jumper
Senior Member
- Location
- 3 Hr 2 Min from Winged Horses
Stat guys, probability of getting rotation right?
- Thread starter chris kennedy
- Start date
- Status
ptonsparky
Tom
- Occupation
- EC - retired
You must be one of the lucky young fellas that has never seen all phases the same color, been distracted mid stream or just had a BF.
mivey
Senior Member
I plead the 5th.
- Location
- Lockport, IL
- Occupation
- Retired Electrical Engineer
zog
Senior Member
- Location
- Charlotte, NC
50-50-90 was a term with many applicationsI recall that 50-50-90 rule as meaning that if you gave one of us a 50-50 guess, we would guess wrong 90% of the time.
A phase rotation meter is a very useful tool. After you determine the phase rotation on your system, IF all of the power distribution and motor circuits are marked correctly, a person should be able to be 100% correct on the motors rotation. Tip: All standard three phase squirrel cage motors rotate clockwise- facing the non-drive end, when made up A-L1, B-L2 and C-L3. ( Yes, I know that the motor is marked T1,2,3,4,5,6,7,8,9 , but your ABC connection would vary depending on voltage.)
Last edited:
RichB
Senior Member
- Location
- Tacoma, Wa
- Occupation
- Electrician/Electrical Inspector
50/50--reasons why--after you make the first connection you have 2 more to make--if it is wrong then you swap the last two connections--after the first one there are only 2 possabilities left correct rotation or backwards
Barndog
Senior Member
- Location
- Spring Creek Pa
I admit i have wired it wrong twice. I had my boss and his boss watching and i had all three wires the same color took off both wirenuts and looked up to talk to them and forgot what wire went where. So with outside interferance it is possiable to get it wrong twice.
Yes, but as with phase rotation, there are just two possible outcomes for each toss of the coin regardless of how many times you do it.
You and i are the only two out of all the posts to buck the 50-50 rule. I have been 100% correct or Just plain lucky in the past with the motors i have wired. The only exception i will admit to is when it comes to DC motors.
My theory is, (1) if your at the butt end of the motor looking towards the shaft, and you see the direction the shaft wants to turn in order to run the device, Lets say clockwise, Then (2)you need to check the incoming feeder rotation and see what that is, Lets say ccw. Then (3) when you land the feeder, you reverse the feeder at any point depending on the devices in line to make the motor run clockwise.. keeping (L1to T1)(L2 to T2) and (L3 to T3)
With VFD's, they are phase rotation sensitive and need to be 1st, clock wise rotation incoming then ccw going out to the motor.
Rick
Open Neutral
Senior Member
- Location
- Inside the Beltway
- Occupation
- Engineer
The odds are only 50-50 if there's no harm to anything but pride if it's bass-ackwards.
If, however, the fragulator will turn to junk if you do; they go up to 93%...
If, however, the fragulator will turn to junk if you do; they go up to 93%...
That depends on the kind of VSD. Most of the common inverters have a plain uncontrolled rectifier at the front end. B6U.
This configuration is insensitive to phase rotation.
mtfallsmikey
Senior Member
Musings from the house plumber..
Musings from the house plumber..
Had my HVAC contractor install a new 15 ton compressor in one of the self-contained units here, he had just bought a new fancy-schmancy phase rotation detector. Hooked up the compressor just like it had come out, the new instrument indicated a wrong connection, but he fired the unit up, ran well. Then, he hooked up the compressor the way the meter indicated it should be and "grrrrr".......no start.....wrong! He returned the meter the next day.
Musings from the house plumber..
Had my HVAC contractor install a new 15 ton compressor in one of the self-contained units here, he had just bought a new fancy-schmancy phase rotation detector. Hooked up the compressor just like it had come out, the new instrument indicated a wrong connection, but he fired the unit up, ran well. Then, he hooked up the compressor the way the meter indicated it should be and "grrrrr".......no start.....wrong! He returned the meter the next day.
- Location
- Mission Viejo, CA
- Occupation
- Professional Electrical Engineer
If rotation alone is the concern 50% is correct (only two ways to rotate, cw & ccw); however, if phasing is a concern because of a potential paralleling of sources the odds of being correct by random is 16.7%. Assuming one side is the "reference", there are six ways to wire the other side, only one is correct.
I did it the hard way, I flipped a nickel 10 times and wrote down the face hit rate. T, T, T, T, H, H, T, H, H, T.
Only one run was enough to support the premise. T is 6/10 and if I had arbitrarily chosen a smaller sample size, T could easily be 3/3 or 4/4.
T is 1/2 on paper, but in practice not 1/2, and only converges to 1/2 with large sample size.
Only one run was enough to support the premise. T is 6/10 and if I had arbitrarily chosen a smaller sample size, T could easily be 3/3 or 4/4.
T is 1/2 on paper, but in practice not 1/2, and only converges to 1/2 with large sample size.
mivey
Senior Member
Incorrect. I just flipped a penny (I'm not rich like you :grin
and got H, T. I got convergence in two flips.The truth is, the sample size required for convergence is random and is not necessarily large.
But, as your "test" shows, there are only two possible outcomes for each flip.
The number of times you do it has no bearing on that. And, an important point, the coin has no memory. The outcome of the next flip has no bearing on the previous flip or flips.
Yes.
So, I flip a coin twice and the first result is H. On the second flip the coin has a 1/2 probability of H, matching the first flip and yielding 2/2 for the run and a 1/2 probability of T, balancing the first flip and yielding 1/2 for the run.
That test, on paper, yields 2/2 for half of the runs (with large sample size).
Saying the run will be 1/2 has the impression of mathematical certainty, deterministic. The next flip 'should be' X to yield 1/2 for the run. No, the next flip will be what you need it to be only 1/2 of the time and incorrect the other half.
If you physically sample the runs, the results plot a distribution around 1/2. By observation, most of the run results are not 1/2, close to, but not exactly 1/2.
My point was that the outcome next toss is not affected by what went before. If you got ten or a hundred heads in a row you still have a one in two chance that you'll get another.
- Status