A Revised Inductance Calculator

[DarkMatter]

The absolute maximum Q at a freq is the reactance of the coil at that
freq divided by the DC resistance. But the Q will always be lower
than that because the actual AC resistance will be higher because of
skin effect, when much of the inner copper wire has no current.
So you could do a simple calc based on reactance / DC res and say that
the max Q is that value, but expect less due to skin effect.

Unless said "skin" is a silver cladding... !!! Hehehehehee.....

Also, testing at 1kHz may NOT show this discrepancy, but higher freqs
will.

The word for today is LITZ

 

[DarkMatter]

Does it do flat spirals?

Mark L. Fergerson

I'll give you three guesses, and the first two don't count, and the
third one doesn't either if the answer is yes.

 

[Tom Bruhns]

When I have some time I might try comparing your web script with a
program I've learned to trust.

Just did a spot check: asked for 0.01mH in a 1" diameter, 1" long
coil, using 26 AWG wire. Got 0.0108mH at 26 turns.

The program I use tells me it should be very close to 0.0120mH.

For the RF work I do, I'll stay with a program that talks to me in
nanohenries and estimates Q and SRF as well. I also find it most
convenient to enter a winding pitch, instead of coil length, since
highest Q will generally be for a wire diameter about half the winding
pitch. I'd expect the web scrip would be useful for multi-layer
air-core audio coils for crossovers and the like.

Cheers,
Tom

 

[John Jardine]

[snip]

Coils of two or three turns are likely to be used only by radio amateurs who
insist on making magloops out of them because they mistakenly imagine they
must be more efficient than the usual single turn versions using the same
amount of copper.

my understanding is that the length and diameter of Hi-Q coils are
equal. my interest in the one turn coil stems from a program from
distinti.com. the documentation shows a Faraday triple integral and
their double integral (not Neumann's - it's from their IEL or Inertial
Electrical Law), the answers of which (for M) disagree slightly, but
their method works out the same as the M calculation in MANDK which i
mentioned in my other reply to you.

anyway, the distinti program calculates M for a single turn coil which
can then be scaled by multiplying by N1 and N2 for comparison to
measured data from multi-turn coils (don't ask why they didn't just
write that in to the program), or so they'd have me beleive. i haven't
had the time to evaluate all this and i have no measurements of real
coils available.

i'm currently digging though my old physics text to try to put it all
together and may be close to a double integral solution - not sure. a
double integral would require less computing time than the Fareaday
method. I'm inscenced that the calculation of M isn't included in my
physics text, just

v1 = M(di2/dt) ... you have to measure it.

i am willing to wrap GUIs around DOS apps in exchange for help. i have a
nice platform independant GUI framework in C++.

regards,
mike

The whole of this discussion is a storm in a tea cup. Everybody ignores the
lengths of a coil's connecting wires and their mutual inductive coupling
with the coil itself. It's equivalent to a fractional increase in the number
of turns. We should all be grateful for the existence of variable tuning
capacitors.

SOLNOID3 also calculates the inductance of long coils of widely-spaced
turns, better described as coarse helicals, which if stretched out enough
become straight lengths of wire. You will notice the calculated
self-resonant frequency of a coil of wire streched out straight is the same
as that of a half-wave dipole of the same wire length. Clever program, eh?
Free to USA citizens. ;o)
----
=======================
Regards from Reg, G4FGQ
For Free Radio Design Software
go to http://www.g4fgq.com
=======================

It looks like you are moving towards a finite element field solver. The
computer can supply good answers on various 3D coil geometries, by solving
for mutual inductance using Maxwell's 'Geometrical mean distance'
repetatively applied to large numbers of points on a mathematical model of
the 3D coil structure. The method is effective but slow and a ballache to
programme.
regards
john

 

[Active8]

[snip]

Coils of two or three turns are likely to be used only by radio amateurs who
insist on making magloops out of them because they mistakenly imagine they
must be more efficient than the usual single turn versions using the same
amount of copper.

my understanding is that the length and diameter of Hi-Q coils are
equal. my interest in the one turn coil stems from a program from
distinti.com. the documentation shows a Faraday triple integral and
their double integral (not Neumann's - it's from their IEL or Inertial
Electrical Law), the answers of which (for M) disagree slightly, but
their method works out the same as the M calculation in MANDK which i
mentioned in my other reply to you.

anyway, the distinti program calculates M for a single turn coil which
can then be scaled by multiplying by N1 and N2 for comparison to
measured data from multi-turn coils (don't ask why they didn't just
write that in to the program), or so they'd have me beleive. i haven't
had the time to evaluate all this and i have no measurements of real
coils available.

i'm currently digging though my old physics text to try to put it all
together and may be close to a double integral solution - not sure. a
double integral would require less computing time than the Fareaday
method. I'm inscenced that the calculation of M isn't included in my
physics text, just

v1 = M(di2/dt) ... you have to measure it.

i am willing to wrap GUIs around DOS apps in exchange for help. i have a
nice platform independant GUI framework in C++.

regards,
mike

The whole of this discussion is a storm in a tea cup. Everybody ignores the
lengths of a coil's connecting wires and their mutual inductive coupling
with the coil itself. It's equivalent to a fractional increase in the number
of turns. We should all be grateful for the existence of variable tuning
capacitors.

SOLNOID3 also calculates the inductance of long coils of widely-spaced
turns, better described as coarse helicals, which if stretched out enough
become straight lengths of wire. You will notice the calculated
self-resonant frequency of a coil of wire streched out straight is the same
as that of a half-wave dipole of the same wire length. Clever program, eh?
Free to USA citizens. ;o)
----
=======================
Regards from Reg, G4FGQ
For Free Radio Design Software
go to http://www.g4fgq.com
=======================

It looks like you are moving towards a finite element field solver. The
computer can supply good answers on various 3D coil geometries, by solving
for mutual inductance using Maxwell's 'Geometrical mean distance'
repetatively applied to large numbers of points on a mathematical model of
the 3D coil structure. The method is effective but slow and a ballache to
programme.
regards
john

i downloaded Ansoft's (?) Maxwell SV -- a 2-solver where you input a
cross section, but i don't see how it would be used to calculate M. it
does a bunch of other stuff.

you're right, i think. it is a type of FEM solver. i can even see the
solution or how the integral would be set up for a triple integral, i
just can't get the equation. haven't had much time to really look at
deriving my own, either.

BRs,
mike

 

[Fred Abse]

No. You don't get that feeling unless you fart in public, and are
noticed as the culprit. Silent and deadly doesn't count unless you are
stupid and admit to it.

Never could manage silent and deadly. Deadly, yes.

 

[Mark Fergerson]

There's a small? formula (Spielrein's) to do these. If of any interest I'll
scan the page.
regards
john

Yes, please.

Mark L. Fergerson

 

[Phantom]

Does it do flat spirals?

Mark L. Fergerson

This site:
http://members.aol.com/marctt/Technical/Inductance_References.htm

has a lot of references about the general subject, including some
for spiral inductors.

 

[Gary Lecomte]

[snip]

Coils of two or three turns are likely to be used only by radio amateurs who
insist on making magloops out of them because they mistakenly imagine they
must be more efficient than the usual single turn versions using the same
amount of copper.

my understanding is that the length and diameter of Hi-Q coils are
equal. my interest in the one turn coil stems from a program from
distinti.com. the documentation shows a Faraday triple integral and
their double integral (not Neumann's - it's from their IEL or Inertial
Electrical Law), the answers of which (for M) disagree slightly, but
their method works out the same as the M calculation in MANDK which i
mentioned in my other reply to you.

anyway, the distinti program calculates M for a single turn coil which
can then be scaled by multiplying by N1 and N2 for comparison to
measured data from multi-turn coils (don't ask why they didn't just
write that in to the program), or so they'd have me beleive. i haven't
had the time to evaluate all this and i have no measurements of real
coils available.

i'm currently digging though my old physics text to try to put it all
together and may be close to a double integral solution - not sure. a
double integral would require less computing time than the Fareaday
method. I'm inscenced that the calculation of M isn't included in my
physics text, just

v1 = M(di2/dt) ... you have to measure it.

i am willing to wrap GUIs around DOS apps in exchange for help. i have a
nice platform independant GUI framework in C++.

regards,
mike

The whole of this discussion is a storm in a tea cup. Everybody ignores the
lengths of a coil's connecting wires and their mutual inductive coupling
with the coil itself. It's equivalent to a fractional increase in the number
of turns. We should all be grateful for the existence of variable tuning
capacitors.

SOLNOID3 also calculates the inductance of long coils of widely-spaced
turns, better described as coarse helicals, which if stretched out enough
become straight lengths of wire. You will notice the calculated
self-resonant frequency of a coil of wire streched out straight is the same
as that of a half-wave dipole of the same wire length. Clever program, eh?
Free to USA citizens. ;o)
----
=======================
Regards from Reg, G4FGQ
For Free Radio Design Software
go to http://www.g4fgq.com
=======================

It looks like you are moving towards a finite element field solver. The
computer can supply good answers on various 3D coil geometries, by solving
for mutual inductance using Maxwell's 'Geometrical mean distance'
repetatively applied to large numbers of points on a mathematical model of
the 3D coil structure. The method is effective but slow and a ballache to
programme.
regards
john

i downloaded Ansoft's (?) Maxwell SV -- a 2-solver where you input a
cross section, but i don't see how it would be used to calculate M. it
does a bunch of other stuff.

you're right, i think. it is a type of FEM solver. i can even see the
solution or how the integral would be set up for a triple integral, i
just can't get the equation. haven't had much time to really look at
deriving my own, either.

BRs,
mike

Totally OFF TOPIC:

I Couldn't get any pictures of the fire, about 2 miles away, too high
of a mountain blocking my view.

But I got some of one of the Hellicopters that was hauling water to
the fire.
Pictures 1 to 8 are taken from town site (Elev 728 Meters), the
balance are from near the top of another mountain. Elevations are
shown in names. Some nice scenery pics also.

http://www3.telus.net/chemelec/Fire

If interested, take a look.

Take care....Gary

 

[Active8]

nice looking hills.
BR,
mike

 

[Brian]

There's a small? formula (Spielrein's) to do these. If of any interest I'll
scan the page.
regards
john

Me Also.....Brian

 

[John Jardine]

news:[email protected]...

I've stuck 2 Gif's on A.B.S.E. Any problems and I'll be happy to email 'em
direct.
regards
john

 

[Active8]

Piss and moan a little, why doncha?

well, sorry. the OP was pretty much blowing off suggestions to make the
thing better. same with me after i was decent enough to try it and
submit my 2 pfennig.

You were smart, up until that moment.

got smart again, real soon and tried Gary's twocoils and solnoid3
pretty decent little dos apps. looks like twocoils returns a good enough
approx of air core M to finish off an analysis and build. it just
doesn't agree with solnoid3 on inductance for a given coil, but if you
put those numbers in the OP's calculator, his gives a result close to
twocoils, but would be better with nH accuracy.

Sheesh, just get online at any wire manufacturer, and use their
online calculators. They are much more comprehensive, and have
compensations built in, I'm sure.

wire manufacturers. never checked. it'd be a good way to test other
calculators, i'd think. of course, there's a fair amount of other little
sites where you can find stuff like that. but a mfg's tools would be
easier to find and on the money. magnetics mfgs have a bunch of
resources, also. thanks.

br,
mike

 

[Brian Kraft]

Since astraweb went pay-only I'm lacking binary access. Know any free
news servers with binaries? Living without a cc can have its downsides
but makes life easier in other respects.

Astraweb is still free 50 MB/day, but if you try to use more than one
free account at the same time, there will be an IP address match and
both/all of your free Astraweb accounts will be de-activated.
They say "you are only allowed one free account" and they mean it.

Here's a place to get a free account at Newscene, no CC needed:
http://www.eastbaytech.com/usenet-acct-setup.asp?P=U

 

[Winfield Hill]

John Jardine wrote...

I've stuck 2 Gif's on A.B.S.E. Any problems and I'll be happy
to email 'em direct.

Strange, I only see one gif there, page 181 from Grover.

Thanks,
- Win

 

[Terry Pinnell]

John Jardine wrote...

Strange, I only see one gif there, page 181 from Grover.

I don't even see one yet (14:32 UK time).

 

[Winfield Hill]

Terry wrote...

But now I see both. (17:00 UK time).

Are they contained in one post or two?

Thanks,
- Win

 

[Jonathan Kirwan]

Strange, I only see one gif there, page 181 from Grover.

I see two attachments. Sadly, they are both in GIF and Acrobat
assumes there is 72 pixels/inch in this case, so it won't OCR
the text for me. If these were scanned into TIFF, with the
resolution per inch at say 300 DPI, I could try and OCR scan
this text and clean it up some. But other than that, they are
quite legible.

Jon

 

[John Jardine]

I see two attachments. Sadly, they are both in GIF and Acrobat
assumes there is 72 pixels/inch in this case, so it won't OCR
the text for me. If these were scanned into TIFF, with the
resolution per inch at say 300 DPI, I could try and OCR scan
this text and clean it up some. But other than that, they are
quite legible.

Jon

Jon. The GIF pages were scanned 'line art' at 300dpi. I'll see if I've got a
bitmap converter that'll do a GIF to TIFF.
regards
john

 

[Jonathan Kirwan]

Jon. The GIF pages were scanned 'line art' at 300dpi. I'll see if I've got a
bitmap converter that'll do a GIF to TIFF.

I know. My Acrobat tools has the ability to do OCR (so they
say, anyway) for anything done between 200 and 400 DPI. Which
is exactly what you did. But GIF doesn't contain the number of
pixels for each inch in its format, so Adobe (apparently)
assumes that the DPI is 72. Natually, it claims you scanned in
a "darn big page" and claims that the pixels per inch is bad.

TIFF includes that information in it. But I'm not sure what
happens if you just convert GIF to TIFF. Probably nothing good.

Jon