serial inductors
- Thread starter brianvino
- Start date
Help Support GroupDIY Audio Forum:
What are you trying to do? Apparently you want to make a 2H tapped coil with taps at 1.2 and 0.7H, but the math doesn't work. With the largest inductor, 47mH, the first tap works out at 0.47H.brianvino said:
Would be MUCH easier and better to wind some wire onto a pot core.
If you go for e.g. a P30x19 core made with the T38 material (Al=28000nH), some 280 turns of wire (0.1 or 0.2mm) would get you to 2H. That should be handwindable alright - and with much lower resistance..
https://www.elfaelektronik.dk/elfa3~dk_en/elfa/init.do?item=58-619-43&toc=20241
https://www.buerklin.com/default.asp?event=ShowArtikel%2882D464%29&l=d&jump=ArtNr_82D464
http://dk.mouser.com/ProductDetail/EPCOS/B65701W0000Y038/?qs=sGAEpiMZZMvgExXaNlWje9LtVyHWJuPKPj1bWrCVh9c%3d
Jakob E.
If you go for e.g. a P30x19 core made with the T38 material (Al=28000nH), some 280 turns of wire (0.1 or 0.2mm) would get you to 2H. That should be handwindable alright - and with much lower resistance..
https://www.elfaelektronik.dk/elfa3~dk_en/elfa/init.do?item=58-619-43&toc=20241
https://www.buerklin.com/default.asp?event=ShowArtikel%2882D464%29&l=d&jump=ArtNr_82D464
http://dk.mouser.com/ProductDetail/EPCOS/B65701W0000Y038/?qs=sGAEpiMZZMvgExXaNlWje9LtVyHWJuPKPj1bWrCVh9c%3d
Jakob E.
I agree, and it would be much more DIY.gyraf said:
It is tempting to use the higher Al cores, but these are ungapped, with a rather high tolerance (-30/+40%), which doesn't make them suitable for reproductible performance. In addition, the T38 material has a magnetisation curve that makes the inductance fluctuate with level. That's the reason other ferrite formulations are generally preferred in tuned filters, typically N41. This type of formulation, in combination with a gapped core, requires a much larger number of turns (10-15 times more for a given core size), but in the end, the results are better.
wind each tap as an individual coil for less leakage C.
takes less time in the long run as you do not have the other sections messing with the error in turns, which requires constant rewinding,
wind 300 turns, check, wind 10 turns and your done.
wind 500 turns, check, un wind 7 turns and you are done with tap 2.
other way is wind 100 turns, wind 50 turns, rewind 100 turn tap because 50 turn tap added somehow screwed up inductance of first coil,
imagine doing this all day, see what i'm sayin?
multiple coils better than stock, better Q, you can use bigger wire for the small coils,
companies thought it was cheaper to use 1 coil, less space, lower the parts count, but at a slight price in high end and phase shift,
takes less time in the long run as you do not have the other sections messing with the error in turns, which requires constant rewinding,
wind 300 turns, check, wind 10 turns and your done.
wind 500 turns, check, un wind 7 turns and you are done with tap 2.
other way is wind 100 turns, wind 50 turns, rewind 100 turn tap because 50 turn tap added somehow screwed up inductance of first coil,
imagine doing this all day, see what i'm sayin?
multiple coils better than stock, better Q, you can use bigger wire for the small coils,
companies thought it was cheaper to use 1 coil, less space, lower the parts count, but at a slight price in high end and phase shift,
yes, check or measure the inductance after making changes to the turns,
overlapped windings?
you can bunch wind torroids,
four pies, located at N S E W,
this will cut down on leakage between windings,
overlapped windings?
you can bunch wind torroids,
four pies, located at N S E W,
this will cut down on leakage between windings,
Leakage capacitance is a problem...or not?
It is a problem when the intrinsic resonance is interfering with the primary role of the inductor.
Let's say you use the inductor with a cap in parallels, as in some fixed equalizers, the leakage capacitance will only shift the resonant frequency by a fraction; it is then easy to compensate by adjusting the parallel cap.
But in most audio applications, inductors are used in a series arrangement, where the intrinsic resonance creates an unwanted twist in the frequency response.
One must remember that the highest the impedance the inductor is to work in, the higher the inductance, the higher the number of turns and the higher the leakage capacitance, and the lower and the most pronounced the parasitic resonance. Most solid-state design do not put much constraints on the design of inductors, but vacuum-tube circuits are a challenge in this respect.
Designers of tube EQ's have to resort to "violent" techniques in order to minimize the node impedance in their designs.
It is a problem when the intrinsic resonance is interfering with the primary role of the inductor.
Let's say you use the inductor with a cap in parallels, as in some fixed equalizers, the leakage capacitance will only shift the resonant frequency by a fraction; it is then easy to compensate by adjusting the parallel cap.
But in most audio applications, inductors are used in a series arrangement, where the intrinsic resonance creates an unwanted twist in the frequency response.
One must remember that the highest the impedance the inductor is to work in, the higher the inductance, the higher the number of turns and the higher the leakage capacitance, and the lower and the most pronounced the parasitic resonance. Most solid-state design do not put much constraints on the design of inductors, but vacuum-tube circuits are a challenge in this respect.
Designers of tube EQ's have to resort to "violent" techniques in order to minimize the node impedance in their designs.
Well.... winding thousand turns manually over a toroid core is not a pushover...maybe he could try reverse winding or insulating tape for each tap on a pot core if leakage capacitance is a problem but I though it was a problem in RF (hence the basket winding)...are we talking about some pF ? http://d4magnetics.com/page6/page6.html
I don't get it...if you don't use the mutual inductance property of two winding coils on the same core then you gonna need more turns for the second for the same value (L proportional to N²) then more capacitance leakage.
I don't get it...if you don't use the mutual inductance property of two winding coils on the same core then you gonna need more turns for the second for the same value (L proportional to N²) then more capacitance leakage.
Reverse winding or insulating tape won't help about leakage capacitance; if you want to reduce leakage cap with insulating tape, it must be quite thick, then the coupling factor will decrease, more turns will be needed and leakage inductance will increase too. Reverse-winding may make things worse even, if the end of the coil is close to the start.Mzaar said:
The leakage capacitance will be distributed differently. When everything is wound on a single core, all the elements are lumped together, primary inductance and capacitance. That creates a resonance at at given frequency. Splitting the inductor in two halves results in two smaller inductances and two smaller capacitances, each one having a higher resonant frequency. That may just be the ticket for shifting the resonance outside the useful range.
shabtek
Well-known member
is CJ referring to multiple cores and coils; or a chambered bobbin affair
i was talikn about a toroid core inductor,
for some reason, leakage C is a bigger problem with the toroid cores,
probably due to votage between windings being high when you come back around the circle to meet the first leads, where as in a bobbin wound coil, the last turn is further away from the first turn because of the layers,
not just a few pf, more like nf, enuff C to completely wipe out the right hand side of a bell curve,
take a 10 KC inductor.
it might work from 5 to 15 KC when forming the bell curve,
well, the 5 KC to 10 KC might ramp up just fine, but unless you put a scope on the equalizer, it is hard to tell if the coil is ramping down the high end or not,
so you might say, "gee, my home rolled inductor is working in my eq circuit"
but what is really happening is the cap is rolling off the low end so you hear that, but the high end roll off is not there, so you think the eq is working, but you are actually only obtaing a shelf curve instead of the bell curve due to the leaky inductor, but a shelf curve sounds like it is doing somehting, so you get fooled, get me?
see, the cap handles the left side of the bell curve, but the inductor handles the right,
this is for a series resonant circuit,
so the cap will be fine, no problems with a commercial cap, but a home grown inductor might start to roll back up due to leakage
i have wound coils that i thought were fine, put them in circuit, and found that the high end roll off was just a straight line, my ears were rolling off the high end, not the equalizer.
so check any coils you wind for XL at the highest freq it will handle,
anything over about 500 turns and you will see leakage problems,
so i used a big core for the low end, thus keeping the turns down,
for the high end you can use a smaller core as the millihenries required will be quite small for a 10 KC inductor.
so a smaller core can be used, which cuts down on leakage.
so if you customize each inductor, you can optimize the performance by improving the bell curve for the higher freqs,
OT: what kind of fuzz box on Spirit in the Sky?
for some reason, leakage C is a bigger problem with the toroid cores,
probably due to votage between windings being high when you come back around the circle to meet the first leads, where as in a bobbin wound coil, the last turn is further away from the first turn because of the layers,
not just a few pf, more like nf, enuff C to completely wipe out the right hand side of a bell curve,
take a 10 KC inductor.
it might work from 5 to 15 KC when forming the bell curve,
well, the 5 KC to 10 KC might ramp up just fine, but unless you put a scope on the equalizer, it is hard to tell if the coil is ramping down the high end or not,
so you might say, "gee, my home rolled inductor is working in my eq circuit"
but what is really happening is the cap is rolling off the low end so you hear that, but the high end roll off is not there, so you think the eq is working, but you are actually only obtaing a shelf curve instead of the bell curve due to the leaky inductor, but a shelf curve sounds like it is doing somehting, so you get fooled, get me?
see, the cap handles the left side of the bell curve, but the inductor handles the right,
this is for a series resonant circuit,
so the cap will be fine, no problems with a commercial cap, but a home grown inductor might start to roll back up due to leakage
i have wound coils that i thought were fine, put them in circuit, and found that the high end roll off was just a straight line, my ears were rolling off the high end, not the equalizer.
so check any coils you wind for XL at the highest freq it will handle,
anything over about 500 turns and you will see leakage problems,
so i used a big core for the low end, thus keeping the turns down,
for the high end you can use a smaller core as the millihenries required will be quite small for a 10 KC inductor.
so a smaller core can be used, which cuts down on leakage.
so if you customize each inductor, you can optimize the performance by improving the bell curve for the higher freqs,
OT: what kind of fuzz box on Spirit in the Sky?
