EEVblog Electronics Community Forum
General => General Technical Chat => Topic started by: xzswq21 on April 26, 2021, 05:31:35 am
-
Hello
I have designed a 10KHz low pass filter so I need to build a 680uH Toroidal inductor, but I'm skeptical to use Ferrite core or Iron Powder core,
the input and output impedance of the system is 50 Ohm and the signal level is only 1Vpeak but the Q should be higher than 20 and DCR should be lower than 2 Ohm.
could you please help me to choose a suitable Material and core type?
This is the response of my 10KHz LPF with Q=20 and DCR=2 Ohm in simulation, the response is good enough for my application:
(https://www.eevblog.com/forum/chat/ferrite-or-iron-powder-toroid-core-for-a-low-pass-filter/?action=dlattach;attach=1214767)
I usually use "mini ring calculator" to design Toroidal inductors.
-
i recommend you buy both and have a few rings on hand to get a good feeling of how they work along with how the windings effect them. It's really gonna cut down on the bullshit you have to deal with if you get a few spools of wire and a few rings, it won't be more then like $40 and your gonna know a whole bunch of shit people don't touch ever. I made a inductor box and a magnet wire box that has a bunch of different gauges of magnet wire and cuttings of other wire in there so I can rapidly make a whole bunch of things. I also keep some tape in there.
how do you want to build your filter? CLC?
Seriously, getting your mind over inductor calculations is like hard and confusing. If you get a few materials that you actually know you will learn ALOT.
-
What's wrong with buying one, probably of ferrite bobbin/spool type?
What about distortion, is this something that needs to be really clean, or does it not matter much?
The Q more or less rules out the lossiest powdered iron (#26, #52) but the rest are fine. Something #8 or #18 or any of the brand names (Kool-Mu, Sendust, MPP, etc.) will do. Gapped ferrite of pretty much any sort will do.
Tim
-
how do you want to build your filter? CLC?
Yes, it is a 7th order CLC filter
What about distortion, is this something that needs to be really clean, or does it not matter much?
The Q more or less rules out the lossiest powdered iron (#26, #52) but the rest are fine. Something #8 or #18 or any of the brand names (Kool-Mu, Sendust, MPP, etc.) will do. Gapped ferrite of pretty much any sort will do.
Tim
Actually I want to attach the filter at the output of a signal generator to clean the signal and then apply this clean signal to a sensor, finally I will check the output distortion by using a Spectrum Analyzer :) in fact I want to report the linearity of some new designed sensors.
the signal generator's distortion is -50dBc@10KHz but I need -90dBc@10KHz! so I need to build a 10KHz Low pass filter to reduce the distortions
-
i think a 7th order filter will be difficult to get nice response.
what you want to do, since its a low frequency signal, is to buy a low distortion oscillator.
-
i think a 7th order filter will be difficult to get nice response.
what you want to do, since its a low frequency signal, is to buy a low distortion oscillator.
This is the response of the LPF with Q=20 and DCR=2 Ohm in simulation: (it's good)
(https://www.eevblog.com/forum/chat/ferrite-or-iron-powder-toroid-core-for-a-low-pass-filter/?action=dlattach;attach=1214767)
I only need to have two of 470uH and one 680uH inductors. DCR should be about 2 Ohm or lower and Q should be higher than 15~20
-
try realizing it though, those simulations are a bit deceptive, some people say I am putting this very lightly for such a high order filter
-
The main advantage if iron powder cores is a relatively high saturation magnetization. So this would be relevant for high power, high current, not for a relatively low power signal. In anyway the magnetic core material is potentially nonlinear. For linear response one may have to limit the magnetization to a relatively low value. Chancs are a ferrite core with air gap may be more linear. The added advanatge is that one can adjust the air gap in some shapes, though not so easy with toroids.
With ferrites one has some magnetostriction and this leads to a relatively sharp resonance(s) somewhere at the mechanical resonance(s) of the core. Compared to a LC resonance this can be with an unusual high Q (e.g. some 1000). One can dampen the resonance from the nechanical side, as least to some degree.
With a closed ferrite ring stress to the core can change the AL value a lot, especially in the more softer ferrites.
-
The main advantage if iron powder cores is a relatively high saturation magnetization. So this would be relevant for high power, high current, not for a relatively low power signal. In anyway the magnetic core material is potentially nonlinear. For linear response one may have to limit the magnetization to a relatively low value. Chancs are a ferrite core with air gap may be more linear. The added advanatge is that one can adjust the air gap in some shapes, though not so easy with toroids.
With ferrites one has some magnetostriction and this leads to a relatively sharp resonance(s) somewhere at the mechanical resonance(s) of the core. Compared to a LC resonance this can be with an unusual high Q (e.g. some 1000). One can dampen the resonance from the nechanical side, as least to some degree.
With a closed ferrite ring stress to the core can change the AL value a lot, especially in the more softer ferrites.
Thank you for your attentions and recommendations, so I looked at Mouser.com and found some inductors:
https://www.mouser.com/ProductDetail/EPCOS-TDK/B82145A1684J000?qs=MD70nW2XBGhko3S1rqCadg%3D%3D (https://www.mouser.com/ProductDetail/EPCOS-TDK/B82145A1684J000?qs=MD70nW2XBGhko3S1rqCadg%3D%3D)
(https://www.eevblog.com/forum/chat/ferrite-or-iron-powder-toroid-core-for-a-low-pass-filter/?action=dlattach;attach=1214923)
(https://www.eevblog.com/forum/chat/ferrite-or-iron-powder-toroid-core-for-a-low-pass-filter/?action=dlattach;attach=1214925)
another inductor:
(But unfortunately it's non-stocked)
https://www.mouser.com/ProductDetail/Fastron/XHBCC-681J-01?qs=R2jSSvul3fBkvVec%2FCk3UQ%3D%3D (https://www.mouser.com/ProductDetail/Fastron/XHBCC-681J-01?qs=R2jSSvul3fBkvVec%2FCk3UQ%3D%3D)
I've extracted the below pic from the datasheet:
(https://www.eevblog.com/forum/chat/ferrite-or-iron-powder-toroid-core-for-a-low-pass-filter/?action=dlattach;attach=1214919)
is it a good choice for the LPF? I think the behavior of L and Q vs Frequency is good.
-
Not too bad. Look for the next size or two up, if you find the Q isn't enough.
Tim
-
try realizing it though, those simulations are a bit deceptive, some people say I am putting this very lightly for such a high order filter
I agree. Simulations can give you a false sense of security regarding the performance of the filter unless your component simulation models very closely represent the performance of the real world components.
The very least you should do is perform another simulation using the Monte-Carlo technique incorporating the effect of all the component tolerances. With a high order elliptic-function filter or inverse-Chebyshev filter selection of close tolerance components is vital to achieve the desired response, particularly in the transition region between pass band and stop band.