HK 7050 (10AWG / 2.58mm, 5+5T, YY)

[modisc731]

A record?

Just finished winding this beast. Some winding techniques are used to make it happen. Basic rule is "symmetry".

winding done with 10AWG wire (2.58mm copper diameter; ~2.65mm outer diameter).
Powercroco's 6-wire technique
5+5T.
Connected as YY, maybe. Or D for exact symmetry.

The 7050 has air gap of ~0.8mm, which is kinda disappointing....

STSD is possible and may add one more turn, but too hard to wind; too much loose ends will be generated from STSD, and may not be practical to arrange nicely for this 2.6mm wire...
All 12 loose ends are packed inside, no chance for loosing up, but still need extra coating to prevent vibration.

For winding techniques:
First: do 2/1, 2/1, 2/1, 2/1, 2/1, 2/1
Then adds up: 3/2....
Then 4/3
then 4/4
Then 5/5 for the last inserted turns.

The two 5-cents coin is just used for reference of SIze.

[Jmann841]

Holy moly! What will you be using that for!?

With such thick gauge wire, do you think there will be any issues with the skin effect?

[modisc731]

Originally Posted by Jmann841

Holy moly! What will you be using that for!?

With such thick gauge wire, do you think there will be any issues with the skin effect?

http://diyaudioprojects.com/Technica...an-Wire-Gauge/
10 AWG: 2600Hz is the limit.
for 2600Hz, electric rpm = 2600X60
For this 10 pole motor: actual RPM = 2600X60/10 = 15600.

for scorpion factory winding: 5T delta = 340Kv.
so 10T YY should be 196Kv

So this motor should be good even using 20S, if connected as YY.

[Jmann841]

Very nice! Is this going in a heli?

I need to learn more about skin effect, I honestly don't know too much about it, other than it can be an issue with larger gauge wire. Looks great!

[powercroco]

in partloaded situation with 8/16/32kHz (depending on esc) you just are deep within the dangerous zone. also for eddy currents into the copper.
we found this by measuring this (only winded for show in 2010) 4+4 x 2,5mm YY 7050:



this is a 7050 with 9+9x1,8mm YY.
n spec is 132/V:



thisone works very well and with good efficiency.

[omerco]

Originally Posted by modisc731

http://diyaudioprojects.com/Technica...an-Wire-Gauge/
For this 10 pole motor: actual RPM = 2600X60/10 = 15600.

Modisc:

Could you explain the 60 in the calculation please? 60 for cogging?

I think you invented a new winding technique (?) if i understand you were stepping up with both teeth together from bottom to top?

[omerco]

Originally Posted by powercroco

in partloaded situation with 8/16/32kHz (depending on esc) you just are deep within the dangerous zone. also for eddy currents into the copper.
.

Ralph: Why does it meter if it is partial load or not? does Modisc have a mistake in his calculation, or only experiments show different results than the calculation?

Edit: OK i think i understood: when throttle is not 100% the ESC uses it's set switching and then you need to multiply Modisk's calculation with the switching freq. is this correct?

[powercroco]

part load in block commutation divides the broad full block into a lot of smaller ones.
they only switch between max and 0, not between max+ and max- but anyway we found, that this ripple matters in very thick wires.
we went the bifilar winding way to prevent this effect.
for modiscs winding 2 x 1,8 in parallel instead of the 2,56 single would imo be the better choise.
but of course may be, my measurements are wrong.

[modisc731]

Originally Posted by powercroco

in partloaded situation with 8/16/32kHz (depending on esc) you just are deep within the dangerous zone. also for eddy currents into the copper.
we found this by measuring this (only winded for show in 2010) 4+4 x 2,5mm YY 7050:

this is a 7050 with 9+9x1,8mm YY.
n spec is 132/V:

thisone works very well and with good efficiency.

Good to know the reference specs: 18T YY = Kv132. Then 10T YY should be ~Kv240...

As powercroco said, smaller gauge would be better for overall performance, and this 2.6mm 10T YY is probably only good for full throttle application. I trust real experiment result then theory
Actually, by CAD drawing, dual 1.9mm or 1.8mm can achieve higher copper fill (~15-20% more) than single 2.58mm. but that would result in multiple layers of wires and may look messy. I am a guy more into "picture" than "performance", too bad for me.

STSD is not at all my invention. Though i come up with this winding technique myself, but it turns out it has been used by many motor winding ancestors long ago, and it is extensively used in "servo motor with distributed windings".

[modisc731]

Originally Posted by omerco

Modisc:

Could you explain the 60 in the calculation please? 60 for cogging?

I think you invented a new winding technique (?) if i understand you were stepping up with both teeth together from bottom to top?

the 60 is simply Hz to RPM conversion

For the winding technique question:
Assume you use wire_A to wind on tooth 1 and tooth 2:
After you wind tooth 2 with half of wire_A, it is very difficult to wind the first turn on tooth_1 with the other half of wire_A, because the other half of wire_A is too close to the bottom of the slot, and hard to clamp without scratching the adjacent coils on tooth_2.
This is, in my opinion, the hardest part of using Powercroco's 6 wire technique.

To avoid this, you should not complete the windings on tooth_2. Instead, you only do one TURN on tooth_2 first. Then move the coil outwards, and do the first turn on Tooth_1. In this way, the first turn on Tooth_1 will be much easier, and can be clamped flat easily. Then you continue winding one turn each time, on each tooth.

In this way, every winding are basically done in exact same condition: same room for winding. So, if you can do tooth_1 and tooth_2, you surely can do the rest tooth without much problem.
This is what i mean by symmetry. But this is only necessary for extra-compact thick wire winding, or maximized copper fill.
If winding single 1.9mm on this 7050, i believe it is not at all necessary, to wind like this.

The downside of this technique, is you are going to wind, with all the wires hooked up on the stator...can be kinda difficult to manipulate: very heavy, and messy sometimes

The last turn of my "2.58mm, 5+5T", the turn as the 2nd layer, is my invention, i believe. That is the turn, "inserted" in the slot: all loose ends will be packed inside by one another. In this way, all coils are tightly packed and secured by coils themselves.
same technique is used in a 5035 "1.9mm, 6+6T"

[modisc731]

Originally Posted by modisc731

the 60 is simply Hz to RPM conversion

I actually am not so sure about one thing:

for 10 pole motor, therefore: 5 pairs of pole
if electric RPM is 156000, the actual RPM should be 156000/10, or 156000/5???

[Jmann841]

So if using a single strand thick wire, a lower PWM setting on the ESC is better to reduce any skin effect issues?

Is PWM something totally different and does not have any play with skin effect or does it play a role?

My Savox 5055-550kv motor gets very hot, even without much load so I keep trying a higher PWM to lower the heat, but, should I try lower? I'm going to give 8khz a try and see how it does. I tried 16khz yesterday and it aws still very hot. Not sure if I should go to 8 or to 24khz? ESC temps are no issue.

I tested 8khz just now and it seemed cooler, Could this be helping it?

[1BOHO]

Originally Posted by modisc731

the 60 is simply Hz to RPM conversion

For the winding technique question:
Assume you use wire_A to wind on tooth 1 and tooth 2:
After you wind tooth 2 with half of wire_A, it is very difficult to wind the first turn on tooth_1 with the other half of wire_A, because the other half of wire_A is too close to the bottom of the slot, and hard to clamp without scratching the adjacent coils on tooth_2.
This is, in my opinion, the hardest part of using Powercroco's 6 wire technique.

To avoid this, you should not complete the windings on tooth_2. Instead, you only do one TURN on tooth_2 first. Then move the coil outwards, and do the first turn on Tooth_1. In this way, the first turn on Tooth_1 will be much easier, and can be clamped flat easily. Then you continue winding one turn each time, on each tooth.

In this way, every winding are basically done in exact same condition: same room for winding. So, if you can do tooth_1 and tooth_2, you surely can do the rest tooth without much problem.
This is what i mean by symmetry. But this is only necessary for extra-compact thick wire winding, or maximized copper fill.
If winding single 1.9mm on this 7050, i believe it is not at all necessary, to wind like this.

The downside of this technique, is you are going to wind, with all the wires hooked up on the stator...can be kinda difficult to manipulate: very heavy, and messy sometimes

The last turn of my "2.58mm, 5+5T", the turn as the 2nd layer, is my invention, i believe. That is the turn, "inserted" in the slot: all loose ends will be packed inside by one another. In this way, all coils are tightly packed and secured by coils themselves.
same technique is used in a 5035 "1.9mm, 6+6T"

Cool wind Mo!

[powercroco]

modisc,

your single tooth winding is very familiar to ABC.... in "1-2-3-short" winding mode.
also in this mode you wind only a single tooth before you have to jump over the next 2 coils.

some experiences from this about your problem (all thought for single tooth winding).
you also could use a short wire (only one arc) as dummie for turn 1 on tooth 12.
then wind the complete coil on t1.
so you save the needed room also and it's imo easier to handle.

also bend the turn number 1 of t2 outside, use the pliers and a dummie a then insert it on the stator.
you also can clamp the backside of turn number 1 with the pliers outside to the tooth and then move it to the center.
also you could use the same mode for the first 2 or 3 turns, depending on the insulation of your stator.

try it....

about the other thing:
number of polepairs is the right option.
field rpm divided by 5 gives the motor rpm for 10P motor.

[modisc731]

Originally Posted by powercroco

modisc,

your single tooth winding is very familiar to ABC.... in "1-2-3-short" winding mode.
also in this mode you wind only a single tooth before you have to jump over the next 2 coils.

some experiences from this about your problem (all thought for single tooth winding).
you also could use a short wire (only one arc) as dummie for turn 1 on tooth 12.
then wind the complete coil on t1.
so you save the needed room also and it's imo easier to handle.

also bend the turn number 1 of t2 outside, use the pliers and a dummie a then insert it on the stator.
you also can clamp the backside of turn number 1 with the pliers outside to the tooth and then move it to the center.
also you could use the same mode for the first 2 or 3 turns, depending on the insulation of your stator.

try it....

about the other thing:
number of polepairs is the right option.
field rpm divided by 5 gives the motor rpm for 10P motor.

I used to think STSD is best for all, but for dLRK, experience told me that your 6-wire technique is the best option. In my opinion, STSD is still best for ABCXn type of winding though, mainly because the loose ends can be arranged nicely at the bottom.

if i understand you correctly: you are using some short wires to "preoccupy the room", to make the actual turns away from the bottom of the slot: therefore, easier to clamp.
If so, i am actually doing exactly the same as you described. instead, i did not use :short wire" to preoccupy the room. The reason "i wind only 1/2, 1/2.....1/2 on the 12 tooth first, then adding one T on each tooth, then go on...", is that i could push the turns away from bottom of the slots first, then wind and clamp.
But i believe your way could make the winding process much simplier

It is good to know that electric RPM = number of pairs of poles X actual RPM. Then 2.58mm should have a limit of ~30k rpm, to ignore skin effect. if not considering control frequency from PWM signals.

Does PWM signals really count in skin effect? i should do some research...

[powercroco]

Does PWM signals really count in skin effect? i should do some research...

in fact in our motor sizes there is a win in efficiency by increasing copperfill up to 1,8mm wire diameter, 1,9mm is break even, 2,0mm or more brings more losses than wins.
using 2x1,4 II instead of 2,0 will give the better motor, as far it is just tested.

but I can't say, what is the detailed reason for the found result!
in 100% we discuss the eddy currents into the copper wires, for part load with high swithcing frequencies we think about something just like the skin effect.

"preoccupy the room" is xact the wording I was searching for

[modisc731]

Originally Posted by powercroco

in fact in our motor sizes there is a win in efficiency by increasing copperfill up to 1,8mm wire diameter, 1,9mm is break even, 2,0mm or more brings more losses than wins.
using 2x1,4 II instead of 2,0 will give the better motor, as far it is just tested.

but I can't say, what is the detailed reason for the found result!
in 100% we discuss the eddy currents into the copper wires, for part load with high swithcing frequencies we think about something just like the skin effect.

"preoccupy the room" is xact the wording I was searching for

it is easy to understand one another, if we share same thoughts.

even at max throttle, 2X1.4 is still better than single 2.0?

[powercroco]

yes.

[modisc731]

a few aluminium part inside the rotor has to be sander off, or it will scratch the wires...





Classic Powercroco's arrangement of wires.


Dirty marker pan...
http://photo1.5imxbbs.com/forum/2015...fd5ftaixz5.jpg

Modified rotor part.

[powercroco]

crazy!