- Location
- Wisconsin
- Occupation
- PE (Retired) - Power Systems
Youngster, trying to show off?I would guess if I told you I once built a radio with a "cat's whisker" you would know what I meant huh?
How big of an antenna did you use?
Flatened top sine wave, does it mean anything?
- Thread starter iwire
- Start date
- Status
How big of an antenna did you use?
- Location
- Massachusetts
- Location
- Massachusetts
OK, I picked up the meter and downloaded the data into a pdf file. I hope this link works.
https://docs.google.com/open?id=0B5wiR6abvbiJYTY1MzUyMWYtN2ZlMi00OTNlLTk4MGYtYjU5MzBiOWQ5ZmQ2
(Thanks Dennis!)
https://docs.google.com/open?id=0B5wiR6abvbiJYTY1MzUyMWYtN2ZlMi00OTNlLTk4MGYtYjU5MzBiOWQ5ZmQ2
(Thanks Dennis!)
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
120121-0826 EST
iwire:
Since I have not used or studied your instrument I have to guess at some aspects.
A quick look seems to indicate mostly normal operation.
The general results are different on Sat, Sun, and Wed.
There seems to be a lot of 5th harmonic.
One day, 1-20-2012 during the 8 AM hour there was a voltage transient on phase B, 156 V. I do not know what this really means because it is quite obviously some sort of average over possibly at least one cycle since it is less than the 120 V sine wave peak. So in actuality this transient could have been 5000 V for a short time. Or it could have been a general increase in the sine wave for one cycle, almost impossible.
Apparently this transient occurred on only one phase. By any chance have the component failures at this facility been only associated with phase B?
,
iwire:
Since I have not used or studied your instrument I have to guess at some aspects.
A quick look seems to indicate mostly normal operation.
The general results are different on Sat, Sun, and Wed.
There seems to be a lot of 5th harmonic.
One day, 1-20-2012 during the 8 AM hour there was a voltage transient on phase B, 156 V. I do not know what this really means because it is quite obviously some sort of average over possibly at least one cycle since it is less than the 120 V sine wave peak. So in actuality this transient could have been 5000 V for a short time. Or it could have been a general increase in the sine wave for one cycle, almost impossible.
Apparently this transient occurred on only one phase. By any chance have the component failures at this facility been only associated with phase B?
,
- Location
- Massachusetts
While I was there I talked to a few employees, I asked about failed ballasts and it turns out that is not really an issue. But they did have a pile of capacitors that they had replaced on HVAC unit motors.
Let me just ask a what if.
Lets say these HVAC units where rated for 240 supply but someone relabeled them 277 and they where supplied at 277.
Could I expect capacitor failures or would you expect another mode of failure?
Let me just ask a what if.
Lets say these HVAC units where rated for 240 supply but someone relabeled them 277 and they where supplied at 277.
Could I expect capacitor failures or would you expect another mode of failure?
How are the capacitors connected?
I don't have much experience of single-phase motors and had assumed that they were connected across the supply.
Gar has stated that they are in series with (motor winding) inductance.
Perhaps either of you could clarify this?
- Location
- Massachusetts
I am sorry I can't answer that.
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
120121-1125 EST
In a single phase motor a capacitor is used in series with a winding to produce a phase shift to one coil of what is essentially a two phase motor while the other coil is connected directly across the line. This provides the necessary two different magnetic phases to produce a sort of rotating magnetic field, and therefore the ability to start the rotor rotating.
In some motors a centrifugally actuated switch disconnects the capacitor after sufficient speed is reached. This would be a start only capacitor. In other motors the capacitor is always connected. Then some are a combination.
Effectively the start capacitor and motor coil form a series resonant circuit, but not tuned to line frequency. Thus, the shift in phase angle of the start magnetic field. Depending upon where you are on the resonance curve will determine to what the capacitor voltage will rise.
Since the ballasts are not failing, then it looks less like a large voltage spike problem.
I would suggest putting a voltmeter across a motor capacitor and see what is the voltage reading. Compare this with the capacitor rating. Look at where the capacitor was made.
.
In a single phase motor a capacitor is used in series with a winding to produce a phase shift to one coil of what is essentially a two phase motor while the other coil is connected directly across the line. This provides the necessary two different magnetic phases to produce a sort of rotating magnetic field, and therefore the ability to start the rotor rotating.
In some motors a centrifugally actuated switch disconnects the capacitor after sufficient speed is reached. This would be a start only capacitor. In other motors the capacitor is always connected. Then some are a combination.
Effectively the start capacitor and motor coil form a series resonant circuit, but not tuned to line frequency. Thus, the shift in phase angle of the start magnetic field. Depending upon where you are on the resonance curve will determine to what the capacitor voltage will rise.
Since the ballasts are not failing, then it looks less like a large voltage spike problem.
I would suggest putting a voltmeter across a motor capacitor and see what is the voltage reading. Compare this with the capacitor rating. Look at where the capacitor was made.
.
OK. Did a bit of digging and I believe that Gar is correct for single phase motors.
Apologies if I caused any confusion.
Capacitors are voltage sensitive as well as temperature sensitive. That said, capacitors used for starting single-phase motors will have voltage rating greater than the supply voltage. I hope the replacement capacitors were properly rated (around 250 V for start-caps and 370 V for the run-caps).
Remember that choosing a larger capacitance than required will result into a higher current in the auxiliary winding and the capacitor, the phase shift will be lesser and the torque affected. (If a start cap is switched and the run-cap is used, the capacitance we are talking about will be the total capacitance with both caps on-line, i.e. sum of the two capacitance). High capacitor currents will heat up that component and will ultimately short the capacitor out, damaging them. Or burn the auxiliary windings.
But, from your recorded data, those HVAC motor capacitors can easily get damaged if the voltage spikes continue,even with properly rated capacitors.
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
120121-1532 EST
Reading topgone's post suggests this question --- for new failures are these of original capacitors, replacement capacitors, or a mix of original and replacement?
If it is a mix, then what is the percentage distribution between original and replacement?
How long does a capacitor last before failure?
How are the failures distributed between the phases?
.
Reading topgone's post suggests this question --- for new failures are these of original capacitors, replacement capacitors, or a mix of original and replacement?
If it is a mix, then what is the percentage distribution between original and replacement?
How long does a capacitor last before failure?
How are the failures distributed between the phases?
.
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
120122-1632 EST
iwire:
The one event in your data of 156 V on phase B on 1-20-2012 needs some further study that may not be possible.
A question I have is what does it mean? What is the measurement? First, I have to assume it is line B to neutral. Then what is it?
From dictionary.com transient is defined as:
adjective
1. not lasting, enduring, or permanent; transitory.
2. lasting only a short time; existing briefly; temporary: transient authority.
3. staying only a short time: the transient guests at a hotel.
4. Philosophy . transeunt.
noun
5. a person or thing that is transient, especially a temporary guest, boarder, laborer, or the like.
6. Mathematics .
a. a function that tends to zero as the independent variable tends to infinity.
b. a solution, especially of a differential equation, having this property.
7. Physics .
a. a nonperiodic signal of short duration.
b. a decaying signal, wave, or oscillation.
8. Electricity . a sudden pulse of voltage or current.
So this word can cover many different types of events.
What does RMS represent? Root Mean Square. Mathematically I can apply an RMS calculation to a sub part of a sine wave cycle. But unlikely to be of much use. More useful is a measurement over an integral number of cycles, including the possibility of just one cycle. If the measurement is over many cycles, then random start and stopping times of the measurement won't make much difference.
The definition of RMS includes an averaging element, that is the mean part of RMS.
If you perform an RMS measurement on a steady-state sine wave, and independent of whether you start the measurement at a zero crossing, or any arbitrary point, then if the measurement is performed over exactly one full cycle the answer will be the same.
A chip, like the Cirrus power monitor, uses a free running, but stable, oscillator at a frequency well above the signal to be measured to control sampling of the waveform being measured. Typical for their chip is 4.096 MHz. This is divided by 1024 resulting in 4000 samples per second. If you use 4000 of these samples to perform and RMS measurement, then 60 integral cycles create the measurement.
We shall assume your instrument is measuring an RMS value of the waveform. Probably from a much higher sampling rate than what Cirrus uses. What we don't know is the RMS averaging time. Suppose it is one cycle, then approximately it might be expected that an actual peak voltage of 156/0.707 = 220 V occurred. However, a shorter pulse of 5000 V could have produced the same result. This would not have been of the base 60 Hz frequency.
So I don't really know what that 156 V transient means. But it only occurred on the one phase and once within your test time frame.
A maximum peak pulse voltage would be a useful measurement.
.
iwire:
The one event in your data of 156 V on phase B on 1-20-2012 needs some further study that may not be possible.
A question I have is what does it mean? What is the measurement? First, I have to assume it is line B to neutral. Then what is it?
From dictionary.com transient is defined as:
adjective
1. not lasting, enduring, or permanent; transitory.
2. lasting only a short time; existing briefly; temporary: transient authority.
3. staying only a short time: the transient guests at a hotel.
4. Philosophy . transeunt.
noun
5. a person or thing that is transient, especially a temporary guest, boarder, laborer, or the like.
6. Mathematics .
a. a function that tends to zero as the independent variable tends to infinity.
b. a solution, especially of a differential equation, having this property.
7. Physics .
a. a nonperiodic signal of short duration.
b. a decaying signal, wave, or oscillation.
8. Electricity . a sudden pulse of voltage or current.
So this word can cover many different types of events.
What does RMS represent? Root Mean Square. Mathematically I can apply an RMS calculation to a sub part of a sine wave cycle. But unlikely to be of much use. More useful is a measurement over an integral number of cycles, including the possibility of just one cycle. If the measurement is over many cycles, then random start and stopping times of the measurement won't make much difference.
The definition of RMS includes an averaging element, that is the mean part of RMS.
If you perform an RMS measurement on a steady-state sine wave, and independent of whether you start the measurement at a zero crossing, or any arbitrary point, then if the measurement is performed over exactly one full cycle the answer will be the same.
A chip, like the Cirrus power monitor, uses a free running, but stable, oscillator at a frequency well above the signal to be measured to control sampling of the waveform being measured. Typical for their chip is 4.096 MHz. This is divided by 1024 resulting in 4000 samples per second. If you use 4000 of these samples to perform and RMS measurement, then 60 integral cycles create the measurement.
We shall assume your instrument is measuring an RMS value of the waveform. Probably from a much higher sampling rate than what Cirrus uses. What we don't know is the RMS averaging time. Suppose it is one cycle, then approximately it might be expected that an actual peak voltage of 156/0.707 = 220 V occurred. However, a shorter pulse of 5000 V could have produced the same result. This would not have been of the base 60 Hz frequency.
So I don't really know what that 156 V transient means. But it only occurred on the one phase and once within your test time frame.
A maximum peak pulse voltage would be a useful measurement.
.
PetrosA
Senior Member
- Location
- Lancaster, Pennsylvania
I wondered the same thing. When playing around with my clamp meter some time ago, I realized the peak hold (inrush) seemed much higher that I expected for voltage. A quick email to Agilent revealed that the peak function is not RMS while MaxMinAvg are. I wonder how the Dranetz functions?
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
120122-2241 EST
iwire:
You can find two capacitor charging plots at http://beta-a2.com/EE-curves-2.html .
These are for the 10 V, 1 Meg, and 1 ufd circuit previously mentioned. The plots are more readable if you print them.
.
iwire:
You can find two capacitor charging plots at http://beta-a2.com/EE-curves-2.html .
These are for the 10 V, 1 Meg, and 1 ufd circuit previously mentioned. The plots are more readable if you print them.
.
wirenut1980
Senior Member
- Location
- Plainfield, IN
Iwire's monitor looks to be a Dranetz PX-5. I have one as well. It samples at 256 times per electrical cycle, which translates to about 66 micro-second resolution. For transients, I believe that resolution is better at 1 micro-second.
I am not sure I trust that transient of 156 V is real and here is why. I had a case where I monitored at the 208Y/120 V main for a radio broadcast tower. I was connected phase to phase. The report said I had a transient of 821.8 V peak lasting 2 milliseconds on 10/31/2011 at 10:32:29.02. I had a waveform capture during that moment in time and did should have been able to see something, but did not see anything. In fact, the waveform capture occurred right before the utility breaker opened up due to a fault on the 12 KV line feeding this customer. Waveform capture is attached. Sorry it is a little hard to see.
I saw the waveform Iwire posted and it does not look like there is a transient on the waveform. Does it say how long the transient lasted for?
I have not had a chance to talk with Dranetz yet about my seemingly false transient.
I am not sure I trust that transient of 156 V is real and here is why. I had a case where I monitored at the 208Y/120 V main for a radio broadcast tower. I was connected phase to phase. The report said I had a transient of 821.8 V peak lasting 2 milliseconds on 10/31/2011 at 10:32:29.02. I had a waveform capture during that moment in time and did should have been able to see something, but did not see anything. In fact, the waveform capture occurred right before the utility breaker opened up due to a fault on the 12 KV line feeding this customer. Waveform capture is attached. Sorry it is a little hard to see.
I saw the waveform Iwire posted and it does not look like there is a transient on the waveform. Does it say how long the transient lasted for?
I have not had a chance to talk with Dranetz yet about my seemingly false transient.
Speedskater
Senior Member
- Location
- Cleveland, Ohio
- Occupation
- retired broadcast, audio and industrial R&D engineering
A scientist that sometimes posts on the audio forums refers to that AC input current pulse as a "haversine".
This was sort of what I wanted to point out in my post #29 some time ago noting what current wave may be as opposed to the voltage drawing as shown.
A lot of good invaluable information and experiences have been pointed out along the way since.
This series of post is a good illustration of the vast amount of brain trust that there is on this forum. It's been one of those things where you say to yourself, "Gee, I didn't think of that." Education in the class room is one thing giving you a good foundation but be being educated by real world experiences is priceless which often times involves trouble shooting when trying to find and answer to a problem. Then comes an attempt to provide a solution.
A lot of good invaluable information and experiences have been pointed out along the way since.
This series of post is a good illustration of the vast amount of brain trust that there is on this forum. It's been one of those things where you say to yourself, "Gee, I didn't think of that." Education in the class room is one thing giving you a good foundation but be being educated by real world experiences is priceless which often times involves trouble shooting when trying to find and answer to a problem. Then comes an attempt to provide a solution.
Last edited:
- Status