I have a stator from a 1/2 hp, four speed, 3-phase motor. Coming from the stator windings there are 12 wires (two separate sets of six on opposite ends of the motor). Unfortunately these are identified in what appears to be a non-standard numbering system and I have no diagram. One set of wires is numbered (5,6,9,11,14,15) and the other (7,8,12,13,16,17). I am trying to identify pairs so that I can wire the rest of the motor. I have determined the resistance between all possible two wire connections. Between the two groups all possible connections are open. Within each group, all possible connections show some small resistance. (I would post a table but I don’t see how to insert a picture or attach a file.)
Can someone explain how to determine which two wires are a pair?
Please give the make and model of the motor, along with the specifications on its tag.
Measure the resistance of one wire to all others, identify if there are isolated groups.
You may have four separate 3PH windings each with a different pole pitch.
Those 3PH windings could be simple delta, or to make it more difficult, in star with a floating common point.
You might find that common neutral point is inside the motor.
If you connect a low voltage DC to two wires, you can then use a magnetic compass to identify the pole pitch and polarity for that winding.
Are there 48 slots for wire in the field? Maybe that gives you speeds of 2, 3, 4 and 6.
Be brave, start measuring continuity or resistance, tabulate the results.
We can then reassess the strategy.
The motor is 230v, 3 ph, 1/2 hp vintage motor madeby Oliver Machinery Company. It originally had a drum switch that provided speeds of 600-1200-1800-3600 rpm. Speculation is that it is most likely a double Dahlander winding configuration.
I suspect that there are more than 48 slots as the motor is huge, but I will have to look at it tomorrow and count them.
Here are the resistance readings. Any help in interpreting them would be greatly appreciated.
The Wiki terminals are: 1U, 2U, 1V, 2V, 1W, 2W.
Likewise, one set of your terminals are connected in a ring with this order:
7, 8, 17, 12, 13, 16, 7.
From the measurements alone;
I cannot identify which is the first or the second connection of a phase, does it matter?
I cannot identify the direction of rotation, you may need to reverse the ring by swapping two lines.
Those need an experimental test.
You have something wrong with the 5, 6, 9, 11, 14, 15 group readings.
The ~8.4 ohm pairs show which other two terminals are adjacent in that sequence.
You have a ring of only 5 terminals.
It appears that terminal 14 is only connected through a single coil to 15.
While 15 is connected to terminals 9, 11 and 14 through single coils.
I suspect that there are more than 48 slots as the motor is huge, but I will have to look at it tomorrow and count them.
I just checked and there are actually 48 slots in the stator
Grubbs1960 said:
Here are the resistance readings. Any help in interpreting them would be greatly appreciated. View attachment 373748
As pointed out by Baluncore, there is an error in this table. The junction of the cell representing the row 15, column 14 should read 8.3 instead of open.
Thanks for your response and that link. I’m getting more confident my motor is a double Dahlander, i.e. it has two independent sets of windings. The motor has a complicated mechanical rotary switch that I believe, depending on which position it was in, provided power to only one of those sets of windings, with two switch positions for each set of windings. One of those positions engages contacts that arranged the selected winding set in the low speed configuration, the other position connecting the high speed configuration. Does this seem correct?
Baluncore said:
The Wiki terminals are: 1U, 2U, 1V, 2V, 1W, 2W.
Likewise, one set of your terminals are connected in a ring with this order:
7, 8, 17, 12, 13, 16, 7.
Am I to interpret this to mean that the wires labeled 7&8 are the ends of one winding, 17&12 another and 13&16 the third?
Baluncore said:
From the measurements alone;
I cannot identify which is the first or the second connection of a phase, does it matter?
I’m not sure I understand what you’re asking. Assuming that it does matter, what test could I perform to determine which is the first or second connection of a phase?
Baluncore said:
I cannot identify the direction of rotation, you may need to reverse the ring by swapping two lines.
Those need an experimental test.
I realize that when I connect a set of windings to power that there is a 50% chance the direction of rotation, but that simply exchanging any two of the power leads will reverse the direction. Is that the experimental test you are referring to?
Baluncore said:
You have something wrong with the 5, 6, 9, 11, 14, 15 group readings.
The ~8.4 ohm pairs show which other two terminals are adjacent in that sequence.
You have a ring of only 5 terminals.
It appears that terminal 14 is only connected through a single coil to 15. While 15 is connected to terminals 9, 11 and 14 through single coils.
You are correct. The junction of the cell representing the row 15, column 14 should read 8.3 instead of open.
Again, thanks for your help and any further assistance.
Am I to interpret this to mean that the wires labeled 7&8 are the ends of one winding, 17&12 another and 13&16 the third?
7, 8, 17, 12, 13, 16, 7.
There is one winding between 7&8, another between 8&17, 17&12, 12&13, 13&16, 16&7
That makes a circuit of six windings in a loop. Have you seen the Wiki diagram?
Grubbs1960 said:
Is that the experimental test you are referring to?
It is half the test. I don't trust my guess that terminals 7 and 8 for example are the same when making the double Y.
Grubbs1960 said:
You are correct. The junction of the cell representing the row 15, column 14 should read 8.3 instead of open.
You need to look deeper. Terminal 14 now only has one 8.3 ohm coil attached, it needs two.
Check that every terminal has two low-resistance mates.
6, 5, 9, 15, 11 is only five coils, it needs six, where is terminal 14 in the loop.
I have hooked windings up to mains power in series with a suitable load to limit current. Probing around carefully with a screw driver can go a long way to identification. The last one I did was a 15 HP motor so I had the luxury of not worrying about current getting too high. Windings would carry whatever I could throw at it on a twenty amp circuit. I used a right angle grinder which gave a nice surge in current when turned on. Easily detectable by the screw driver. Of course this assumes the rotor has been removed.
7, 8, 17, 12, 13, 16, 7.
There is one winding between 7&8, another between 8&17, 17&12, 12&13, 13&16, 16&7
That makes a circuit of six windings in a loop. Have you seen the Wiki diagram?
Yes, but I when I did I was under the (apparently mistaken) impression that each set of six wires represented the ends of three different winding sets. And that bias caused my cursory examination of the diagrams to recognize only three windings.
I now realize after reading what you wrote and re-examining the diagrams that the twelve wire ends probably are singe ends of 12 windings, the opposite ends being connected internally inside the stator. Is that correct?
Baluncore said:
It is half the test. I don't trust my guess that terminals 7 and 8 for example are the same when making the double Y.
You need to look deeper. Terminal 14 now only has one 8.3 ohm coil attached, it needs two.
Check that every terminal has two low-resistance mates.
6, 5, 9, 15, 11 is only five coils, it needs six, where is terminal 14 in the loop.
Is this still true if the 8.3 ohm value is added as described in my previous post? If so, help me understand how to “look deeper.”
In addition to the choice of speeds, it is usual to use wye-delta starting for large induction motors.
Luckily, this is not a large and powerful motor, because in a Dahlander pole changing motor, the Y configuration is the fast mode, while the Δ is the slow one, at half the Y speed.
gah said:
But it is supposed to be 1:2:3:6.
One set of Dahlander windings only offers a factor of two. That is why there are two independent sets of windings.
Power the first, and you get a speed of: 1 in Y, or 2 in Δ.
Power the second, and you get a speed of: 3 in Y, or 6 in Δ.
Luckily, this is not a large and powerful motor, because in a Dahlander pole changing motor, the Y configuration is the fast mode, while the Δ is the slow one, at half the Y speed.
One set of Dahlander windings only offers a factor of two. That is why there are two independent sets of windings.
Power the first, and you get a speed of: 1 in Y, or 2 in Δ.
Power the second, and you get a speed of: 3 in Y, or 6 in Δ.
This is my understanding as well.
@Baluncore: My post #9 was a response to some of the previous observations you’ve made and I had a couple of additional questions that I am hoping you can answer.
Sorry about the delay. It was 3:15AM here when your post #8 arrived, I must have nodded off.
I don't trust our Dahlander assumptions, so I need to go back to check everything.
No need to apologize. I was just afraid that you might have missed that particular post since you responded to one that was a few posts later.
If there is some additional tests that I can conduct to help verify this is a Dahlander and/or better understand the twelve loose winding ends, please advise.
If there is some additional tests that I can conduct to help verify this is a Dahlander and/or better understand the twelve loose winding ends, please advise.
I have tried to find the problems by pattern, but I am too confused, so I hallucinate false data.
Draw up your data in two separate tables, then consider each alone.
Table A, with rows and columns of: 5, 6, 9, 11, 14, 15.
Table B, with rows and columns of: 7, 8, 12, 13, 16, 17.
See the bad example at the bottom of this post.
Check the symmetry of your data by comparing all the rows against their columns.
Next practice your false Sudoku.
Find the pairs of lowest numbers in each row and column.
There should only be two distinct low values in each column.
There should only be two distinct low values in each row.
You should be able to follow those pairs of lowest numbers in a closed chain of six coils. If that cannot be done, then the data is faulty, or it is not Dahlander connected.
As an example of my Dahlander confusion.
Lookup column 6, rows 5 and 11 are low.
Lookup column 5, rows 6 and 9 are low, but we just came from 6;
So lookup column 9, rows 5 and 15 are low.
Lookup column 15, where three rows are low, 9, 11 and 14.
That is a problem because the chain must branch to 11 or 14.
11 must be wrong, since the loop would close with only 5 elements, with element 14 missing.
We have visited 11 earlier, so follow 14.
Lookup column 14, only 15 is the low, but it should close the loop of six coils to column 6.
This data is bad, it is asymmetric, or not Dahlander connected.
Code:
# 5 6 9 11 14 15
5 X 8.6 8.4 13.1 14.5 12.8
6 8.6 X 12.8 8.4 12.8 14.3
9 8.4 12.8 X 14.3 12.7 8.3
11 13.1 8.4 14.3 X 12.8 8.4
14 14.5 12.8 12.7 12.8 X 8.3
15 12.8 14.3 8.3 8.4 8.3 X
# 7 8 12 13 16 17
7 X 6.5 11.0 9.8 6.5 9.8
8 6.5 X 9.7 10.9 9.7 6.4
12 11.0 9.7 X 6.6 7.7 6.5
13 9.8 10.9 6.6 X 7.7 9.8
16 6.5 9.7 7.7 6.5 X 6.5
17 9.8 6.4 7.7 9.8 6.5 X
I have tried to find the problems by pattern, but I am too confused, so I hallucinate false data.
Draw up your data in two separate tables, then consider each alone.
Table A, with rows and columns of: 5, 6, 9, 11, 14, 15.
Table B, with rows and columns of: 7, 8, 12, 13, 16, 17.
See the bad example at the bottom of this post.
Check the symmetry of your data by comparing all the rows against their columns.
Next practice your false Sudoku.
Find the pairs of lowest numbers in each row and column.
There should only be two distinct low values in each column.
There should only be two distinct low values in each row.
...
I have rearranged the data into two tables, as suggested. Neither conforms to the two distinct low values in each column and row. I am going to retest the leads the next time I am in the shop and will post my results. In the meantime, please find below the two tables just in case you are able to tell that the issue is not faulty data, but instead suggest some other configuration.
In the meantime, please find below the two tables just in case you are able to tell that the issue is not faulty data, but instead suggest some other configuration.
It appears the motors are Dahlander connected, but you have three or more errors in the data.
If I assume there is only one fault in the black table, then it seems you have swapped the measurements of 14 and 15. That could have been done when measuring against wire 9, or against wire 11. Undoing one of those swaps, gives two low in each column and two in each row, with a Dahlander sequence of six wires.
If the swap was during wire 9 measurements, then the order is; 5, 6, 11, 15, 14, 9, 5.
If the swap was during wire 11 measurements, then the order is; 5, 6, 11, 14, 15, 9, 5.
In the red table, you have 14 low values, but there should only be 12. Also, the table is not symmetrical.
Column 17, row 13, should probably read 9.8, not 6.5.
But that still leaves wire 17 connected to three others, namely 8, 12 and 16.
One of those three is probably a wrong value.
To solve problems with multiple wires, you must be obsessive, continuously doubt and methodically check your work.
Below find new data, which does include only two distinct low readings in each row and column. My apologies for posting incorrect data in prior posts. Any insights this new data provides will be greatly appreciated.
(Edited to add: I believe if I have gained a correct understand of what I have learned from you that the first table indicates that the coils are connected as 5-6-11-14-15-9-5
and that the second table shows coils connected as 7-8-17-12-17-16-7. Please confirm or correct.)
Thank you profusely for getting me on the right track! It will probably be next week before I actually get a chance to actually test this, but I will post results when I do.