Winding a strong electromagnet

[Xaverius]

Amp turns per watt:

Some time ago I ran some numbers to try to find the "efficiency" of an EM winding.  How to get the most AT (ampturns) per watt.

I used the awesome coil calculator to determine wire length, # turns and ohms at:
http://www.coilgun.info/mark2/inductorsim.htm

The attached picture is of an excel spreadsheet showing these.
I'll run some more in the near future - and post the thing as a .pdf (I hate downloading an unknown .xls)

This is a winding comparison using the same winding length - 1.5" (not wire length) and core.  The actual gauss/tesla would depend on the permeability/width of the core?

Note the appreciable difference between the 1st and 2nd winds.  #2 is MUCH more efficient!

Comments?  Additional math to apply? etc?

Sorry, I can't seem to make the coilgun calculator work.   I don't know what the problem is.  A few observations on the excel spreadsheet, for a 1/2 inch diameter bolt the number of turns for 125 feet should be approximately 960.  Also, 10000 gauss = 1 Tesla, therefore .1T should be 1000 gauss. 

Gauss/Tesla is a function of H (magnetic field strength, Ampere-turns/meter) multiplied by permeability.  The width of the core determines the cross-sectional area which determines the amount of flux, a wider core produces more flux.

[capthook]

I now would take double the length 740 ft ( not double the AT) of this #2 calculation and make two separate coils a 370 ft and put these two coils in parallel. Now you have the same input-wattage and much more Tesla.

Kator

Interesting idea.  I'm not sure how to figure what the possible boost might be.  Is it something like:
The amps is half - but you are now influencing two cores - so the relative boost of each core is 50%, but you now have two cores influenced so you might see a total increase of 25% because you now have 2 magnets and are at the low end of permeability curve for each?  Not sure how to state what I'm thinking.  Could you explain your thinking a little more?

- -
Hoptoad!  Thanks for the link!  I've seen some stuff here and Utube that shows something similar to the effect in that link.  Can't remember the threads or links or titles etc.  Interesting idea to ponder.....

- -
Xaverius: yea - guess I spaced on the Gauss=Tesla conversion - used 1,000 instead of 10,000 

Gauss/Tesla is a function of H (magnetic field strength, Ampere-turns/meter) multiplied by permeability.  The width of the core determines the cross-sectional area which determines the amount of flux, a wider core produces more flux.

The actual gauss/tesla would depend on the permeability/width of the core

Yes - the Gauss listed is just a reference point - and for an absolute - would in fact be dependant on core permeability/width.  I guess a grade 2 hex bolt is somewhere in the range of 50-100u. I could convert the AT to AT/meter (show me please) and I could (throw a dart at the wall and) pick a number between 50-100 for a multiplier.    100 it is! (of course the goal is what - 2,000u electrical steel?)(and how does the permeability curve affect the calculations? Isn't 1/2 watt going to be way different than 10 watts?)(and which permeablity figure would you use? A materials specs is listed as intial, max, or the other one (can't remember it))
Now you see why it's listed as relative - you are the math guy 
The calculation I gave in the chart was a simplied one presented by a reed switch manufacturer as a way to give an approximation for AT/gauss conversion needed to activate the switch.
So in line 1 of the chart - please show me the calculation for converting 298 AT to the projected gauss. (tx)

I re-checked the # of turns for 125' ( I'm assuming you are ref. line 1 - the 22AWG).
Same result (actually 616 instead of 620)

1" = 25.4 mm
1' = .3048 m

1. Choose wire size: 22
2. input inner dimension (1/2" core): .5" x 25.4mm = 13mm
3. input coil length: 1.5" x 25.4 = 38mm
4. convert 125' to meters = 125 x .3048 = 38m
5. adjust the outer dimension slider until the wire length line shows about 38m wire length
6. observe: # turns, # of layers, resistance, and coil outer dimensions etc.

[capthook]

X - I maybe misunderstood you comment of can't get coil calculator to work?  It doesn't load?

It is a java applet - you may need to update your java.  The website says this:

"This program requires Java Runtime Environment (JRE) 1.4 or above. Please visit www.java.com/en/download/index.jsp to download and install the JRE."

[i_ron]

X - I maybe misunderstood you comment of can't get coil calculator to work?  It doesn't load?

It is a java applet - you may need to update your java.  The website says this:

"This program requires Java Runtime Environment (JRE) 1.4 or above. Please visit www.java.com/en/download/index.jsp to download and install the JRE."

The applet is very handy, well worth the trouble.... but... installing java also installs the Yahoo tool bar!
in Firefox go to "tools", "add-ons" and delete this screen waster, if you so desire... lol

Ron

[Kator01]

Capthook,

no, no two cores. Your calculation was triggering my mind of just combining what was said earlier - by xaverius reply # 22 - you wind one coil on one core,  but you first wind 370 ft on the core than take another sperate 370 ft and wind this above the first coil. Now you connect both the beginnings and endings of each coil. Now you have these two cores in parallel . But this would not double the AT, so forget this idea. It is just a means of reducing the resistance and inductance.You are right , it only will make sense if it is put on two seperate cores.  I was mistaken - sorry - too long computer-sessions until early in the morning.

Regards

Kator01