Hi,
I am going to wind a toroid inductor for this circuit,
but would a high permeability core of this type have any downsides?
And is 34 AWG wire fine for this?
Info on core:
Core type MPP
OD 39.90 mm
ID 24.1 mm
HT 14.5 mm
AL value: 740 +/- 8% (nH)
MPP Permeability: 550 (μ)
Wire gauge: 34 AWG
Turns:
16 KHz - 29.9 mH 195 turns
10 kHz - 31.4 mH add 5 turns
8 kHz - 48.2 mH add 48 turns
5 kHz - 67.5 mH add 45 turns
4 kHz - 87.9 mH add 41 turns
3 kHz - 156 mH add 111 turns
Help is much appreciated!
I am going to wind a toroid inductor for this circuit,
but would a high permeability core of this type have any downsides?
And is 34 AWG wire fine for this?
Info on core:
Core type MPP
OD 39.90 mm
ID 24.1 mm
HT 14.5 mm
AL value: 740 +/- 8% (nH)
MPP Permeability: 550 (μ)
Wire gauge: 34 AWG
Turns:
16 KHz - 29.9 mH 195 turns
10 kHz - 31.4 mH add 5 turns
8 kHz - 48.2 mH add 48 turns
5 kHz - 67.5 mH add 45 turns
4 kHz - 87.9 mH add 41 turns
3 kHz - 156 mH add 111 turns
Help is much appreciated!
I never wound inductors for an equ and certainly will never do this. For the targeted bandwidth you require the appropriate Q-factor of your coils as well - this gives you the copper resistance suitable - and further computation how to achieve these with give number of turns etc.
Are you comfortable with winding 200~300 turns through your toroids? I never did more than 50 turns that way.
Are you comfortable with winding 200~300 turns through your toroids? I never did more than 50 turns that way.
You might be better of with gapped ferrite pot cores - self-shielding and much easier to wind. You can buy them pre-gapped, which takes care of one big pain.
+1 for gapped ferrite - lower losses and generally better tolerances than MPP. Likely cheaper too. 156mH on a toroid I suspect will be a PITA to wind.
A high μ toroid with a lot of turns is going to saturate very easily. You have not said anything about the maximum current.
But if i did go with an MPP core with this high permeability, would it then affect the high frequency? What i understand is that the Q would be better, as the resistance in the wires would be minimal, due to the shorter length of the wire.
Q is better when the total wire length is shorter yes, but this assumes equal core losses. Gapped ferrite can be had with AL = 1000 (RM10 cores) which is significantly higher than your MPP (740).
I'd suggest you could achieve higher Q with a gapped RM10 than an MPP toroid. The disadvantage of AL=1000 is the 10% tolerance.
I'd suggest you could achieve higher Q with a gapped RM10 than an MPP toroid. The disadvantage of AL=1000 is the 10% tolerance.
Last edited:
Hi abraxalito,
Thank you for the info!
I see, i could get a better Q with RM10, but i think i would like to go with MPP just to follow some of the the original specs on the core, used in this design.
I found this table showing that a higher perm MPP core of 550u would actually not be suited for this, or am i wrong?
I found out as well that this coil would have to be wound in a special way like full 360° windings for each tap, plus reverse winding and
double reverse winding to limit capacitive losses.
If someone has any info on the winding techniques used to make this kind of coil would be appreciated a lot!
Best regards 🙂
According to that table, your 550u core (AL=740) is only recommended for making inductors working up to 6kHz. I'm puzzled why they'd stipulate such a low frequency - it could be due to core losses or it could be due to parasitic capacitance of the windings. My guess is more likely the former than the latter.
That would be a good reason to switch to a gapped RM core I'd think. Based on my experience in making inductors with gapped P cores (14 and 18mm diameter) they're perfectly useful over the entire audio band below 200mH or so. In the high hundreds of mH, the SRF starts to get within an octave of the top of the band.
It could be that they are thinking high-Q circuits.
By RF standards, even narrow audio filters are low-Q. Mostly around 1. Rarely over 3.
Note a minor economy. If you use a "too good" core, your copper has few turns and low resistance. You may have to ADD resistance to get the Q down. Yes, a resistor is 12 cents so not a thing in DIY..... in mass production the pennies add-up and the foresighted designer will tight-pencil this if possible (to delay company bankruptcy).
But its worth checking, especially with high turns counts, and remembering that current will increase with decreasing frequency and most AF energy is in the low frequencies...
I'm in the need of making few pot core inductors for my EQ too, but I am not sure how to choose cores and calculate windings. I have found those TDK cores: https://product.tdk.com/system/file.../ferrite-acc/data_sheet/80/db/fer/p_22_13.pdf. It says that they produce them of various materials, amongst which T38 has AL value of 16000! So it makes them very nice choice for this kind of inductors, if I understand those datasheets well. I need inductors with values of 1.5H and 3.0H, so I calculated that I need 306 windings for the first one, or 433 for the second one (http://www.learningaboutelectronics.com/Articles/Ferrite-core-inductor-calculator.php#answer). Is this correct, or I have missed something? And I should use .15mm wire, to achieve better frequency response. Is it OK to calculate Q=2*PI*f*L/R?
Last edited:
Materials like T38 are intended for applications where µ has to be high, but not particularly accurate, like common mode chokes.
They have poor stability, high losses and their linearity isn't very good either.
Anyway, you should never rely on the raw µ: it is mostly indicative, and the tiniest of parasitic gap will easily halve it.
Use a ferrite having a moderate µ (500 ~ 2000), intended for filter applications and use a gap sufficient to reduce µ apparent to <250.
You will need many more turns, but that's the price to be paid for a good quality inductor
They have poor stability, high losses and their linearity isn't very good either.
Anyway, you should never rely on the raw µ: it is mostly indicative, and the tiniest of parasitic gap will easily halve it.
Use a ferrite having a moderate µ (500 ~ 2000), intended for filter applications and use a gap sufficient to reduce µ apparent to <250.
You will need many more turns, but that's the price to be paid for a good quality inductor
So you suggest to use M33 (AL=1600, u=640) or N48 (AL=3800, u=1520)material cores with a gap? May I use ungapped, because there is very low current and no DC?
I need them for an EQ section of a bass amp, so maybe they need not to be so perfect?
Last edited:
With an ungapped inductor your tolerance will be in the region of +/- 30%, assuming you achieve very close contact between the halves. With a gap as Elvee suggests you'll potentially have +/-3%. For your particular high value inductors I'd suggest a gap which gives AL=400.
N48 would be my choice
It depends on what level of imperfection you tolerate: if the value doesn't need to be accurate, you might dispense with the gap (and the turns). At low levels, the THD of the raw ferrite will be in the tenths of percent range
Check the temperature dependence of the material's permeability - one of the many advantages of gapped inductors is the lower tempco as the permeability of the gap is constant with temperature, so if the gap dominates it stabiliizes the thermal variations.
I actually had in mind to use only one half of the core. Is it OK to do it like that, or I was wrong?
Last edited: