On Tue, 20 Aug 2019 14:06:17 -0500, amdx <nojunk@knology.net> wrote:>On 8/20/2019 11:41 AM, legg wrote: >> On Mon, 19 Aug 2019 19:53:28 -0500, amdx <nojunk@knology.net> wrote: >> >>> I'd like to have a Variable 100uh inductor, that is controlled to full >>> saturation with dc through the 100uh coil. >>> I see some two winding coils, but the dc control winding has huge >>> inductance. Working range 500kHz to 4MHz. >>> Is this possible? >>> >>> Thanks for your thoughts, Mikek >> A cpnventional current-controlled inductor has two magnetic paths that >> are configured to carry antiphase windings, with the third 'inductor >> winding sharing both. The control windings' AC flux, from the third >> winding, cancels out. >> >> You can do this with an e-core (control windings on outer arms), two >> toroids (each with control winding - sandwitched to support single >> 'inductor' winding around both) or solenoids (two solenoid cores >> support control windings, 'inductor' wound around both.). >> >> RL >> > > As state previously, My question was not ask well, I only want one >winding with both ac and the dc to saturate it. > > Thanks, MikekAn ideal current source is easy to simulate, but in real life, it's voltage compliance limits, stray loading and real power loss will set limits on what you can do. The conventional multiple winding is the more realistic approach. Parasitics from the lower-powered control circuit are also easier to isolate. RL
How to design an Interesting inductor
Started by ●August 19, 2019
Reply by ●August 21, 20192019-08-21
Reply by ●August 21, 20192019-08-21
On 2019-08-21, legg <legg@nospam.magma.ca> wrote:> On Tue, 20 Aug 2019 14:06:17 -0500, amdx <nojunk@knology.net> wrote: > >>On 8/20/2019 11:41 AM, legg wrote: >>> On Mon, 19 Aug 2019 19:53:28 -0500, amdx <nojunk@knology.net> wrote: >>> >>>> I'd like to have a Variable 100uh inductor, that is controlled to full >>>> saturation with dc through the 100uh coil. >>>> I see some two winding coils, but the dc control winding has huge >>>> inductance. Working range 500kHz to 4MHz. >>>> Is this possible? >>>> >>>> Thanks for your thoughts, Mikek >>> A cpnventional current-controlled inductor has two magnetic paths that >>> are configured to carry antiphase windings, with the third 'inductor >>> winding sharing both. The control windings' AC flux, from the third >>> winding, cancels out. >>> >>> You can do this with an e-core (control windings on outer arms), two >>> toroids (each with control winding - sandwitched to support single >>> 'inductor' winding around both) or solenoids (two solenoid cores >>> support control windings, 'inductor' wound around both.). >>> >>> RL >>> >> >> As state previously, My question was not ask well, I only want one >>winding with both ac and the dc to saturate it. >> >> Thanks, Mikek > > An ideal current source is easy to simulate, but in real life, it's > voltage compliance limits, stray loading and real power loss will set > limits on what you can do. > > The conventional multiple winding is the more realistic approach. > > Parasitics from the lower-powered control circuit are also easier to > isolate. > > RLAIUI this is for a recieve antenna so the AC signal will be failrly small, I expect that the current source only needs enough compliance to overcome earth resistance and a little headroom. -- When I tried casting out nines I made a hash of it.
Reply by ●August 21, 20192019-08-21
On 21.8.19 05:22, Tim Williams wrote:> "amdx" <nojunk@knology.net> wrote in message > news:qjhbo9$uq0$1@dont-email.me... >> �Man, what time did I write the above? >> What I meant to say was, >> I had a BOG before and used a bedds that had a Q of 1 or 2. >> � so if I could get 20 it would be better. > > I still don't know what you're writing.� It ain't English... > > Well, "bog" aside.� But that means something different in English > English. ;-) > > >> >> It's for a receive only antenna, so microwatts, maybe milliwatts. >> 500kHz to 4 MHz. > > What about just low noise JFETs?� It's not like you need any antenna > gain in those bands. > > >> �Another possible solution, >> �Is there a combination of L's and Cs the would approximate such an L >> reduction over frequency? > > Sure.� But not just L and C, R is required.� Necessarily has a very low > Q (from Kramers-Kronig relations).� Ferrite beads for example. > > TimThe OP is attempting to make a tuned circuit with transductor tuning. The linear equations of a tuned circuit (here: an antenna system) fall apart as soon as the reactive element (transductor coil) is saturated into non-linear operating range for control of effective inductance. I just wonder how much unwanted crud such a system will radiate on un-intended freqiencies when a bunch of strong local stations meet the saturating inductor. There ain't such thing as a good MF antenna tuner in small size. -- -TV
Reply by ●August 21, 20192019-08-21
On 8/21/2019 1:55 AM, Jasen Betts wrote:> On 2019-08-21, legg <legg@nospam.magma.ca> wrote: >> On Tue, 20 Aug 2019 14:06:17 -0500, amdx <nojunk@knology.net> wrote: >> >>> On 8/20/2019 11:41 AM, legg wrote: >>>> On Mon, 19 Aug 2019 19:53:28 -0500, amdx <nojunk@knology.net> wrote: >>>> >>>>> I'd like to have a Variable 100uh inductor, that is controlled to full >>>>> saturation with dc through the 100uh coil. >>>>> I see some two winding coils, but the dc control winding has huge >>>>> inductance. Working range 500kHz to 4MHz. >>>>> Is this possible? >>>>> >>>>> Thanks for your thoughts, Mikek >>>> A cpnventional current-controlled inductor has two magnetic paths that >>>> are configured to carry antiphase windings, with the third 'inductor >>>> winding sharing both. The control windings' AC flux, from the third >>>> winding, cancels out. >>>> >>>> You can do this with an e-core (control windings on outer arms), two >>>> toroids (each with control winding - sandwitched to support single >>>> 'inductor' winding around both) or solenoids (two solenoid cores >>>> support control windings, 'inductor' wound around both.). >>>> >>>> RL >>>> >>> >>> As state previously, My question was not ask well, I only want one >>> winding with both ac and the dc to saturate it. >>> >>> Thanks, Mikek >> >> An ideal current source is easy to simulate, but in real life, it's >> voltage compliance limits, stray loading and real power loss will set >> limits on what you can do. >> >> The conventional multiple winding is the more realistic approach. >> >> Parasitics from the lower-powered control circuit are also easier to >> isolate. >> >> RL > > AIUI this is for a recieve antenna so the AC signal will be failrly > small, I expect that the current source only needs enough compliance > to overcome earth resistance and a little headroom. >It is a receive only antenna. And if you are suggesting a dc return through the earth, that was not part of my plan. I think electrolysis would be a problem.. So a two wire dc system. Mikek
Reply by ●August 21, 20192019-08-21
On 8/21/2019 3:26 AM, Tauno Voipio wrote:> On 21.8.19 05:22, Tim Williams wrote: >> "amdx" <nojunk@knology.net> wrote in message >> news:qjhbo9$uq0$1@dont-email.me... >>> �Man, what time did I write the above? >>> What I meant to say was, >>> I had a BOG before and used a bedds that had a Q of 1 or 2. >>> � so if I could get 20 it would be better. >> >> I still don't know what you're writing.� It ain't English... >> >> Well, "bog" aside.� But that means something different in English >> English. ;-) >> >> >>> >>> It's for a receive only antenna, so microwatts, maybe milliwatts. >>> 500kHz to 4 MHz. >> >> What about just low noise JFETs?� It's not like you need any antenna >> gain in those bands. >> >> >>> �Another possible solution, >>> �Is there a combination of L's and Cs the would approximate such an L >>> reduction over frequency? >> >> Sure.� But not just L and C, R is required.� Necessarily has a very >> low Q (from Kramers-Kronig relations).� Ferrite beads for example. >> >> TimI appreciate your input but not sure I understand.> The OP is attempting to make a tuned circuit with transductor tuning.The circuit is a classic tuned circuit, it is not at resonance.> > The linear equations of a tuned circuit (here: an antenna system) fall > apart as soon as the reactive element (transductor coil) is saturated > into non-linear operating range for control of effective inductance.I'm Sorry, I don't have any understanding of the "non-linear operating range". Is this like a diode causing mixing of two RF signals creating other products? If that is the case, will millivolt level signals in a non-linear inductor create objectionable mixing? Aren't all ferrite inductors non-linear, even in normal operation? Thanks, Mikek> I just wonder how much unwanted crud such a system will radiate on > un-intended freqiencies when a bunch of strong local stations meet > the saturating inductor. > > There ain't such thing as a good MF antenna tuner in small size. >
Reply by ●August 21, 20192019-08-21
On Wednesday, August 21, 2019 at 11:16:41 PM UTC+10, amdx wrote:> On 8/21/2019 3:26 AM, Tauno Voipio wrote: > > On 21.8.19 05:22, Tim Williams wrote: > >> "amdx" <nojunk@knology.net> wrote in message > >> news:qjhbo9$uq0$1@dont-email.me...<snip>> I'm Sorry, I don't have any understanding of the "non-linear operating > range". Is this like a diode causing mixing of two RF signals creating > other products? > If that is the case, will millivolt level signals in a non-linear > inductor create objectionable mixing? > Aren't all ferrite inductors non-linear, even in normal operation?Sure but they are mostly built with gapped ferrites, and the harmonic products tend to be 80 or 90dB below the fundamental. They get a lot worse when driven close to saturation. LTSpice offers the John Chan model to work out how hysteresis actually affects the voltage-current relationship - just for inductors as opposed to coupled inductors and transformers. It gave the right answer, the one time I used it. -- Bill Sloman, Sydney
Reply by ●August 21, 20192019-08-21
On Wed, 21 Aug 2019 07:59:39 -0500, amdx <nojunk@knology.net> wrote:>On 8/21/2019 1:55 AM, Jasen Betts wrote: >> On 2019-08-21, legg <legg@nospam.magma.ca> wrote: >>> On Tue, 20 Aug 2019 14:06:17 -0500, amdx <nojunk@knology.net> wrote: >>> >>>> On 8/20/2019 11:41 AM, legg wrote: >>>>> On Mon, 19 Aug 2019 19:53:28 -0500, amdx <nojunk@knology.net> wrote: >>>>> >>>>>> I'd like to have a Variable 100uh inductor, that is controlled to full >>>>>> saturation with dc through the 100uh coil. >>>>>> I see some two winding coils, but the dc control winding has huge >>>>>> inductance. Working range 500kHz to 4MHz. >>>>>> Is this possible? >>>>>> >>>>>> Thanks for your thoughts, Mikek >>>>> A cpnventional current-controlled inductor has two magnetic paths that >>>>> are configured to carry antiphase windings, with the third 'inductor >>>>> winding sharing both. The control windings' AC flux, from the third >>>>> winding, cancels out. >>>>> >>>>> You can do this with an e-core (control windings on outer arms), two >>>>> toroids (each with control winding - sandwitched to support single >>>>> 'inductor' winding around both) or solenoids (two solenoid cores >>>>> support control windings, 'inductor' wound around both.). >>>>> >>>>> RL >>>>> >>>> >>>> As state previously, My question was not ask well, I only want one >>>> winding with both ac and the dc to saturate it. >>>> >>>> Thanks, Mikek >>> >>> An ideal current source is easy to simulate, but in real life, it's >>> voltage compliance limits, stray loading and real power loss will set >>> limits on what you can do. >>> >>> The conventional multiple winding is the more realistic approach. >>> >>> Parasitics from the lower-powered control circuit are also easier to >>> isolate. >>> >>> RL >> >> AIUI this is for a recieve antenna so the AC signal will be failrly >> small, I expect that the current source only needs enough compliance >> to overcome earth resistance and a little headroom. >> >It is a receive only antenna. >And if you are suggesting a dc return through the earth, that was not >part of my plan. I think electrolysis would be a problem.. > So a two wire dc system. > MikekI don't see such a restriction to be a valid design goal in a new product. What's the selling point? Electrolysis????? Even in a retrofit, the addition of an external field-generating winding might be easiest. A modest current source can be decoupled from a tuned/tuning circuit, over a known frequency range using a series choke. In small signal circuitry, it's an issue you'd try to avoid, as resistors do the job pretty well at low current. A tuned circuit with an ungrounded terminal might benefit from a re-examination of the schematic. Ground is, after all, everywhere. You should use that feature, rather than compound it's potential problems. If the aim is repeatable, calibratable 'L' values, I forsee a host of issues with this technique. Closed circuit, low frequency magnetics have to be re-evaluated for function, when located within close proximity to large magnets, but a magnet on a stick doesn't fit in well with typical solid-state circuit aims. RL
Reply by ●August 21, 20192019-08-21
On 21.8.19 16:16, amdx wrote:> On 8/21/2019 3:26 AM, Tauno Voipio wrote: >> On 21.8.19 05:22, Tim Williams wrote: >>> "amdx" <nojunk@knology.net> wrote in message >>> news:qjhbo9$uq0$1@dont-email.me... >>>> �Man, what time did I write the above? >>>> What I meant to say was, >>>> I had a BOG before and used a bedds that had a Q of 1 or 2. >>>> � so if I could get 20 it would be better. >>> >>> I still don't know what you're writing.� It ain't English... >>> >>> Well, "bog" aside.� But that means something different in English >>> English. ;-) >>> >>> >>>> >>>> It's for a receive only antenna, so microwatts, maybe milliwatts. >>>> 500kHz to 4 MHz. >>> >>> What about just low noise JFETs?� It's not like you need any antenna >>> gain in those bands. >>> >>> >>>> �Another possible solution, >>>> �Is there a combination of L's and Cs the would approximate such an >>>> L reduction over frequency? >>> >>> Sure.� But not just L and C, R is required.� Necessarily has a very >>> low Q (from Kramers-Kronig relations).� Ferrite beads for example. >>> >>> Tim > > �I appreciate your input but not sure I understand. > > >> The OP is attempting to make a tuned circuit with transductor tuning. > > The circuit is a classic tuned circuit, it is not at resonance. > >> >> The linear equations of a tuned circuit (here: an antenna system) fall >> apart as soon as the reactive element (transductor coil) is saturated >> into non-linear operating range for control of effective inductance. > > I'm Sorry, I don't have any understanding of the "non-linear operating > range". Is this like a diode causing mixing of two RF signals creating > other products? > �If that is the case, will millivolt level signals in a non-linear > inductor create objectionable mixing? > �Aren't all ferrite inductors non-linear, even in normal operation? > ��������������������������������������� Thanks, MikekYes, but they are not intentionally driven hard out of the linear range. The control in a transductor is based to driving the core into saturation with the control current and thus controlling the effective inductance. -- -TV
Reply by ●August 21, 20192019-08-21
On 8/21/2019 11:10 AM, legg wrote:> On Wed, 21 Aug 2019 07:59:39 -0500, amdx <nojunk@knology.net> wrote: > >> On 8/21/2019 1:55 AM, Jasen Betts wrote: >>> On 2019-08-21, legg <legg@nospam.magma.ca> wrote: >>>> On Tue, 20 Aug 2019 14:06:17 -0500, amdx <nojunk@knology.net> wrote: >>>> >>>>> On 8/20/2019 11:41 AM, legg wrote: >>>>>> On Mon, 19 Aug 2019 19:53:28 -0500, amdx <nojunk@knology.net> wrote: >>>>>> >>>>>>> I'd like to have a Variable 100uh inductor, that is controlled to full >>>>>>> saturation with dc through the 100uh coil. >>>>>>> I see some two winding coils, but the dc control winding has huge >>>>>>> inductance. Working range 500kHz to 4MHz. >>>>>>> Is this possible? >>>>>>> >>>>>>> Thanks for your thoughts, Mikek >>>>>> A cpnventional current-controlled inductor has two magnetic paths that >>>>>> are configured to carry antiphase windings, with the third 'inductor >>>>>> winding sharing both. The control windings' AC flux, from the third >>>>>> winding, cancels out. >>>>>> >>>>>> You can do this with an e-core (control windings on outer arms), two >>>>>> toroids (each with control winding - sandwitched to support single >>>>>> 'inductor' winding around both) or solenoids (two solenoid cores >>>>>> support control windings, 'inductor' wound around both.). >>>>>> >>>>>> RL >>>>>> >>>>> >>>>> As state previously, My question was not ask well, I only want one >>>>> winding with both ac and the dc to saturate it. >>>>> >>>>> Thanks, Mikek >>>> >>>> An ideal current source is easy to simulate, but in real life, it's >>>> voltage compliance limits, stray loading and real power loss will set >>>> limits on what you can do. >>>> >>>> The conventional multiple winding is the more realistic approach. >>>> >>>> Parasitics from the lower-powered control circuit are also easier to >>>> isolate. >>>> >>>> RL >>> >>> AIUI this is for a recieve antenna so the AC signal will be failrly >>> small, I expect that the current source only needs enough compliance >>> to overcome earth resistance and a little headroom. >>> >> It is a receive only antenna. >> And if you are suggesting a dc return through the earth, that was not >> part of my plan. I think electrolysis would be a problem.. >> So a two wire dc system. >> Mikek > > > I don't see such a restriction to be a valid design goal in a new > product. What's the selling point? Electrolysis????? Even in a > retrofit, the addition of an external field-generating winding might > be easiest. > > A modest current source can be decoupled from a tuned/tuning circuit, > over a known frequency range using a series choke. In small signal > circuitry, it's an issue you'd try to avoid, as resistors do the job > pretty well at low current. > > A tuned circuit with an ungrounded terminal might benefit from a > re-examination of the schematic. Ground is, after all, everywhere. You > should use that feature, rather than compound it's potential problems. > > If the aim is repeatable, calibratable 'L' values, I forsee a host of > issues with this technique. > > Closed circuit, low frequency magnetics have to be re-evaluated for > function, when located within close proximity to large magnets, but a > magnet on a stick doesn't fit in well with typical solid-state circuit > aims. > > RL >Start here.> https://www.w8ji.com/beverages.htmThen see Loaded beverages here,> https://www.w8ji.com/slinky_and_loaded_beverages.htmNow, I want to load a beverage on ground. I don't have enough room for BOG long enough to have good directionality in the BCB, therefore, I want to load it, reducing the VF and changing the pattern. If it is loaded to make it directional at 1MHz then it doesn't look good at 4 MHz, So I want variable inductors. It doesn't look good, the mixing products concerns me and finding a material or combination that I can cause to reduce inductance as frequency increases is not jumping out. Thanks, Mikek
Reply by ●August 21, 20192019-08-21
On 8/20/19 3:04 PM, amdx wrote:> On 8/20/2019 11:28 AM, Phil Hobbs wrote: >> On 8/20/19 4:41 AM, Clive Arthur wrote: >>> On 20/08/2019 01:53, amdx wrote: >>>> I'd like to have a Variable 100uh inductor, that is controlled >>>> to full saturation with dc through the 100uh coil. I see some >>>> two winding coils, but the dc control winding has huge >>>> inductance. Working range 500kHz to 4MHz. Is this possible? >>>> >>>> Thanks for your thoughts, Mikek >>> >>> Some RM cores and pot cores have a central hole through which you >>> could wind your 100uH, and the DC winding goes on the bobbin... >>> >>> https://uk.farnell.com/ferroxcube/rm10-i-3c95/ferrite-core-rm-i-3c95/dp/2103458?st=rm%20core >>> >>> >>> >>> ...this is a 5.5uH core so approx 4 turns for 100uH if it were >>> wound on the bobbin, I don't know how many if wound through the >>> hole, but I'd guess it would be similar. I'd try to grind the >>> hole edges to take off the sharp edge if possible. >>> >>> Cheers >> >> One minor tweak to that idea is to run an odd number of half-turns >> on the bobbin, and use a high-mu pot core with an adjustable centre >> gap. >> >> The extra half-turn will generate a large field going round the >> outside of the core (avoiding the post) and so will saturate it >> fairly readily. >> >> Cheers >> >> Phil Hobbs >> >> > Can you explain physically how that would look?With pot cores you usually twist the two leads together and bring them out through a slot in the side of the core. If there are two or more slots, the wire goes in through one slot and comes out through a second one, 180 degrees away.> My only experience with half turns is, we put a 4-1/2 turn inductor > in a medium power circuit. The coil over heated, 4 turns were fine, 5 > turns were fine, we never did a half turn inductor again. OH wait you > only mean a single half turn as we did above, so if I need 12 turns > make it 12-1/2 turns. Is that correct? Any way to do 1/2 turn on a > toroid?Nope. It's just the threading count. Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com
