Understanding of suitabily of ferrite grade materials for resonant tuned circuit

[Sparks_nz]

Thank you all for your advice. I will try a gapped potcore first, and see how it goes. Then perhaps a powdered iron toroid, if I can get one a suitable size to put wire bobbin through it. I appreciate all the comments, and have learnt a lot. Still reading material you guys have suggested. This forum is awesome !!

[jonpaul]

std 22, 26 mm pot core u 1600, 2200 etc
gap yourself with paper or G10
use bobbin

Toroids winding is a pain, and have high self C thus lower SRF

J

[Sparks_nz]

Thanks for this idea Jon. I was looking at the two grades of gapped P26 potcores which are recommended for resonant circuits. EPCOS N48 and M33.
N48 ui = 2,300 (10-100kHz)
M33 ui = 750 (0.2-1MHz)
From what youre saying, I presume the N48 ferrite can operate well over a much wider freq range than what is recommended above.
I never thought about putting an insulating gap between the core halves that is brilliant. I do wonder about the exact thickness needed. I presume this critical to getting the correct amount of ue reduction?
JEFF

[Sparks_nz]

Another thing Im not sure about is I presume I measure the SRF using no added capacitor, and if it is suitably high enough by adjusting turns (say twice the required frequency) then I add a suitable capacitor for resonance at the desired frequency. I have seen using 50% of SRF mentioned as the best percentage for maximum Q?

[T3sl4co1l]

Yeah, shimming the cores is the second best way to airgap it; grinding the center peg is the best way.  Which can be done with a rotary tool and grinding wheel, but it's a bit messy and dirty.  Ferrite is brittle, so go carefully.  You have to grind away a bit, check inductance, and back and forth until it's the required value.

There are various ways to measure SRF.  The most basic is to use the component in a series impedance divider.  Adjust frequency until the phase shift goes away: at this point the component has the same phase as the other element.  In a divider with a resistor, this means the component is resistive, which means it's at resonance (L and C cancel out).

So:
Signal generator from Vi to GND
Component from Vi to Vo
Resistor from Vo to GND
Scope probe from Vo to GND

Use a 10x probe, and use a smallish value resistor (a few k max).  This minimizes loading from the probe (which is capacitive in the ~MHz range).

It could also be tested with a grid-dip method, using a coupling link (i.e., a small coil) and whatever coupling gets through the ferrite core (pot cores are well shielded, but not perfect; or a small loop could be put into the winding window or near the air gap to get more coupling).  In this case you're testing for the frequency where the link/loop/antenna sees a dip (notch) in amplitude response.  A sweep generator and oscilloscope (set to peak-detect mode) is one way to do this.  (This isn't as easy a method overall, so I'm not bothering to get very specific; it used to be more common, with the "grid dip meter" being fairly common equipment.  Which would be easy enough to make, it's true, but not so easy to calibrate, is the problem.)

As for measuring inductor parameters with limited equipment, I use these for various things:
https://www.seventransistorlabs.com/Calc/RLC.html

Tim

[jonpaul]

Rebonjour

gaps....in 55 years we never ground a gap.

a. paper or G10 can easily be adjusted, once ground out a gap cannot be reduced.

b. certain pot cores have adjusters to trim the inductance eg tune a resonance

SRF..SELF resonant frequency mean the inductance by itself NO  added cap! it's the resonance of the L and winding parasitic C.

Measurement SRF...any oscillator,audio, function generator RF, that covers the approx frequency range.

GEN>>series R 10k..100k>>L under test.

observe V across the inductance with 10x scope probe.

Sweep F and find maximum

Alternative demo revers L and R, observe voltage on  resistance

Easy to make a Vi lissajous figure.

J

[Sparks_nz]

Hi,
Thanks. Yes, I see that the EPCOS N48 potcore has an adjustable slug that can be screwed in to adjust inductance, the spec sheet says it has 10% range or so.

[Sparks_nz]

One thing that is confusing is that the P26 potcore spec sheet lists various variants of the N48 material type, depending on the gap distance (s). See attached snap.The AL value varies considerably due to this, and Im unsure what or if there is an optimum one to choose from.
Does a large gap mean it is more broadband in its overall Q value, compared to a smaller gap, or are there other properties that the gap size will influence in regards a resonant circuit. I know you have mentioned energy storage in relation to the gap.

[T3sl4co1l]

Q generally goes up with gap, because air has no loss.  But at low enough mu_eff, the amount of wire required to take advantage of the core goes way up, and eventually overall Q starts going down again.  mu_eff in the 30-120 range is generally best with copper windings.

Tim

[Sparks_nz]

Thanks Tim, I think Ive got a good handle on things now, thanks to yourself and the other guys.
J