mathematically calculating k
Ben Solon, Wed Oct 24 2012, 08:41PMdoes anyone know how to calculate the k of two coils(pri and sec) if they have weird dimensions? all i've found are formulas for two solenoids of the same radius. i need to do are two helical coils(maybe one angled helix) of different diameters. all i need are formula's. thanks.
Re: mathematically calculating k
Dr. Brownout, Wed Oct 24 2012, 10:10PM
Have you checked this
Not sure if its what your looking for but at least it gives you a starting point.
Dr. Brownout, Wed Oct 24 2012, 10:10PM
Have you checked this
Not sure if its what your looking for but at least it gives you a starting point.
Re: mathematically calculating k
Ben Solon, Wed Oct 24 2012, 11:10PM
No, I've found plenty of these types of sites. i'm trying to calculate the k. no measurements other than dimensions. the main problem is that no formula I've found so far factors in coil positions relative to each other and coils with different dimensions.
Ben Solon, Wed Oct 24 2012, 11:10PM
No, I've found plenty of these types of sites. i'm trying to calculate the k. no measurements other than dimensions. the main problem is that no formula I've found so far factors in coil positions relative to each other and coils with different dimensions.
Re: mathematically calculating k
Hazmatt_(The Underdog), Thu Oct 25 2012, 12:40AM
you'll need some books. I was interested in this too, but I haven't had much time to read.
Terman, Radio Engineering Handbook
Welsby, The Theory and design of inductance coils
Grover, inductance calculations
and theres probably a few more out there
Hazmatt_(The Underdog), Thu Oct 25 2012, 12:40AM
you'll need some books. I was interested in this too, but I haven't had much time to read.
Terman, Radio Engineering Handbook
Welsby, The Theory and design of inductance coils
Grover, inductance calculations
and theres probably a few more out there
Re: mathematically calculating k
Ben Solon, Thu Oct 25 2012, 01:00AM
Thanks! Guess I'll have to get some of those on amazon or something. I'm really surprised that there's nothing around the internet for this sort of thing. Sure there's javatc, but no equations. I even tried going through javatc's source to get at the javascript functions, but they're located elsewhere.
Ben Solon, Thu Oct 25 2012, 01:00AM
Thanks! Guess I'll have to get some of those on amazon or something. I'm really surprised that there's nothing around the internet for this sort of thing. Sure there's javatc, but no equations. I even tried going through javatc's source to get at the javascript functions, but they're located elsewhere.
Re: mathematically calculating k
Newton Brawn, Thu Oct 25 2012, 03:39AM
Ben:
Try these :
Inca (ICCDTR) program
Installation - The program works in a PC under Windows.
Just copy the files Inca.exe and inca.xpl to the same directory.
available at
Author : Antonio Carlos M. de Queiroz - 21/12/2003
regards
Newton
Newton Brawn, Thu Oct 25 2012, 03:39AM
Ben:
Try these :
Inca (ICCDTR) program
Installation - The program works in a PC under Windows.
Just copy the files Inca.exe and inca.xpl to the same directory.
available at
Author : Antonio Carlos M. de Queiroz - 21/12/2003
regards
Newton
Re: mathematically calculating k
dex, Thu Oct 25 2012, 09:29AM
IMHO,there are no closed formulae taking in account all parameters of the geometry between two coils.To calculate mutual inductance precisely enough,you must use numerical methods.
dex, Thu Oct 25 2012, 09:29AM
IMHO,there are no closed formulae taking in account all parameters of the geometry between two coils.To calculate mutual inductance precisely enough,you must use numerical methods.
Re: mathematically calculating k
Steve Conner, Thu Oct 25 2012, 09:39AM
Yes, you can use FANTC and maybe JavaTC too.
Steve Conner, Thu Oct 25 2012, 09:39AM
Yes, you can use FANTC and maybe JavaTC too.
Re: mathematically calculating k
Ben Solon, Thu Oct 25 2012, 11:32PM
but what are the driving calculations behind javatc and inca?(just need the formula for a code function on a uC) it can obviously be done because these programs exist, but it seems the equations are missing from everywhere.
Ben Solon, Thu Oct 25 2012, 11:32PM
but what are the driving calculations behind javatc and inca?(just need the formula for a code function on a uC) it can obviously be done because these programs exist, but it seems the equations are missing from everywhere.
Re: mathematically calculating k
dex, Fri Oct 26 2012, 06:59AM
i guess they are based on systems of elliptic integrals and various equivalent threatments of them.
dex, Fri Oct 26 2012, 06:59AM
ben123324 wrote ...
but what are the driving calculations behind javatc and inca?(just need the formula for a code function on a uC) it can obviously be done because these programs exist, but it seems the equations are missing from everywhere.
but what are the driving calculations behind javatc and inca?(just need the formula for a code function on a uC) it can obviously be done because these programs exist, but it seems the equations are missing from everywhere.
i guess they are based on systems of elliptic integrals and various equivalent threatments of them.
Re: mathematically calculating k
Steve Conner, Fri Oct 26 2012, 08:08AM
The workings inside FanTC/JavaTC are explained in Paul Nicholson's TSSP papers. They are numerical methods, not "formulas". (I got through an entire EE math curriculum without understanding the difference, but I eventually got it.) You don't want to be implementing them on a microcontroller.
Reframe the problem: why do you want a uC to calculate k in the first place? How will the user input the coil geometry to the uC?
Say you were designing a motor-driven variable coupling system, and you wanted the user to set k directly. I would use JavaTC to calculate the coupling in lots of little steps over the travel of the mechanism, then use that data to make a lookup table for the uC.
Steve Conner, Fri Oct 26 2012, 08:08AM
The workings inside FanTC/JavaTC are explained in Paul Nicholson's TSSP papers. They are numerical methods, not "formulas". (I got through an entire EE math curriculum without understanding the difference, but I eventually got it.) You don't want to be implementing them on a microcontroller.
Reframe the problem: why do you want a uC to calculate k in the first place? How will the user input the coil geometry to the uC?
Say you were designing a motor-driven variable coupling system, and you wanted the user to set k directly. I would use JavaTC to calculate the coupling in lots of little steps over the travel of the mechanism, then use that data to make a lookup table for the uC.
Re: mathematically calculating k
Ben Solon, Sat Oct 27 2012, 12:56AM
that's what I would do, but javatc has me doubtful. the secondary geometry gave me whacko numbers which I proved wrong with my own calculations and another calculator. the coupling would've been off due to that. it's really strange, and a one time occurrence(don't ask how that works with a pre-written script).
Ben Solon, Sat Oct 27 2012, 12:56AM
that's what I would do, but javatc has me doubtful. the secondary geometry gave me whacko numbers which I proved wrong with my own calculations and another calculator. the coupling would've been off due to that. it's really strange, and a one time occurrence(don't ask how that works with a pre-written script).
Re: mathematically calculating k
Steve Conner, Sat Oct 27 2012, 08:46AM
JavaTC and FANTC use the same engine. I've used FANTC to design my last three coils and they all turned out fine. I'm using JavaTC for my next coil and I have no reason to doubt the results so far... Are you sure you entered all the figures right?
Hit the "Draw System" button and see if the picture looks like what you expect.
Steve Conner, Sat Oct 27 2012, 08:46AM
JavaTC and FANTC use the same engine. I've used FANTC to design my last three coils and they all turned out fine. I'm using JavaTC for my next coil and I have no reason to doubt the results so far... Are you sure you entered all the figures right?
Hit the "Draw System" button and see if the picture looks like what you expect.Re: mathematically calculating k
Ben Solon, Sat Oct 27 2012, 03:07PM
yes, all the dimensions where in cm, and the coil somehow had a higher inductance and lower ~fres than some of the giant brick coils, without a topload. it should be around 60-80khz in reality with a topload. the picture looked good, but you can't tell much about the secondary other than it's dimensions. i'll give it another try though.
Ben Solon, Sat Oct 27 2012, 03:07PM
yes, all the dimensions where in cm, and the coil somehow had a higher inductance and lower ~fres than some of the giant brick coils, without a topload. it should be around 60-80khz in reality with a topload. the picture looked good, but you can't tell much about the secondary other than it's dimensions. i'll give it another try though.
Print this page