Author Topic: Coil Testing  (Read 4072 times)

0 Members and 1 Guest are viewing this topic.

Boondocker

  • Full Member
  • ***
  • Posts: 193
Coil Testing
« on: January 24, 2007, 11:35:33 PM »



With much interest I've been following the posts on testing coils and load matching the alternator/prop.  This is the 1st attempt in applying the information to my project.  Thanks to all who contribute to this form. Please review and point out errors, omissions and any improvements that can be made.  Boy, I'm having fun learning this stuff.


Here is background information on the project/experiment:


Magnets:        24 - ¼" wedge (Windstuffnow)

Poles:          12

Rotor size:      12"

Magnet Wire       14 awg

Stator:          9 coils, 3 phase wired star

System Volts      24


Below is a graph plotting data from of the 1st test coil of 78 turns.  Open volts were calculated from multimeter rms readings.  Open volts= (volts rms)3*1.41*1.73-1.4





Next, 6 turns were removed off the coil.  This time the air gap between the magnets were varied and data was collected.  The gap between the magnets were 0.843",0.770" and 0.715".





The resistance of the test coil was first estimated based on weight.  14 awg wire is 0.203 ohms/lb and the 72 turn coil weighs 301.7g.  (301.7 g / 453.6 g/lb) * 0.203 ohms/lb = 0.135 ohms.    The resistance was also checked by voltage drop across the coil on a measured load.  When the coil was at room temperature a resistance of .137 ohms was obtained.    Then the coil was heated-up by shorting the leads with the battery charger for a minute or so, until the coil was fairly warm to the touch.  The resistance of the warm coil increased to .180 ohms.





Now for the complicated mysterious part, "Estimating" alternator performance and then matching a prop.  Here are some definitions and assumptions used.


Input power          = Output + Losses

Stator Loss           =  Current^2 * Winding R

Star Winding R      = Sum of the resistance of 2 phases   

Output                 = [(Open dcv - Charge dcv)(Winding R * cf)] *  Charge dcv

Cf (Correction factor  )= 1.3  (based on Flux's experience)


Prop Cp at different TSR,

 .12        4

.20        5

.35        6

.4        7


Taking the above assumptions and creating a spreadsheet produced the next graph.





The Excel spreadsheet is uploaded into my files for review.


The cells in blue represent data measured off the test coils.  The cells in yellow are set-up as variable inputs.


I have no idea how correct and/or accurate the spreadsheet came out.  Hopefully, with some of your input it can be improved.   The spreadsheet has been educational for me to see graphically the effect changing a variable has.  Most notable, how adding resistance or opening the magnet gap can move the system out of a slow speed stall.  Initially I had a hard time grasping this concept, which is mentioned in several post.


Here is some explanation on the spreadsheet:


-Test coil vac      based on test coil regression analysis (acv/rpm input)

-Open vdc           based on 3 coils/phase in star, rms conversion, and

                            then subtracts off rectifier and line loss.

-Output amp       (Open vdc - Charging vdc)
resistance

-Battery Watts    amps  
charging vdc

-System Heat        (amp^2)* corrected resistance   

-Stator Input        Battery + System Heat


One of the short coming of the spreadsheet it does not factor in resistance will be increasing as the stator heats up.


Questions so far:


1. Is the Cp vs TSR assumption reasonable for a 2.4m prop?  Does anyone have any

   real world data?


2. Does the rectifier and line losses need to be added independently as part of the

   Stator Input, or are they accounted for in the Open vdc calculation?


3. If a boost 12vdc converter is used, what is recommended for the cut-in rpm

   on the main rectifier?


Thanks,


Boondocker

« Last Edit: January 24, 2007, 11:35:33 PM by (unknown) »

willib

  • SuperHero Member
  • ******
  • Posts: 2414
  • Country: us
Re: Coil Testing
« Reply #1 on: January 24, 2007, 04:47:00 PM »
in the first graph , what was the actual coil voltage obtained? at the RPMs stated?
« Last Edit: January 24, 2007, 04:47:00 PM by willib »
Carpe Ventum (Seize the Wind)

Boondocker

  • Full Member
  • ***
  • Posts: 193
Re: Coil Testing
« Reply #2 on: January 24, 2007, 05:07:27 PM »
Multimeter        Est

RMS acv    Rpm    Multi


  1.     213    35.4
  2. 4    200    31.0
  3.     180    28.0
  4. 3    150    22.9
  5.     140    20.7
  6.     93    13.3


« Last Edit: January 24, 2007, 05:07:27 PM by Boondocker »

willib

  • SuperHero Member
  • ******
  • Posts: 2414
  • Country: us
Re: Coil Testing
« Reply #3 on: January 24, 2007, 06:43:12 PM »
for my machine (the current mini gen) my target voltage was 14.5 V , chosen as a cut in.

i have 4 coils per phase so 14.5/4 = 3.625V per coil divided by root 2 gave me 2.56V ac per coil as a target , per coil .

« Last Edit: January 24, 2007, 06:43:12 PM by willib »
Carpe Ventum (Seize the Wind)

Tom in NH

  • Full Member
  • ***
  • Posts: 191
Re: Coil Testing
« Reply #4 on: January 24, 2007, 08:15:41 PM »
Nice work, Boondocker. It looks like you've been very careful and thorough. Thanks for sharing it. --tom
« Last Edit: January 24, 2007, 08:15:41 PM by Tom in NH »

Flux

  • Super Hero Member Plus
  • *******
  • Posts: 6275
Re: Coil Testing
« Reply #5 on: January 25, 2007, 01:50:36 AM »
This is some of the best work that I have seen. I haven't looked at your spreadsheet ( probably won't be able to understand it anyway, but I will look when I have time).


The totally unknown issue is the prop Cp.


I think your figure of .4 is too high, my experiments suggest that you can get .35 in lower winds with tsr about 6 to 7. Things start to fall rapidly below tsr5.


In higher winds the optimum tsr seems to fall so your peak will be nearer tsr6. Below tsr 4 you hit stall quite sharply.


I did start some tests on this a while ago but the test site was too turbulent for the results to mean anything, There should be a better site coming up but may be a while.


If you want to use a boost converter then cut in can be 200 to 220 rpm and you should be able to boost from about 150 rpm.


Without the converter cut in at 150 will kill your higher wind performance. In this case cut in 180 to 200 would give a better overall result. In desperately poor wind areas drop to 170 and accept less in higher winds.


keep up the good work.

Flux

« Last Edit: January 25, 2007, 01:50:36 AM by Flux »

Flux

  • Super Hero Member Plus
  • *******
  • Posts: 6275
Re: Coil Testing
« Reply #6 on: January 25, 2007, 10:12:54 AM »
I have had a bit more time to look at this now.


I assume your .715" gap is the smallest that will take your coil thickness. I wouldn't use a larger gap with those thin magnets.


Now your 72 turns is a bit on the slow side for direct operation at 24v but near.


If you want to run with a boost converter then if you raise cut in to 220 rpm you can manage with 50 turns. You seem to have got 76 turns in originally so you can use a fair bit thicker wire for 50 turns. In view of the large air gap and close magnet spacing, I would take a chance on smaller holes in the coils ( width wise, same radial size).This could get you even thicker wire in.


You can probably end up with half the resistance of your present winding which will help your efficiency a lot. With the higher cut in you will not likely stall so you can run a much higher full load efficiency with much less stator heat and the same output at significantly lower wind speed.


At the low end you have the converter loss but better prop matching so you shouldn't be any worse.


The change over point for maximum energy capture depends on local wind conditions, in a really good wind area I would take main cut in to over 220 rpm but that is probably a good compromise.


Flux

« Last Edit: January 25, 2007, 10:12:54 AM by Flux »

SparWeb

  • Global Moderator
  • Super Hero Member Plus
  • *****
  • Posts: 5452
  • Country: ca
    • Wind Turbine Project Field Notes
Re: Coil Testing
« Reply #7 on: January 25, 2007, 01:11:31 PM »
Boondocker,


That's great work so far.  Looking at your power in/out graph, the prop's collected power does not exceed the power required by the stator.  This may not be a problem; the mill just runs slower than you predict, I bet.  I played with your numbers a bit - notice that the curves line up better if you assume 28V charging voltage instead of 25V.


I know of the V(open) - V(charging) equation that the guys here use, but I still view it with suspicion... It's probably close enough for our purposes, though.


As you have already noticed, there are a LOT of variables that affect your results.  I, too, am finding it difficult to engineer a theoretical model that points to any conclusions.


These tests are a lot of fun, though, aren't they!




« Last Edit: January 25, 2007, 01:11:31 PM by SparWeb »
No one believes the theory except the one who developed it. Everyone believes the experiment except the one who ran it.
System spec: 135w BP multicrystalline panels, Xantrex C40, DIY 10ft (3m) diameter wind turbine, Tri-Star TS60, 800AH x 24V AGM Battery, Xantrex SW4024
www.sparweb.ca

Boondocker

  • Full Member
  • ***
  • Posts: 193
Re: Coil Testing
« Reply #8 on: January 25, 2007, 05:36:39 PM »
Thanks for the comments.


Yes, a 0.715" magnet gap is as narrow as they can be placed with the test coil set in plywood.  The rotor plates are not exactly flat and some float room is required.


I'll aim for 220 rpm cut-in on the "boost stator".  I believe 50 turns, 2 in hand, 15 awg wire will fit.  Though it will be tight.  Would a single 12 awg wire be a better choice?


Boondocker

« Last Edit: January 25, 2007, 05:36:39 PM by Boondocker »

Flux

  • Super Hero Member Plus
  • *******
  • Posts: 6275
Re: Coil Testing
« Reply #9 on: January 26, 2007, 02:15:22 AM »
I think you will find #12 a pain to wind and the stacking factor is likely to be better with 2 in hand #15.


Flux

« Last Edit: January 26, 2007, 02:15:22 AM by Flux »

Boondocker

  • Full Member
  • ***
  • Posts: 193
Re: Coil Testing
« Reply #10 on: January 28, 2007, 12:27:00 PM »
Made a new test coil, 51 turns two-in-hand out of #14 wire.  The resistance dropped considerably.  Measuring a single coil the highest reading was .068 ohms.  With this coil the trick of heating by shorting it out across the battery charger doesn't work.  The charger kicks out; not enough resistance.


The size of the coil will not be an issue and fits nicely in the stator mold.


Below are the new graphs:







« Last Edit: January 28, 2007, 12:27:00 PM by Boondocker »

TomW

  • Super Hero Member Plus
  • *******
  • Posts: 5130
  • Country: us
Ouch.....
« Reply #11 on: January 28, 2007, 12:59:27 PM »
boondocker;


That sure is a pair of huge files for those charts.


T

« Last Edit: January 28, 2007, 12:59:27 PM by TomW »

Boondocker

  • Full Member
  • ***
  • Posts: 193
Re: Ouch.....
« Reply #12 on: January 28, 2007, 03:21:13 PM »
Sorry about that, this should be better.







« Last Edit: January 28, 2007, 03:21:13 PM by Boondocker »

Boondocker

  • Full Member
  • ***
  • Posts: 193
Re: Ouch.....
« Reply #13 on: June 28, 2007, 04:15:53 PM »
Need to make a correction the coils are wound from 2 in-hand No. 15 wire not 14.
« Last Edit: June 28, 2007, 04:15:53 PM by Boondocker »