Topics

Toroid winding numbers and measurements

 

This may be of help to other n00bs as well as those without test gear. Fortunately a ham friend not too far away has an inductance measuring tool. The results are interesting, and I think important and likely to affect others.

My output power at 12V was just around or a bit less than 1W instead of 3W.

I thought that I wound the toroids L1-L4 quite well. I thought they we fairly reasonably tight and well spaced. However, here are the readings I got and the effects of removing turns, all of this applies to 30m:

L1: Required or ideal value 1.1uH. 19 turns actual value was 1.22uH, removing turns I got 18 turns: 1.16uH and 17 turns 1.07uH, I thus had to remove 2 turns.

L3: same as L1, expected value 1.1uH with 19 turns, actual value was 1.32uH. Removed turns, 18 gave 1.24uH, 17 turns 1.18uH and 16 turns 1.08uH. . I thus required 16 turns, removing a total of three turns! This is quite a lot, so I suspect that a "normal amateur-amateur first time toroid winder" doesn't manage to wind it as tight as an expert ie non-amateur amateur. Assuming that tighter windings means lower inductance.

L2: expected value is 1.3uH with 20, actual was measured at 1.6uH. Removing turns, 19 gave 1.34uH, 18 was 1.24, so I settled on 19, thus removing one turn.

L4: expected value is 0.78uH with 14 turns. Mine measured 0.88uH. 13 turns gave 0.86, 12 gave 0.79 and 11 gave 0.76, so I settled on 12 turns, thus removing 2 turns.

So on average, I had to remove 2 turns, or 1-3 turns. I have yet to put them all back and measure the power, but hopefully it'll be closer to the expected 3W at 12V.

Another reason I may have had low power output was that my torroid connections may not have been the best... now the ends are properly tinned and will be properly connected, then let's see!

Hopefully the receiver will sprint back to life once I have redone T1...



L2: expected value 1.3uH. 20 turns actual value was 1.6uH. 19 turns: 1.34uH. Here I had to remove only 1 winding.

L3: same as L1, expected value 1.1uH.

Alan G4ZFQ
 


This may be of help to other n00bs as well as those without test gear.
Or it might confuse them more:-) Do I do it this way or as the instructions say?
The design should allow for tolerances, most constructors do not have reliable measuring gear. I agree, most inductors I wind test higher than specified but how many of them will significantly affect performance? (I believe tolerances in the core material are the reason.)

If you find a single critical winding then report it.
I'm not sure that rewinding all in one go is going to be particularly helpful, Hans should have factored in usual tolerances but maybe got one a little too close.

My output power at 12V was just around or a bit less than 1W instead of 3W.
It would be good to know a single reason, if there is one.

Assuming that tighter windings means lower inductance.
Not sure, I've never heard that one, spreading/compressing to fine tune, yes.

73 Alan G4ZFQ

Nick Norman
 

Morning all,

This is down to tolerance in the core material, the tightness will make (approximately) 0 difference to the inductance - coverage of the circumference will change it (or is that what you meant by tightness?)
I think the specs are for 80% coverage.
HTH

Nick
M0HGU

--
73 Nick M0HGU

 

Sorry I stand corrected, n00b's pay not attention! Follow the manual... it does say if the power is low, remove a turn or two...
Now, having done everything at once... new problem: when I turn on and go to BPF menu I get only "1" for 30m, and "4" for 20m... but I tried to get something that may work on both 20m and 30m, so I wound this time 31:3:3:3 -- priority is to get this working no 30m, I'll try removing 1 to 30, and hope I don't have to make 4, 4, 4 as the BPF does say for 20m it will cover down to 9.6 MHz without too much loss I think. I also removed some turns on L1-L4 also with aim of getting it working for 30/20m.

Not tried on receive antenna yet, but, when I get to do the adjustments of I-Q bal frequency, Phase Lo and Hi frequency, I now have NO bars moving at all! So something appears to be very much amiss... I will do some more checking tomorrow with the help of someone experienced, e.g. connections of those toroids, and various voltages etc but, I'm just wondering if anyone knows that no data at all (no bars, nothing moving) in those adjustment menus may indicate as to a possible fault?

 

Scrap that... with headphones plugged in those configuration items are now working, though it doesn't make a nice pure sound on 20m but a loud pulsing sound. We'll get there, I should just have stuck to making it 30m only :-)

Jim Sheldon
 

I'm still waiting for my 40 meter version (#943) but shipping has resumed again as of yesterday so hopefully by the end of this month.

I have all the proper cores for all the toroids called for in the construction manual, and after reading this thread I decided to wind a set and measure the inductance with my AADE L/C meter which has proven to be quite accurate.

To achieve the proper inductance for L4, I had to use 1 turn less than called for so I wound L2 next - same thing - 1 turn less to get the right inductance for the band.  I wound L1 and L3 with 1 turn less and they came out very close to the called for value.

Now, a caveat - NOT ALL TOROID CORES ARE CREATED EQUAL!  Especially if they are Chinese made - LOL. Depending on the actual permeability of the core, the inductance will most likely vary a bit with the same turns count.  Also I saw something earlier about spread of turns VS inductance.  Close winding gives the MOST inductance and if you spread the turns out you will reduce the inductance somewhat.  Not by much but it will be less with wider spaced turns by the theory of inductors and I confirmed this by measuring each of the above coils.  I first used the called for number of turns and they all measured up to a half microhenry high.  Spreading the turns got it down about .2 microhenry but they still were too high so I removed a turn from each which with close to even spacing over the core got them within .02 microhenry which I think should be close enough.  Just my experience with the set of cores I had on hand.  Your mileage will most certainly vary so don't take this as gospel, just as a point of information and a starting point for troubleshooting.

I'm looking forward to receiving my QCX so I can guit it built and hopefully confirm my observations on the LPF coils.

Jim - W0EB

Leland L. Bahr
 

One last caveat:  The measured inductance will very with the frequency used to measure the inductance.  Hans has not told us if he just calculated it or if he measured the best inductance with an instrument.  If he used an instrument we would have to know the frequency used.  In other words if you remove or add turns to acquire the exact inductance needed but did this at a different frequency because your meter is on a different frequency, your results will not be what is required.  In my estimation one is playing with fire trying to match inductance without knowing the standard's test parameters.  A lot of hams don't realize inductance of an inductor changes with frequency.  An inductor tested at one frequency will give a different inductance if tested at a different frequency.

Lee, w0vt


On 11/1/2017 11:37 AM, w0eb@... wrote:
I'm still waiting for my 40 meter version (#943) but shipping has resumed again as of yesterday so hopefully by the end of this month.

I have all the proper cores for all the toroids called for in the construction manual, and after reading this thread I decided to wind a set and measure the inductance with my AADE L/C meter which has proven to be quite accurate.

To achieve the proper inductance for L4, I had to use 1 turn less than called for so I wound L2 next - same thing - 1 turn less to get the right inductance for the band.  I wound L1 and L3 with 1 turn less and they came out very close to the called for value.

Now, a caveat - NOT ALL TOROID CORES ARE CREATED EQUAL!  Especially if they are Chinese made - LOL. Depending on the actual permeability of the core, the inductance will most likely vary a bit with the same turns count.  Also I saw something earlier about spread of turns VS inductance.  Close winding gives the MOST inductance and if you spread the turns out you will reduce the inductance somewhat.  Not by much but it will be less with wider spaced turns by the theory of inductors and I confirmed this by measuring each of the above coils.  I first used the called for number of turns and they all measured up to a half microhenry high.  Spreading the turns got it down about .2 microhenry but they still were too high so I removed a turn from each which with close to even spacing over the core got them within .02 microhenry which I think should be close enough.  Just my experience with the set of cores I had on hand.  Your mileage will most certainly vary so don't take this as gospel, just as a point of information and a starting point for troubleshooting.

I'm looking forward to receiving my QCX so I can guit it built and hopefully confirm my observations on the LPF coils.

Jim - W0EB


Virus-free. www.avast.com

Chris Wilson
 

Hello Lee,

This is very true, I am experimenting with a new power amp and also
a big LPF bank for it. If I measure with a Fluke LCR mter that's still
in calibration it measures at (from memory) 100kHz. But if I use my
AIM4170 analyser I can measure inductance at the frequency the amp
will be run at, and the inductance value is somewhat different. This
variance is of course more marked the higher infrequency above the
Fluke test frequency the inductor under test will ACTUALLY see in use.
I suppose a spectrum analyser with a tracking generator would be a
better tool still for messing with them.

Wednesday, November 1, 2017

One last caveat:  The measured inductance will very with the
frequency used to measure the inductance.  Hans has not told us if
he just calculated it or if he measured the best inductance with
an instrument.  If he used an instrument we would have to know the
frequency used.  In other words if you remove or add turns to
acquire the exact inductance needed but did this at a different
frequency because your meter is on a different frequency, your
results will not be what is required.  In my estimation one is
playing with fire trying to match inductance without knowing the
standard's test parameters.  A lot of hams don't realize
inductance of an inductor changes with frequency.  An inductor
tested at one frequency will give a different inductance if tested at a different frequency.
Lee, w0vt





--

2E0ILY
Best regards,
Chris mailto:chris@...
--
Best regards, Chris Wilson (2E0ILY)

Jim Sheldon
 

Very good point.  Been over 50 years since I had my theory classes and I had forgotten this point.  I do have a Midnight Design Solutions Scalar Network Analyzer which can measure L at frequencies up to at least 35 MHz with the right fixture.  I'll look into this and see what I can comeup with.

W0EB

Alan G4ZFQ
 

An inductor tested at
one frequency will give a different inductance if tested at a different frequency.
Lee,

Have you any examples of this?
All inductors I have tried with a lowish frequency tester and a GDO at working frequency have coincided with a reasonable accuracy.

73 Alan G4ZFQ

Alan G4ZFQ
 

< If I measure with a Fluke LCR mter that's still
in calibration it measures at (from memory) 100kHz. But if I use my
AIM4170 analyser I can measure inductance at the frequency the amp
will be run at, and the inductance value is somewhat different.
Chris,

Somewhat? How much?
What type of inductor?

I've seem Lee's warning many times, I want to know when the variation might be significant.

73 Alan G4ZFQ


This
variance is of course more marked the higher infrequency above the
Fluke test frequency the inductor under test will ACTUALLY see in use.
I suppose a spectrum analyser with a tracking generator would be a
better tool still for messing with them.

J68HZ
 

Its important to test your inductors against known standards (inductor of a known inductance).  Just reading your LC meter and believing it can be very deceiving.


Dr. William J. Schmidt - K9HZ J68HZ 8P6HK ZF2HZ PJ4/K9HZ VP5/K9HZ PJ2/K9HZ

 

Owner - Operator

Big Signal Ranch – K9ZC

Staunton, Illinois

 

Owner – Operator

Villa Grand Piton - J68HZ

Soufriere, St. Lucia W.I.

Rent it: www.VillaGrandPiton.com


email:  bill@...

 


On Nov 1, 2017, at 2:26 PM, Alan G4ZFQ <alan4alan@...> wrote:

< If I measure with a Fluke LCR mter that's still
in  calibration  it  measures at (from memory) 100kHz. But if I use my
AIM4170  analyser  I  can  measure inductance at the frequency the amp
will  be  run at, and the inductance value is somewhat different.

Chris,

Somewhat? How much?
What type of inductor?

I've seem Lee's warning many times, I want to know when the variation might be significant.

73 Alan G4ZFQ


This
variance  is  of  course  more marked the higher infrequency above the
Fluke test frequency the inductor under test will ACTUALLY see in use.
I  suppose  a  spectrum  analyser with a tracking generator would be a
better tool still for messing with them.



Leland L. Bahr
 

Yes, a scientific fact.

Lee, w0vt

On 11/1/2017 2:19 PM, Alan G4ZFQ wrote:
 An inductor tested at
one frequency will give a different inductance if tested at a different frequency.
 Lee,

Have you any examples of this?
All inductors I have tried with a lowish frequency tester and a GDO at working frequency have coincided with a reasonable accuracy.

73 Alan G4ZFQ

-

G0DZB
 

On Wed, 2017-11-01 at 15:14 -0500, Leland L. Bahr wrote:
Yes, a scientific fact.
Does this "fact" only apply for iron dust cored inductors because the
equations for the inductance of air cored coils do not include a
frequency term.

PeterO
G0DZB

Alan G4ZFQ
 

Yes, a scientific fact.
Does this "fact" only apply for iron dust cored inductors because the
equations for the inductance of air cored coils do not include a
frequency term.
Peter

So do iron dust cored inductors include a frequency term?
If so just how much difference?

My tests on air and ferrite cored coils have been with this https://sites.google.com/site/vk3bhr/home/lcm1 and using 1% capacitors to make these coils resonant at around working frequency and careful GDO measurements.

People quote that frequency makes a difference, say it's a fact but then do not seem to be able to give examples of when any differences become significant, say more than a few per cent.

73 Alan G4ZFQ

Paul Jorgenson KE7HR
 

An example of different amounts of core coverage with a given number of turns vs. inductance is at:
http://www.vkham.com/Info/ferro/wind_deg.gif

As a practical matter, different mixes (stuff the toroid is made of) have frequency sensitivity. Look at the data sheets or use the calculator available at
http://toroids.info/T50-2.php

Will it make a difference to your project? You have to experiment. The range of possible values of the capacitors and inductors make it difficult for the designer to cover all the bases in a low cost, high performance, complex circuit. You have to experiment.

The case of having 20 or so turns and having to remove one or two is well within the range of the actual permeability range of the cores - 5% to 10%. I was just working on a circuit (not a QRP Labs circuit) that has three equal inductors in the filter. To get the peak response where I need it, I had to remove one turn from two of them and three from the other. In this circuit, I had modeled (Experimental Methods in RF Design software - sadly no longer available) the response to have an insertion loss of about 3 dB. I ended up with about 20 db insertion loss. I had modeled the Q of the inductors to be 200 and had to change the modeled Q to be 50 to match what had been built. Time for much larger cores with much larger wire on that one to get the Q up!

So, go out and get a cheap inductance meter (eBay $10 to $50) if you are going to participate in busting on the design and make some of your own measurements and provide data and not just a complaint.


---------

Paul KE7HR

John Backo <iam74@...>
 

All:

Note that the only thing that will change inductance on a toroid
winding is THE NUMBER OF TURNS.

Pushing the turns together or apart seems to change it a little,
but all one is really doing is changing the distributed inter-winding
capacitance. Yes, that does affect frequency response.

As noted elsewhere, the is some variation in the cores. Amidon and
Micrometals try to keep these as uniform as possible, but even they admit
there is some variation. This, of course, means that the same type of core
from different manufacturers will give different results...maybe.
Different core variations mean different inductance and temperature
characteristics. But a type 2 core is still a type 2, and so on. Two cores
may give different results, but they will still be in the ballpark.

The only sure way to know is to wind your coil., and then measure it with a good
inductance meter. Which is what you were going to do anyway. Right?

Hope this helps.

john
AD5YE

Ben Bangerter, K0IKR
 

Paul -  KE7HR:  
I have the first edition of EMRFD, with the CD (though I have not opened the CD), so could provide programs on the CD to those interested.  Email me at bwbangerter-at-yahoo-dot-com if interested and I will try to accomadate.

Ben

Graham, VE3GTC
 

Amidon is not a manufacturer but rather only a distributor of Micrometals materials.

Micrometals has very interesting website.

An interesting design program here: http://www.micrometals.com/software_index.html

some selected app notes:

App note on Iron Powder Cores for High Q inductors http://www.micrometals.com/appnotes/appnotedownloads/ipc4hqi.pdf

App note for Iron Powder core selection for RF power applications: http://www.micrometals.com/appnotes/appnotedownloads/ipcs4rfp.pdf

Plus others. There is also some good documentation to be found on the Amidon site as well: http://www.amidoncorp.com/

In the first Micrometals app note you will note that there is no frequency term in the formula for calculating inductance of a toroid inductor BUT cross sectional area is one term which does effect inductance.

It is really not that difficult - calculate, build, measure and test, trim and repeat as many steps as necessary to get where you want to go all the while keeping good notes.

Of course, you can also get lost in the deep details. For some light bed time reading:

http://www.ferroxcube.com/FerroxcubeCorporateReception/datasheet/FXC_HB2013.pdf (1108 pages)

https://www.fer.unizg.hr/_download/repository/Philips_ferrites_databook.pdf (848 pages)

cheers, Graham ve3gtc

On 2017-11-01 23:20, John Backo via Groups.Io wrote:
All:

Note that the only thing that will change inductance on a toroid
winding is THE NUMBER OF TURNS.

Pushing the turns together or apart seems to change it a little,
but all one is really doing is changing the distributed inter-winding
capacitance. Yes, that does affect frequency response.

As noted elsewhere, the is some variation in the cores. Amidon and
Micrometals try to keep these as uniform as possible, but even they admit
there is some variation. This, of course, means that the same type of core
from different manufacturers will give different results...maybe.
Different core variations mean different inductance and temperature
characteristics. But a type 2 core is still a type 2, and so on. Two cores
may give different results, but they will still be in the ballpark.

The only sure way to know is to wind your coil., and then measure it with a good
inductance meter. Which is what you were going to do anyway. Right?

Hope this helps.

john
AD5YE

Arv Evans
 

The "frequency makes a difference" part of measuring inductance of any coil
is the distributed capacitance factor.  This apparent difference in measured
inductance can be used to calculate the amount of distributed capacitance
between windings, but you cannot do anything about it anyway.

When I need to make things come out exactly where I want them I usually
use a signal generator and oscilloscope to measure the resonance point.
  • Add up all the frequency determining capacitance in your circuit.
    Select a fixed capacitor of that value.

  • Connect the above capacitance across your inductor.

  • Couple the signal generator across the tuned circuit through a 2.2 pf
    capacitor.

  • Couple the oscilloscope across the inductor through a 2.2 pf capacitor.

  • Tune the signal generator through the expected frequency range while
    looking for a peak on the oscilloscope.
Since you know the C-value and the frequency you can now calculate the
inductance of your coil.  This process includes any distributed capacitance that
is inherent in the inductor windings. 

I have a variable capacitor that has a dial to show capacitance settings in pf. 
With this I can add or subtract capacitance until the inductor resonates at the
exact frequency I want.  This helps when working up a new circuit design. 

The ancient Heathkit LC analyzer is no more.  It was a very useful tool for doing
design work.

Arv  K7HKL
_._