Inductance What is it what does it do? Collosus has 8uH!

[Arlo1]

3turns 2 in parallel

031_renamed_30453.jpg (77.65 KiB) Viewed 707 times

OK SOOOOOOOOOOOOOOOOOOOOOOOOOOOO BIIIIIIIIIIGGGGGGGGGGG NEEEEEEEEEEWWWWWWWWWS. Collossus has such low inductance it midas well not have any (well ok almost) I got the first cheep inductance meter today and here is some findings.... OH BTW you guys rewinding motors that are hard to control you would be smarter with one strand of a thicker wire then multiple strands of thinner wire!!! I found way more inductance with 3 turns of a single strand vs 3 turns with 2 strands parellel!

[Lebowski]

OK SOOOOOOOOOOOOOOOOOOOOOOOOOOOO BIIIIIIIIIIGGGGGGGGGGG NEEEEEEEEEEWWWWWWWWWS. Collossus has such low inductance it midas well not have any (well ok almost) I got the first cheep inductance meter today and here is some findings.... OH BTW you guys rewinding motors that are hard to control you would be smarter with one strand of a thicker wire then multiple strands of thinner wire!!! I found way more inductance with 3 turns of a single strand vs 3 parallel strands!

This is probably due to the L being proportional to (windings)^2 ....

[Arlo1]

OK SOOOOOOOOOOOOOOOOOOOOOOOOOOOO BIIIIIIIIIIGGGGGGGGGGG NEEEEEEEEEEWWWWWWWWWS. Collossus has such low inductance it midas well not have any (well ok almost) I got the first cheep inductance meter today and here is some findings.... OH BTW you guys rewinding motors that are hard to control you would be smarter with one strand of a thicker wire then multiple strands of thinner wire!!! I found way more inductance with 3 turns of a single strand vs 3 parallel strands!

This is probably due to the L being proportional to (windings)^2 ....

Sorry I went back and corrected what I said I tested 3turn x 2 parrel and got less inductance then 3turns with just one strand!

[Lebowski]

OK SOOOOOOOOOOOOOOOOOOOOOOOOOOOO BIIIIIIIIIIGGGGGGGGGGG NEEEEEEEEEEWWWWWWWWWS. Collossus has such low inductance it midas well not have any (well ok almost) I got the first cheep inductance meter today and here is some findings.... OH BTW you guys rewinding motors that are hard to control you would be smarter with one strand of a thicker wire then multiple strands of thinner wire!!! I found way more inductance with 3 turns of a single strand vs 3 parallel strands!

This is probably due to the L being proportional to (windings)^2 ....

Sorry I went back and corrected what I said I tested 3turn x 2 parrel and got less inductance then 3turns with just one strand!

Oeh now you're in highly dangerous territory. At work here we got a high precision RLC meter

http://www.testequipmentdepot.com/usede ... /4284a.htm

With this thing you can set the frequency at which it measures the inductance. When set to inductance
it also measures the series resistance of the inductor.

What I've noticed is that for higher frequencies the series resistance of a thin-wired inductor goes up due
to the skin-effect.

What's probably happening is that your inductance meter is running at a pretty high frequency and that it
mistakes the higher series resistance of the single strand wire for extra inductance...

[Arlo1]

Oeh now you're in highly dangerous territory. At work here we got a high precision RLC meter

http://www.testequipmentdepot.com/usede ... /4284a.htm

With this thing you can set the frequency at which it measures the inductance. When set to inductance
it also measures the series resistance of the inductor.

What I've noticed is that for higher frequencies the series resistance of a thin-wired inductor goes up due
to the skin-effect.

What's probably happening is that your inductance meter is running at a pretty high frequency and that it
mistakes the higher series resistance of the single strand wire for extra inductance...

ok??? So then collossus is not low inductance? I have another Low inductance meter coming so I will compare the two. But what I reserched is multiple strands do lower inductance. So...... I will try more tests with two inductors parallel and it should be 1/2 of one of the inductors.

[Arlo1]

IN fact I tell you what I will wind it in such a way that I test the same lengh of wire with 3 turns with one and two strands.... So I will still have the two peices of wire hooked to the meter with one set of turns or 2 parallel sets of turns.

[Lebowski]

IN fact I tell you what I will wind it in such a way that I test the same lengh of wire with 3 turns with one and two strands.... So I will still have the two peices of wire hooked to the meter with one set of turns or 2 parallel sets of turns.

What I would do is make two measurements, both with the same amount of windings but one with a single strand of thick wire and one with a single strand of thin wire.
Since inductance is dependent on winding diameter and the length of the windings (diameter and length on this page:

http://www.66pacific.com/calculators/coil_calc.aspx

) make sure both are equal for the thin and thick wire coil !

[jdb]

Do any of the COTS controllers expose the control loop gains? What controller has emerged as the Survivor for this motor so far? 8uH is really low... but it might be possible to modify an existing controller for the purpose if you can get access to its current control loop gains.

[Arlo1]

IN fact I tell you what I will wind it in such a way that I test the same lengh of wire with 3 turns with one and two strands.... So I will still have the two peices of wire hooked to the meter with one set of turns or 2 parallel sets of turns.

What I would do is make two measurements, both with the same amount of windings but one with a single strand of thick wire and one with a single strand of thin wire.
Since inductance is dependent on winding diameter and the length of the windings (diameter and length on this page:

http://www.66pacific.com/calculators/coil_calc.aspx

) make sure both are equal for the thin and thick wire coil !

What I figured the difference in inductance with thicker and thinner wire was caused by is the simple fact you can't fit the same amount of turns in a given area with thicker wire. I will do 3 turns of 14 awg then 3 of 18awg 3 of 20 awg and 3 of 10 awg on one tooth of collossus to see what the differences are and to be fair I will try to space them all the same as the 10 awg!

[rhitee05]

It would surprise me quite a bit if the skin effect is responsible. A google search on your meter suggests that the L measurement in the "standard" range is probably at 500 kHz. The skin effect is noticeable at this frequency in the larger wire - the AC resistance of 10 AWG at 500 kHz is about 10x higher than the DC resistance... but that still puts it only at around 23 milliohm/meter. The impedance of 1 uH at 500 kHz is about 3 ohms. If the meter is measuring total impedance, then the resistance is still less than 1%.

More likely is just that the geometry of your windings were slightly different and the change in inductance due to the different wire diameters. Even if the geometries were identical there will still be a difference in inductance with different wire diameters.

[Arlo1]

More likely is just that the geometry of your windings were slightly different and the change in inductance due to the different wire diameters. Even if the geometries were identical there will still be a difference in inductance with different wire diameters.

I will do a very tight winding with and without parallel strands but the wire is 100% the same it was a continious peice of 14 awg from the same 900+ foot roll You can watch the inductance go up and down as you wind something. The video I posted at the start of the thread shows how two parallel inductors lowers the inductance as the inductors are close to each other. Parallel inductors will cut the inductance by 1/2 if they are the same value. So I do understand it will be different on the same tooth.

But this leads me to think if you do need to do parallel strands you can wind them one at a time so only one parallel strand is close to its parallel partner and this will alow the other strands to build there feild next to the strands in series with them self!!! I will test this asap!

[rhitee05]

If you do a true multi-strand winding, where all the strands are wound together, then the inductance of the multi-strand winding will be approximately the same as a similar winding using thicker wire. Not exactly the same, but very similar. This seems like parallel inductors, which would reduce the total inductance by 1/N, but it is not. The reason is that the inductors are very tightly coupled together so the mutual inductance between them makes up for the parallel combination.

If you do two separate windings on the same tooth and then connect them in parallel, the inductance will be lower than the multi-strand winding because the mutual inductance will be lower. It might not be a lot lower because the tooth will still keep the coupling fairly high, but it will be lower. The lower the coupling gets, the closer to straight parallel inductors it will become.

[Arlo1]

So my second cheep L meter cam in today and I decided to do some more testing. I found 2 strands next to each other is about the same inductance as 5 turns of one strand then 5 of a second strand and terminated in parallel on each end. But I also found 10 turns is not 4 times the inductance of 5 turns times 2 in parallel ITS MORE..... This makes me wonder is it worth upping the battery voltage to get the same max rpm and wind it with as many turns in series possible? What is the relationship of kv of a brushless motor vs winds?

[Arlo1]

So I compared the inductance at 8uh to 4.55 times that and 2 times the voltage. My reason was I think 2 times the turns means 1/2 the rpm per volt but two times the turns is 4.55 times as much inductance. I do realize the resistance will go up as well when you have 2 times the turns but I can minimize that as much as possible.

[Arlo1]

Ok so this makes me think a motor with 100 kv run at 50 volts would be easier on a controller if you rewound to 50 kv and run at 100 volts by taking the 2 parallel strands and just making them one continious strand eg change it from 5 turns times 2 in parallel to 10 turns in series.

[Arlo1]

OK so Motor A is stock collossus Motor B is a guess at what I would end up with at the same top speed double the voltage and double the turns with only one strand in series with 4 times the resistance because its just one strand. Motor C is my X5

[rhitee05]

Inductance varies with the number of turns squared. Your results pretty much match what would be expected, that twice the turns will be 4x the inductance.

[Arlo1]

Inductance varies with the number of turns squared. Your results pretty much match what would be expected, that twice the turns will be 4x the inductance.

Now what about the rpm per volt?

[rhitee05]

I'm not entirely certain, but I think there's a linear relationship between turns and kV.

[Arlo1]

I'm not entirely certain, but I think there's a linear relationship between turns and kV.

Thats what I thought, so might have a way to help save a controller!

[Arlo1]

So I just realized I never took a reading of the hs35 motor assembled the inductance reading was with the magnet ring off the motor. So here is the second set of readings.

[Arlo1]

Ok guys so I have tried to get as much inductance as I can with different windings ect. I tried taking two strands and clamping one end in the drill and the other in the vice to twist them together with not much increase in inductance. The amount of copper fill you would loose doing that would out weigh the gains!

Then I woke up this morning and i got it! I will wind each tooth with the most amount of continuous turns of the same strand possible. It will either be 8 turns continuous of my 10.5 awg winding wire or 9 turns of the 14 awg wire. Then in series with the next two teeth on the same side of the motor like normal but instead of going around to the other side of the motor for the other three teeth in series I will go out of the case with them. Then wind the three teeth on the other side separately (looking at one phase) SO when finished I will have 1/2 the stator wound separately from the other then I can hook the wires up in series or parallel on the out side of the motor and use a switch if needed. But Even in parallel I will have the same rpm per volt and 2 times the inductance! SO my 8uH motor with 75kv will now be 16uH with 75kv and if that's not enough I will use the series parallel switch set up for 32uH and 37.5kv for the first gear (starts) and then switch to 16uH and 75kv second gear(hi speed)

[Arlo1]

So here is motor A stock collossus
Motor B rewound and connected in parallel for same KV
And motor C wound in series for 1/2 the KV You can see even motor B will be so so so much easier on a controller!

[Thud]

Here is a question for the motor uber guru's here.

Arlo has taken to different winds in an attempt to keep the inductance inside the motor as best he can.

I have taken another path. I have re-configured the pole count to elliminate the counter winds inside the motor.
Look inside this can. I have gone from 7prs to 8prs & now have to wind ABCABCABCABC as opposed to AabBCcaABbcC.
the stock .5 x 2 x .125"t n42's are a perfect fit with .030" spacing


Since I have no lab equipment capable of measuring inductance I would like to here the positives & negatives to such configurations. The only thing i am aware of is the added magnet count will start showing iron loss a little earlyer in the rpm range.

Biff, any thoughts?
Thanks
T

[Arlo1]

Here is a question for the motor uber guru's here.

Arlo has taken to different winds in an attempt to keep the inductance inside the motor as best he can.

I have taken another path. I have re-configured the pole count to elliminate the counter winds inside the motor.
Look inside this can. I have gone from 7prs to 8prs & now have to wind ABCABCABCABC as opposed to AabBCcaABbcC.
the stock .5 x 2 x .125"t n42's are a perfect fit with .030" spacing


Since I have no lab equipment capable of measuring inductance I would like to here the positives & negatives to such configurations. The only thing i am aware of is the added magnet count will start showing iron loss a little earlyer in the rpm range.

Biff, any thoughts?
Thanks
T

Thud I played with leaving the counter wind out and it does not hurt inductance. If you have 4 turns on one tooth then 4 the opposite way on the next tooth and 4 turns the first way on the third tooth in a roe it all adds up so if each tooth gives 5uH then all three will add to 15uH I played leaving the second tooth out and just trying the first and last of the 3 and in that case it would be 10uH!