60 Hz 8.3 and 10 Hz sidebands

Status
Not open for further replies.
G

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
160212-1708 EST

In the recent thread on harmonics I mentioned seeing sideband signals at +/- 8.3 and 10 Hz of the 60 Hz line frequency. The same sidebands show up with the 60 Hz harmonics. This looks like AM modulation of the 60 Hz voltage.

The 10 Hz I could guess might related to a 600 RPM speed for the alternator(s). 3600/600 = 6, 60/6 = 10. I don't have a reasoned source for 8.3 Hz. Do large alternators run as fast as 600 RPM?

A quick internet search did not tell me at what RPM large alternators are being run. Can any of you power company people provide any insight about these sidebands?

.
 
160213-0047 EST

Ingenieur:

Thanks.

I did some more Internet searching and only found a reference to 1800 RPM in one place. Also found that via a clutch some generators are being used as large synchronous motor loads for power factor correction.

Tonight I am not seeing the 10 Hz sidebands.

I am seeing a frequency arount 8.7 Hz. At present I don't have an accurate means to measure this frequency. See attached plot.
2016_02_13-0041-Line-60-Hz-Hamming-1024.jpg

.
Note: that 8.7 Hz or thereabouts is not an integer relation to 60 Hz. So are the two spikes below 60 at about 51.5 and 43 sidebands of 60 and related to the about 8.7 first spike. Note: 60/7 = 8.5714. Is there something in the system that would relate to a factor of 7?

Are these sidebands or other frequencies based on the rotating machinery, or are these some form of resonance in the voltage regulation systems, or the power grid system.

Can anyone make comparable measurements either in my area or other parts of the country? And if so what are your results?

.
 
gar:

I just PM'd you contact info for someone who was asking a few yeas ago about anomalies you described in a recent thread where you are pushing two screwdrivers in the ground. He would love to hear from you. I have his plots somewhere but I don't think I should post someone else's work. I'm sure he'll be happy to share and enjoy some dialogue. Give him a shout when you have time.

What I recall standing out is he would measure anomalies at particular times of day at a given frequency.
 
Since I read this, has anyone found that these anomalies are what is causing random arc fault tripping?
 
mopowr steve said:
Since I read this, has anyone found that these anomalies are what is causing random arc fault tripping?
Click to expand...

It's my less than optimally qualified opinion that the magnitude is insufficient to do much of anything.

But let's ask this question "If you had a frequency generator, could you use it to trip AFCI breakers?" I think that test should happen before you go speculating about cause & effect, to include or exclude the cause as a possibility.
 
160213-0907 EST

mgookin:

Thanks for the reference and information.


mopowr steve:

The Y axis is in db. On a voltage basis db = 20 log(base 10) V1/V2. The largest off frequency spikes (meaning not 60 Hz) are down at least 40 db. Thus, 40/20 = 2, and v1/v2 = 100. These are not more than 1% in voltage, and 0.01% in power. Not likely to affect an AFCI.


New observations this morning:

I found out how to get a numeric value for the frequency values with Pico. Some of the frquencies come and go with time, others are more or less constant.

There is something close to 8.67 Hz and integer multiplies thereof. Note: 8.67 * 7 = 60.7 . Thus, it could be a sub-harmonic of the 60 Hz fundamental. Is there something in an alternator or its mechanical driver to produce a divide by 7?

Also this morning I saw something around 50.71 vs 59.83 or a difference of 9.12 Hz for a short time. I also saw 51.45 at this same time or 8.38 Hz.

For windowing the above were Blackman.

.
 
I recently completed a course on FACTS
basically solid state pf, Z and ph angle manipulation

the switching action of the thyristors etc cause harmonic distortion
filters are incorporated to attenuate the thd
the filters are usually notch type with CL elements
the C >> L so you get some + VAR out of it

we typically only considered the 5th and 7th
the 3rd being trapped/cancelled by a delta iso xfmr and the rest very low
depending on the system line parameters etc sometimes the 11th would need address
again the 9th was not an issue due to the inherent delta filtering

from the text
 
Here's the 5, 11, 13 and 17
I xth harmonic as a % of I 60 Hz at a 20-30 deg overlap (which be adjusted/optimized to minimize tdh)
5th 16%
7th 9%
11th <3%
13th <2%
17th <1%
 
Last edited:
Status
Not open for further replies.
Top