Dec 23, 2012, 03:25 AM Inspiration United States, MI, Ann Arbor Joined Aug 2011 895 Posts Here we go. Step_1: there are many reference scheme for half parallel YY dLRK, but i found the one posted by Manuel V. is very clear and sound (and easier to understand) http://www.rcgroups.com/forums/showp...86&postcount=7 let's use this as our reference for now. Step_2: Forget about the scheme or any other graphs, concentrate on doing a winding as shown in the pic you have posted on 148#: http://www.rcgroups.com/forums/showp...&postcount=148 First, wire number 1 (wire_1) winded on Tooth number 1 and tooth number 2 (tooth_1, tooth_2). Look at the wire_1, starting from the end on tooth_1, towards the end on tooth_2: it is a CW coil on tooth_1, and CCW coil on tooth_2. Then let's wind wire_2 on tooth_3 and tooth_4. Making a CW coil on tooth_3, and a CCW coil on tooth_4. Do the rest of the 6 wires. After you have done with the 6 wires, the 12 teeth should have 6 identical pair of coils winded following the same way. and it is CW-CCW-CW-CCW-CW-CCW-CW-CCW-CW-CCW-CW-CCW. I believe this is what you have already achieved before you solder the 12 ends last time. Am i right? Step_3, since there are one wire end coming from one tooth, so let's label them as tail_1, tail_2....tail_12 (or whatever you want to call them). Tail_1 is coming off tooth_1. Clear? Step_4, Now, you have already put a sticker on each of the 12 ends, as shown in step_3. You have 12 labelled ends now. Now, look at the scheme in step_1 provided by Manuel V. We need to convert our 12 ends to the "12 ends in his scheme". The following is how it should be converted: tail_1 = E1a tail_2 = S1a tail_3 = E3b tail_4 = S3b tail_5 = E2b tail_6 = S2b tail_7 = E1b tail_8 = S1b tail_9 = E3a tail_10 = S3a tail_11 = E2a tail_12 = S2a In Manuel V's scheme, it shows how half parall Y (or called YY) should be connected (as well as DD). Using the above conversion, You will be connecting your 12 ends as shown in Manuel V's scheme. But DONOT connect yet Everything clear? Step_5: Now, you know how to handle the 12 ends. Let's put the 2 ends that should be twisted together and soldered together, and you need to find a good way to arrange these transitions properly on the down size of the stator. You will take some time to find a good way to do this, but it should not be a problem. After you have arranged these ends properly, cut the excessive wire to appropriate length, for "twisting" and "soldering" after scratching off the insulation skin of the wire, details of doing this is shown on 142# here in this thread: http://www.rcgroups.com/forums/showp...&postcount=142 Use some heat shrinking tubes (or whatever you have) to protect the wires. Step_6: After you have twisted those ends together, as shown in Step_4, put the stator on the stator holder, and put the rotor on. (assemble the motor briefly). Then turn the rotor by hand, there should be no "drag" feel at all. Step_7: Now you can solder. Step_8: assemble the motor briefly again, do a no-load test run, eg, at 3S (11V), and measure the no-load current. At 3S, kv around 500-600, the no-load current for a 4035 can be expected be around 2A (1-3A). If everything looks fine, take off the rotor, and glue the coils with some epoxy, or what ever that can withstand some temperature. Step_9: assemble the motor completely, including the retaining collar for keeping the front bearing inside. END. As I said before, when you are doing the winding, remember to check for shorts between the wires and the stator!!!always remember to check for shorts before any big move.
 Dec 23, 2012, 03:37 AM Inspiration United States, MI, Ann Arbor Joined Aug 2011 895 Posts Remember I said Manuel V's scheme is "easier to understand"? Following is why. I am sure after looking at Manuel V's scheme, you noticed there is a big difference between yours (or what Powercroco has done) and his. Manuel's YY only have one big Y-node, containing 6 wire ends, while yours have two smaller Y-nodes, each containing 3 wire ends. Both will work and are correct. The reason why i prefer Manuel V's scheme, is that from his scheme, the YY (or DD) is views as the 4 originally serial connected coils in one phase are now connected as 2 in serial and the 2 serial in parallel. For a normal Y/D for a 12N stator, each phase takes 4 tooth (4 coils). The 4 coils are connected in series. And this becomes a normal Y or D. if you connect the 4 coils, for example, coils on tooth_1, 2, 7, 8, as 1 and 2 in series; 7 and 8 in series; and 1-2 and 7-8 in parallel: then this is a YY or DD. You can view the 9N6P's YYY/DDD, or 12N8P's YYYY or DDDD in the same way. However, in the real life, powercrco's 2 Y-nodes for YY is preferred, as this will save some room (2 smaller node is better than one huge node). And for servo motors, YYYY is a very commonly seen, and it has 4 small Y-nodes, rather than one huge Y-node containing 12 wire ends.
 Dec 23, 2012, 09:08 AM Jack USA, ME, Ellsworth Joined May 2008 19,842 Posts Modisc, When you say "Manuel V's image" in the post linked to above and call it easier to understand, you are referring to the drawn image on the white background, right? Not the photograph of the stator laying on wood and with the lines drawn on it. If that is so, there is a little danger of some confusion arising from your calling that "Manuel's image". In the past those two images, left to right as seen in that post, have come to be known as something like "Manuel V's Half Parallel dLRK-Delta wind image" and "Truglodite's Half Parallel dLRK wind image". So some of us will look at those two images side by side like that and think of the one on the left at Manuel's and not the one on the right. The one on the right, Trug's image, is a file named HalfParallelDLRK.jpg and that has sort of reinforced what it is usually called. Manuel's image is named 2p-AF-ED-BC.jpg and most of us think of it and refer to it as the Half Parallel dLRK for Delta image because it brings the pairs of ends out side by side as they will be terminated. Some of us are much more reliant on the images and the notations on the images than others. You are able to mentally visualize or conceptualize these winds much easier than I am. So I will always want to have and image at hand and, even better for me, is an image that has everything shown or noted on it as far as both winding and terminating. I appreciate your encouragement for me to be as mentally adept as you are about some these things but am not sure that will ever happen... Please read no insult or criticism into this as none is intended, just polite discussion. Jack
Dec 23, 2012, 09:36 AM
Jack
USA, ME, Ellsworth
Joined May 2008
19,842 Posts
And with the risk of raining more confusion than help on the discussion here, this morning I started to make an image for myself to use. It draws on various bits of the good stuff here, Tom's image, your explanations, the explanations of Dr. Ralph Okon over on the Scorpion kit tutorial and more.

I numbered the 12 arms into 6 groups and noted the starts and ends. Then I added notes to the image for the Y bundles and motor terminations.

So I now think I have an image that will help me. I am going to call it the "12n2pdlrk-yy.jpg" image for "Half Parallel dLRK terminated YY" for now. But if it is not right or not a good idea please tell me!

And if you want to me help further, you could tell me how those same 12 ends, using the 1S-1E to 6S-6E lables, would be grouped for a DD terminated motor. I think I have it figured out but would like to see you thoughts on it.

Jack

# Images

Dec 23, 2012, 10:56 AM
Inspiration
United States, MI, Ann Arbor
Joined Aug 2011
895 Posts
Quote:
 Originally Posted by jackerbes Modisc, When you say "Manuel V's image" in the post linked to above and call it easier to understand, you are referring to the drawn image on the white background, right? Not the photograph of the stator laying on wood and with the lines drawn on it. If that is so, there is a little danger of some confusion arising from your calling that "Manuel's image". In the past those two images, left to right as seen in that post, have come to be known as something like "Manuel V's Half Parallel dLRK-Delta wind image" and "Truglodite's Half Parallel dLRK wind image". So some of us will look at those two images side by side like that and think of the one on the left at Manuel's and not the one on the right. The one on the right, Trug's image, is a file named HalfParallelDLRK.jpg and that has sort of reinforced what it is usually called. Manuel's image is named 2p-AF-ED-BC.jpg and most of us think of it and refer to it as the Half Parallel dLRK for Delta image because it brings the pairs of ends out side by side as they will be terminated. Some of us are much more reliant on the images and the notations on the images than others. You are able to mentally visualize or conceptualize these winds much easier than I am. So I will always want to have and image at hand and, even better for me, is an image that has everything shown or noted on it as far as both winding and terminating. I appreciate your encouragement for me to be as mentally adept as you are about some these things but am not sure that will ever happen... Please read no insult or criticism into this as none is intended, just polite discussion. Jack
Thank you Jack, for pointing this out. I have no idea that the two images in Manuel's post were from different persons. Yes, i am referring to the image on the right, which has those ends labelled and the formula for connection.

I was new to this rewinding stuff about a year ago. I learnt a lot during the past year, and what i have been "concepturized" is my own understanding of the rewinding which may be wrong. But these ideas truely helps me a lot in understanding the rewinding, and I just hope everyone else will benefit as well.
 Dec 24, 2012, 04:10 AM Aka: Tom Jenkins United States, FL, West Palm Beach Joined Aug 2008 4,726 Posts There, I found another error in the lables so here it is again, this is labled per Manuel's dwg as reference. Modisc, do you have anything on how you strip the ends to ready for solder? Last edited by ApexAero; Dec 24, 2012 at 04:35 AM.
Dec 24, 2012, 08:02 AM
Inspiration
United States, MI, Ann Arbor
Joined Aug 2011
895 Posts
Quote:
 Originally Posted by ApexAero There, I found another error in the lables so here it is again, this is labled per Manuel's dwg as reference. Modisc, do you have anything on how you strip the ends to ready for solder?
Looks correct this time. But I would prefer to stick to the original reference scheme, cause I also feel confusing looking at these labels sometimes.

I simply use a knife to scratch off the skin. For single strand of thick wire, this is easy. But for multiple fine wires, it would need to use some kind of insulation paint solvent. I personally never used such solvent, since I only do thick wire. Such solvent can be quite hazardous, I am not sure it could be found in US market. The solvent may also work with only certain types of magnetic wire insulation. Anyway I suggest just scratch using knife, easy and dependable.
Last edited by modisc; Dec 24, 2012 at 08:08 AM.
Dec 24, 2012, 08:14 AM
Inspiration
United States, MI, Ann Arbor
Joined Aug 2011
895 Posts
Quote:
 Originally Posted by jackerbes And with the risk of raining more confusion than help on the discussion here, this morning I started to make an image for myself to use. It draws on various bits of the good stuff here, Tom's image, your explanations, the explanations of Dr. Ralph Okon over on the Scorpion kit tutorial and more. I numbered the 12 arms into 6 groups and noted the starts and ends. Then I added notes to the image for the Y bundles and motor terminations. So I now think I have an image that will help me. I am going to call it the "12n2pdlrk-yy.jpg" image for "Half Parallel dLRK terminated YY" for now. But if it is not right or not a good idea please tell me! And if you want to me help further, you could tell me how those same 12 ends, using the 1S-1E to 6S-6E lables, would be grouped for a DD terminated motor . I think I have it figured out but would like to see you thoughts on it. Jack

Using the following convertion,
tail_1 = E1a
tail_2 = S1a
tail_3 = E3b
tail_4 = S3b
tail_5 = E2b
tail_6 = S2b
tail_7 = E1b
tail_8 = S1b
tail_9 = E3a
tail_10 = S3a
tail_11 = E2a
tail_12 = S2a

And use Trug's pic.
Dec 24, 2012, 09:19 AM
Jack
USA, ME, Ellsworth
Joined May 2008
19,842 Posts
@modisc - Thanks for the details on the DD, that will help me a lot.

@all,

On the stripping, even the 32 AWG stuff, I use the following.

1 - Glue (CA, Gorilla, etc.) two tongue depressors back to back to make them a little stiffer and bending resistant. Then clamp that to the bench top so it over hangs the edge of the work bench 2" or so.

2 - Lay one wire end on the wood and rotate it as you scrape lightly with an #11 Exacto knife. Go all the way around and that will remove about 90% of the insulation.

Then take a pink cushion board disposable nail file (280 to 3209 grit typically) and cut it across in strips (heavy duty scissors or tin snips will work) to make a bunch of 3/8" to 1/2" wide by 3/4" long double sided sanding blocks.

3 - With the scraped end still laying on the tongue depressor, polish the scraped end by laying the sanding block on it and sliding it off to the end. it will take about six slight rotations to get all the way around and the end will be bright and bare. The cushioned board will wrap the wire a little and overlap on each pass, especially on the bigger wire.

But even the 32 AWG wire will survive the scraping and polishing without losing much or any copper.

I always dip the scraped end in the can of paste flux and tin it with solder and look at that closely. Any dark spots or un-tinned wire means it is not ready yet. If there is any flux residue a little more work with the cushion board sanding blocks will remove it.

I used to twist the ends together before I soldered them but have stopped doing that in most cases, particularly for a Wye bundle. Now I get the three end stripped and tinned and then get them together in a triangular bundle of parallel strands. I lash the bundle together with a scrap of winding wire that still has the coating on it. Two or three round turns on the three strand bundle (around the still coated wire below the stripped ends) will hold it together for soldering. The I dip the bundle in the flux again.

Then just a small drop of solder on a clean tip will flow quickly and finish the bundle. I trim the length so that the soldered length is 2-3 times the bundle diameter and that's it.

I don't think there is any clear advantage in having a long length of twisted and soldered wire and should I want to unsolder the bundle later all it takes is a touch of the iron to set all ends free. Nothing short of a total motor disaster in progress will get solder hot enough to melt (600F or so) in a running motor.

On a seriously big motor with lots of room in the end plate I might twist the wires. But otherwise I cater to Murphy's Law and the possibility that I might want to unsolder later.

Jack

# Images

Last edited by jackerbes; Jan 12, 2013 at 05:34 PM.
Dec 24, 2012, 10:21 AM
Jack
USA, ME, Ellsworth
Joined May 2008
19,842 Posts
Quote:
 Originally Posted by modisc Hi, about the DD, Using the following convertion, tail_1 = E1a tail_2 = S1a tail_3 = E3b tail_4 = S3b tail_5 = E2b tail_6 = S2b tail_7 = E1b tail_8 = S1b tail_9 = E3a tail_10 = S3a tail_11 = E2a tail_12 = S2a And use Trug's pic.
After I studied this a little I have another question about it. This is the same sequence you used up in post #151 also. When I look at I now see that your tail number sequence does not have the ends in sequential agreement with trug's image or with the arms if numbered in sequence.

tail_1 = E1a = arm #1
tail_2 = S1a = arm #2
tail_3 = E3b = arm #4
tail_4 = S3b = arm #3
tail_5 = E2b = arm #5
tail_6 = S2b = arm #6
tail_7 = E1b = arm #8
tail_8 = S1b = arm #7
tail_9 = E3a = arm #9
tail_10 = S3a = arm #10
tail_11 = E2a = arm #12
tail_12 = S2a = arm #11

It looks like, by the arm numbers, arms 3, 4, 7, 8, 11, and 12 are out of order. Is that intentional and necessary? Or have I missed something here?

Jack
Dec 24, 2012, 08:03 PM
Inspiration
United States, MI, Ann Arbor
Joined Aug 2011
895 Posts
Quote:
 Originally Posted by jackerbes After I studied this a little I have another question about it. This is the same sequence you used up in post #151 also. When I look at I now see that your tail number sequence does not have the ends in sequential agreement with trug's image or with the arms if numbered in sequence. read: your tail_# = trug image marking = sequentially numbered arms tail_1 = E1a = arm #1 tail_2 = S1a = arm #2 tail_3 = E3b = arm #4 tail_4 = S3b = arm #3 tail_5 = E2b = arm #5 tail_6 = S2b = arm #6 tail_7 = E1b = arm #8 tail_8 = S1b = arm #7 tail_9 = E3a = arm #9 tail_10 = S3a = arm #10 tail_11 = E2a = arm #12 tail_12 = S2a = arm #11 It looks like, by the arm numbers, arms 3, 4, 7, 8, 11, and 12 are out of order. Is that intentional and necessary? Or have I missed something here? Jack
I am sorry that I might not be clear enough, but what you have pointed out is correct, and it is intentional and necessary.

Please look at the 2 attachment scheme, one is from Apex and the other is from Trug.

You see, Apex's scheme: assume tooth_1 has a CW coil, then from tooth_1 to tooth_12, the coils will be : CW-CCW-CW-CCW-CW-CCW-CW-CCW-CW-CCW-CW-CCW.
while for Trug's scheme, assume tooth_1 is CW, then the 12 teeth will be CW-CCW-CCW-CW-CW-CCW-CCW-CW-CW-CCW-CCW-CW.

You can see that for tooth_3/4, Tooth_7/8, tooth_11/12, the direction is different as shown in the 2 pics.

If you wind following Trug's scheme, you can just follow the formula shown in his scheme. However, if you wind following Apex's scheme, you will need to change the direction for tooth 3/4, 7/8, 11/12 by exchange the two ends of the wire on the above pairs of tooth. I said I prefer to connect as Trug's scheme, but I prefer to wind as Apex's scheme, because in Apex's scheme, all the 6 pair of teeth can be winded in exactly the same way, and this, i think, is best for achieving good copper fill.

# Images

 Dec 25, 2012, 07:43 AM Jack USA, ME, Ellsworth Joined May 2008 19,842 Posts OK, now I get it. I had not noticed the difference on those arms. And it also explains why Trug's termination schemes will not work with Aero's winding schemes too. I think, all things considered, that a better image would be Aero's image with six pairs of CW-CCW phases and then to also have the ends all labelled and also have the image labelled with how to terminate the six arm pairs/12 wire ends for a YY motor and also for a DD motor. A thing that happened to once in the past was that when I did Trug's wind terminated Wye with a single Wye bundle, I could not get it to run. But when I changed it so that it had two Wye bundles it did run. At that time someone attributed that to being the result of the ESC I was using. They said that some Castle Creations ESCs (as I was using one) would do that. That was some time ago, the details are foggy in memory and I can't find the discussion. Have you ever heard of anything like that? Jack
Dec 25, 2012, 05:51 PM
Inspiration
United States, MI, Ann Arbor
Joined Aug 2011
895 Posts
Quote:
 Originally Posted by jackerbes OK, now I get it. I had not noticed the difference on those arms. And it also explains why Trug's termination schemes will not work with Aero's winding schemes too. I think, all things considered, that a better image would be Aero's image with six pairs of CW-CCW phases and then to also have the ends all labelled and also have the image labelled with how to terminate the six arm pairs/12 wire ends for a YY motor and also for a DD motor. A thing that happened to once in the past was that when I did Trug's wind terminated Wye with a single Wye bundle, I could not get it to run. But when I changed it so that it had two Wye bundles it did run. At that time someone attributed that to being the result of the ESC I was using. They said that some Castle Creations ESCs (as I was using one) would do that. That was some time ago, the details are foggy in memory and I can't find the discussion. Have you ever heard of anything like that? Jack
I have no idea whether one Y bundle or multiple Y bundle for multiple-Y connection has any difference. Personally, i think there is no difference.

But I do notice that multiple Y bundle is preferred, as seen in servo motors. the 12N8P servo motor, used YYYY connection, with 4 Y bundles.

However, there are discussion about YY is superior to other type of connections from HF forum. I doubt it. The only thing i can think of is that because YY is a type of Y, it will not have a drag feel for non-ABC windings (though the "drag" has no influence on the performance).
Dec 26, 2012, 07:37 AM
Aka: Tom Jenkins
United States, FL, West Palm Beach
Joined Aug 2008
4,726 Posts
Here it is again, done by my drawing, sure hope its correct before I trim the tails!
I labeled the stator tops rather than the wire. Also, I still count 11.5 turns?
Tom

# Images

Dec 26, 2012, 09:20 AM
Inspiration
United States, MI, Ann Arbor
Joined Aug 2011
895 Posts
Quote:
 Originally Posted by ApexAero Here it is again, done by my drawing, sure hope its correct before I trim the tails! I labeled the stator tops rather than the wire. Also, I still count 11.5 turns? Tom
looks great. make sure you put on the rotor bell and at least do a "turn the shaft by hand" test after you connect the 2 Y bundle, as well as each pair of wires in the three ESC leads. There should be no "drag" feel.

I am not absolutely sure that i am right but, i always just count the upper surface where there is no wire coming out. So yours looks like 5+6T to me.