussed here:
thread238-217182: ability of clamp-on probe to see exponentially-decaying dc current

Since we call it an ac probe, we should immediately begin to wonder how it will respond to something that is not pure ac... i.e. a waveform that starts suddenly and includes an exponentially decaying transient dc component.

Analsysis of a simple ct circuit with simplifying assumption of purely resistive burden tells us that the step response of a CT is a step followed by an exponential decay to 0 with time constant (call it Tau_meas).  And ASSUMING we are within a range where the circuit acts as LTI (linear time invariant), the step response completely characterizes the system and can be used to determine response to any input.

We have measured the response to a step increase in input dc current for  a clamp-on probe simila to the one used for the motor starting waveform  (except smaller).   Results match the above predicted ideal response very well (step increase followed by dc decay).

Regardless of whether the exact model assumptions are  correct for the clamp-on, as long as the step response is correct, I can use that model to predict response to any input (again as long as remains within the range where the system acts as LTI).

The model is given in more detail in the attachment.

With model in hand,  I can draw two or three conclusions about how the clamp-on will represent the motor starting waveform (under model assumptions):
1 – As long as Tau_meas >> Tau_motor,  the indicated current is very close to actual.
(Tau_meas is L/R of measurement circuit and Tau_mot is L/R of motor/supply circuit).
2 – As Tau_meas get down near Tau_motor, we start to see errors.
3 – One of the errors we see is an apparent oscillating dc component of the indicated waveform (even though nothing of the sort existed in the actual waveform).   This seems to explain the apparent oscillating dc component that we saw in our waveofrms.

Near the end of the attached pdf, several plots of compueted Actual and Indicated motor starting current waveforms are provided which I believe support these conclusions.

What do you think?
 

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

RE: questions about recorded motor starting current waveforms

Quote:

QUESTION 2 (Observation GO3)- What do you think about the current so much lower on A phase than the other phases B and C? (assuming it is real).    I have often seen that the unbalance in current can be very high at no-load and disappears as load increases.   I wonder if for some reason the starting condition might act like the no-load condition in this respect.  Or does this indicate some anomaly in ths system.
What is interesting is Slide 7.  The three phases do not add to zero instantaneously.  If the motor is connected delta or ungrounded wye, then something is wrong.  There is no place for zero-sequence current to flow.

Quote:

QUESTION 3 (Observation S1O1) - What do you think about the voltage behavior in slide 1 where voltage on the line side of the contactor seems to smoothly decrease after contactor opening. Is this the residual voltage generated by the motor, or could it indicate some malfunction of the contactor?
I'd guess residual voltage generated by the motor.  I can't see a smooth decrease resulting from any contactor malfunction.  Restrikes would cause chopping and high frequency components.

RE: questions about recorded motor starting current waveforms

(OP)
Sorry, an error at the end of my file:

Quote:

# Higher sinusoidal frequencies demand lower tau (faster decay) for accurate amplitude

Should have been:

Quote:

# Lower sinusoidal frequencies demand higher tau (slower decay) for accurate amplitude

I had the mindset that fast decay was fast response and accurate reponse.  That was bass-ackwards.

The ideal response is a very high Tau representing slow decay (step response more closely resembles a step).  

The same non-idealities which prevent accurately represinting the quasi-dc exponentially decaying component also prevent accurately representing frequencies below 30hz.  Stated frequency range of the probe is 30hz - 1000hz.

Conclusion is the same, I just wanted to correct those comments that I had typed into the end of the attachment which were out in left field.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

RE: questions about recorded motor starting current waveforms

(OP)
Thanks jghrist - great comments.

About question 3 - Great points. Smooth transition doesn't seem like contactor malfunction but I guess (?) is what we'd expect for motor residual voltage.

About question 2 - I hadn't noticed that anomaly on slide 7.  All 3 currents sum to 0.  That was the only trace we took on this particular motor.  I'm guessing that the CT on that most offset phase was flipped and actual waveform should be negated from what is shown.

It also brings to mind a very useful concept that all the currents should sum to 0 which I just haven't been thinking about lately.   I think I can get access to the numerical data and dump it into excel.  Then I will try summing the three currents and see if that waveform showing the very low A phase current makes sense or not.  That may provide some clues if there is an indication problem or not.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

RE: questions about recorded motor starting current waveforms

(OP)

Quote (ELECTRICPETE):

I'm guessing that the CT on that most offset phase was flipped...
Duh.  I meant to say:

Quote (ELECTRICPETE):

I'm guessing that the clamp-on on that most offset phase was flipped...

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

RE: questions about recorded motor starting current waveforms

ElectricPete.
I would like ask some stuipid Q's.
We had such trips on the MV motors with digital relays.
1. You haven't recorder in trip times, maybe it was more than 1272A?
2. Why you decide that problems in the breaker or starter?
3. Are inst. module of molded breaker is some electronic/digital part ( sorry, Im not familiar with this molded breaker).

What we found in our case.
BTW, it was inst. setting too. Time of trip about 40ms, but all digital device include retradation time. that mean minimum time for hold the trip operation.
I try explained on the other example: Time to trip is 100ms
peak current is decyed after 98ms. Are relay trip or not?
Retradation time is exactly -t range of this parametr.
in this relay it was about 30ms. That mean  before 70ms relay is resets, after will trips, after 100ms. In case of inst. setting was enough peak of current about 10ms.
Small change of setting solve our problem.
I recommens to you check from this point too.
Best Regards.
Slava

RE: questions about recorded motor starting current waveforms

Hi electricpete
It is a very interesting phenomenon, indeed. My opinion is as follows:
1) The starting phenomenon may be divided into 3 parts.
The 1st is the subtransient [0-0.3 sec] where there is an oscillating component.
The oscillating component may be produced by some parasitic capacity included in the circuit with R and L.
Also it is possible a slide uncontrolled imbalance of these capacity which may influence the current unbalance.
The 2nd part [0.3-.07 sec] it is transient phenomenon where the current decays with the slip as expected.
The 3rd part is the rated slip and rated current at 0.7 sec [sharp!].
2) The remaining voltage after breaker trip is due to the residual magnetic flux.
Best Regards
 

RE: questions about recorded motor starting current waveforms

(OP)
Thanks slavag and 7anoter4 for your responses.

7anoter4 - Great point.  While I had focused on the first few cycles, the extent of these oscillations shows up better in slide 9 (comparison of both motors phase C current) where we see a distinct transition.  Looks to me like the oscillation abruptly ends around 0.2 seconds.   My model with measurement should terminate oscillations much more quickly I think (in fact I think it can only create one oscillation up and one down... not repeated up and down).

And I do agree with you 0.7 seconds is the transition to normal running conditions.

Now you mentioned two reasons for the oscillations.
1 - might be a consequence of unbalance.  I hadn't thought of that before.  Interesting.

2 - Parasitic capacitance.  I can see two possibilities:

2A - If you are talking about parasitic capacitance in parallel between cable and ground, I think the curent is at most a few 100 milliamps (what we would see during ac hi-pot), correct?  In that case, it doesn't seem relevant.  

2B - The other place to model capacitance would be as an element in series with the leakage reactance in the equivalent circuit.  Could represent turn to turn capacitance.  At single frequency, we can absorb the capacitance into the inductance and the circuit acts inductively.  But under transient, we cannot make this simplification. Hmmm.  How would one get an estimate of the order of magnitude of capacitance which might be expected?

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

RE: questions about recorded motor starting current waveforms

Hi Pete

My comments are:

1. I believe that the DC oscillation is a function of the probe characteristics. If your probe was "broad band" including DC, you would see the expected decay.
2. I have observed reduced starting current on one phase when there is a problem on the supply to that phase.
When you are running a motor virtually open shaft, there can be an imbalance in current due to a small offset in the supply phase angle on one phase and an imbalance in the magnetizing circuit in the motor on one phase. When these coincide, the imbalance is exaggerated. Rolling the phases reduces this imbalance.
I have not observed this happening during start as the magnetizing current is very small relative to the start current. I have observed an imbalance in the starting current due to one high impedance phase.
I would look hard at the probe and calibration.
3. This looks like regenerated voltage. I expect that if you could see more, you would see the frequency dropping and voltage dropping with time.
Some motors can regenerate for quite a period of time.

Best regards,
 

Mark Empson
L M Photonics Ltd

RE: questions about recorded motor starting current waveforms

(OP)

Quote:

QUESTION 2 (Observation GO3)- What do you think about the current so much lower on A phase than the other phases B and C? (assuming it is real).
Phase A is about 150A lower than the other two phases.  I summed the three waveforms together numerically and they do not sum to 0.  When phase A is at peak, the sum is about 150A in the opposite direction.  As jghrist has mentioned they should of course sum to 0.  The problem appears on a few waveforms and not on others.  No other anomalies noticeable when phase A is low.  

I think it must have been a measurement anomaly.  Perhaps the clamp-on was not firmly clamped closed.  

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

RE: questions about recorded motor starting current waveforms

(OP)
By the way, thanks for the comments Mark and others.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

Red Flag This Post

Please let us know here why this post is inappropriate. Reasons such as off-topic, duplicates, flames, illegal, vulgar, or students posting their homework.

Red Flag Submitted

Thank you for helping keep Eng-Tips Forums free from inappropriate posts.
The Eng-Tips staff will check this out and take appropriate action.

Reply To This Thread

Posting in the Eng-Tips forums is a member-only feature.

Click Here to join Eng-Tips and talk with other members!


Resources