80 meter bpf toroid winding 30 turns givel 4.3 micro Henrys, not 3.8


KI7MWA
 

Been assembling an 80 meter band pass filter, and have consistently had to remove 6 turns during the tuning phase (using a nanovna).  This most recent attempt, I used my Peak LCR45 to measure the initial inductance, before installing the toroid.  With the recommended 34 turns (4 more than the design "design" number, 30), I get 5+ uH.  Pulling the extra 4 turns gives me 4.3 uH.  The design value is 3.8 uH.

Anyone with similar experience?  Or am I doing something wrong here?


KI7MWA
 

So, I wound it again and got the design value (3.83 uH).  The problem, really, is that, tuning with the nanovna, I have to reduce the turns to 28 (I've done this 3 times now), and when I trim the extra wire and check it again, I've passed the tuning point (trimmer cap is no longer about half way open).  I am aiming for a center frequency of 3.75 MHz (too high?).  Best insertion loss I've achieved is around -3.5 dB.


Kees T
 

Try spreading the turns and/or bunching them together and see what the variation is. It may surprise you.

73 Kees K5BCQ 


Bob Benedict, KD8CGH
 

I second Kees advise on tuning by turn spacing.

I have also had to remove turns from LPF torroids. I put a small modeling study in the files section
https://groups.io/g/QRPLabs/files/LP.pdf
that you can use to compare your sweep to.

One issue with a VNA is calibration. You should calibrate at the point of device test. I built a LPF test holder along with plug in calibration devices for open, short, 50 ohm load and through. Here is a post on it
https://groups.io/g/QRPLabs/message/49978
and a nice commercial version here
https://groups.io/g/QRPLabs/message/49980

Finally here are two posts I found helpful from Allison
https://groups.io/g/QRPLabs/message/51788
https://groups.io/g/QRPLabs/message/49853

--
  73
    KD8CGH


KI7MWA
 



Here's my setup.  The two cal standards lying over are the open and through.  If you're familiar with QRP Labs products you will recognize the band pass filter, with the toroids unmounted.  This is an 80 meter kit, 30 turns of #28 wire on a T37 doesn't leave much room for bunching or squeezing.


Bob Benedict, KD8CGH
 

Your PeakLCR45 measures inductance at a max frequency of 200 kHz. The material in torroid inductors has some frequency dependence. I measure my inductors at the intended frequency of use. Here's one of many articles on using a nanoVNA to measure inductance.

https://owenduffy.net/blog/?p=17298

The other point that Allison makes is to check capacitance. I used to just assume that inductors were always the problem.
--
  73
    KD8CGH


Kees T
 

KI7MWA,

If xx turns of #28 wire do not allow moving the turns around a little, try #30 wire.

73 Kees K5BCQ


KI7MWA
 

I thank all for the advice.  Clearly I've f***ed this up, so I've ordered another kit and I'm tossing this one.


Arv Evans
 

As mentioned earlier by someone else,  compressing or spreading 
the turns on a toroid core probably does not change the inductance
very much (the number of turns remains the same).  What it does 
do is to change the distributed capacitance between turns.  

Also mentioned by others is that the apparent inductance of metallic 
cores does change with frequency.  This could be a factor if you are 
operating the core at widely different frequencies.

DC loading also causes toroidal cores to change apparent 
inductance.  While probably not an issue with DC isolated 
filters, it could be problematic for tank circuits that carry DC
current.

Arv
_._



On Mon, Sep 7, 2020 at 8:56 AM Kees T <windy10605@...> wrote:
KI7MWA,

If xx turns of #28 wire do not allow moving the turns around a little, try #30 wire.

73 Kees K5BCQ


@CurtisM
 

take another careful look - if any apparent damage is limited to the low pass filter section of the board, with a little wire you can replace the traces.  many of us struggle somewhere on these kits.  also, note most of us assemble and get things working without a nanoVNA (I need to learn how to use the calibration functions - even though I have been using VNA's since 1979 !).  best to slow down and figure things out, and maybe not presume that things are way off (I did remove one turn from the end toroids on my 80m QCX, but I think I gained less than a half watt - I am thinking they are okay as is - especially when I decide to drive the PA, being assembled just now - but it was doing nicely last evening at 3.5 watts out). 

I hope you are regrouping - and perhaps ordering another QCX+ for a different band?  oh yes my original 40m QCX sits in sick bay awaiting a PLL chip ... best to keep a reasonable pace with these things. 

73 curt wb8yyy