attenuator inductor emf trouble

[Chuck H]

The out-of-phase thing was just a thought, if it works for pickups, I don't see why it shouldn't for attenuator chokes.

My thinking was that a humbucker has two coils "seeing" the same EMI and since they are out of phase, they cancel. But can two "sources" of EMI cancel if they are out of phase? That seemed unlikely to me, which is why I disregarded my consideration. But I don't really know much about it.

The dimmer chokes I'm thinking of are known in the industry, I believe, as "storage chokes", "suppression chokes", or just "dimmer chokes". Either way they're just inductors with toroidal iron powder cores. Though I've seen some with ferrite rod cores, which would get you back where you started.

Almag light dimmer rise-time and emc chokes
DJ - Liniar Storage Chokes - www.nordtek.se

Excellent. I did a search and found some iron powder core inductors that will fit the bill. At least for the offending inductor. Mouser carries them for about $4.00 US.

Note, don't confuse these with the common-mode chokes you asked about a while ago. Those look similar but are on ferrite cores that can hardly store any energy.

Got it. I managed that distiction after you were good enough to help with my earlier post.

Thanks Steve.

Chuck

[Steve Conner]

Hi Chuck

In RF engineering, there is a principle called "reciprocity" or suchlike which says that the directional properties of an antenna will be the same when it's receiving as when it's transmitting. The implication of this is that coils "source" EMI in exactly the same way as they "see" it, so my idea should work.

[Enzo]

Certainly two speakers - as sources of soundwaves - will cancel when oriented out of phase with each other. I have no trouble envisioning a coil sending out a field with another coil next to it sending out a similar field that is reversed in polarity cancelling each other.

[Chuck H]

Great. Makes sense. The point about the speakers is a real world audible for the principle I can relate to.

I'll try both. The iron powder toroid and the series/xphase plain iron core. If they both work I can choose whichever is more cost effective.

Thanks

Chuck

[Chuck H]

It's Alive. BWooaHA HA HAaa...

I had two air core inductors on hand of half the uH for the troubled inductor. Anxious for answers I wired them up series/Xphase and stacked them one atop the other...

The problem is fixed That worked. Even with the air core inductors. WOO HOO.

I'll probably change to iron core inductors or at least try the toroid to cut a little cost. But I'm elated the the problem could be solved.

Thanks again.

Chuck

[Enzo]

Great.

ANother example occurs: They make noise cancelling headphones that have essentially a mic on the outside to pick up ambient noise, then a circuit amplifies that noise and adds it back into your audio stream at reverse polarity to the real sound waves. Thus the noise is cancelled leaving your signal all you hear. I guess that is just a special case of the speakers out of phase example.

[Chuck H]

Schematic

To be decent I feel I should post the schematic since I had alot of help here. I know that no one here is going to buy one anyway

This will be the "built in" unit. There will eventually be a "stand alone" 100W unit with a 4/8/16 ohm impedance range and seperate switches that bridge the active elements for "high" and "low" instead of just the "excite" switch. As you can see this switch just bypasses the active elements. But it seems to sound better at very high attenuation levels.

I wanted to solve four common attenuator problems with this design. First, the most popular attenuators don't have impedance switching. Second, some attenuators present a less than desirable load to the amp at different settings. This design swings about 20% +/-. Third, many attenuators aren't continuously variable. This one is. From about -3db off bypass down to about half a watt. And fourth, attenuators are too damned expensive. I wanted a simple, eloquent design that would solve these other problems so it could be made affordable.

The only rub is that rheostats are a little glitchy sounding at either end. I tried to lessen this with the .5R in series with the rheostat. It helps but it doesn't solve the problem completely.

I hope someone builds one. It really works well and is very transperant.

Chuck

[dai h.]

hey Chuck, thanks for reporting your findings(and Steve C. for the interesting idea). Haven't "tinkered" with my Power Brake guinea pig thing but have thought of some potential ideas for a bit of tweakability, namely a fixed R in series with a rheostat in place of the fixed R in parallel with the LF and HF boosting parts (*if* my thinking is correct--the fixed one so the current handling of the rheostat doesn't have to be so high and thus it wouldn't need to be as big and pricey, plus maybe it doesn't need to go all the way to short, i.e. zero ohms to be useful). Also (idea from a poster on plexi palace bbs--some old post I had saved), when attenuating drastically, trying with 10 inch spks. since the attenuated sound seems to work better with less LF(from the L plus C pair), so maybe just not have spkrs. with the extra lows in the first place (is the idea). Also, thought about replacing the C of the L/C pair with a film instead of alu electrolytic for longevity. Wonder if film alone would sound different also (have tried extra uF using film in addition to the stock alu electro. but not alone). Also, FWIW there was some stuff I remember reading about alu electros and there being non-etched and etched (foil) types and IIRC the non-etched were the ones to use for spk. crossovers since they handled... (wait let me just find it on my desktop, lol).

okay...

(link apparently dead)

http://www.leggott.co.uk/electronics...capacitor.html

1. Aluminium Electrolytic

There are two types of aluminium capacitor, the plain foil type and the etched foil type. The thickness of the aluminium oxide film and high breakdown voltage give these capacitors very high capacitance values for their size. The etched foil type differs from the plain foil type in that the aluminium oxide on the anode and cathode foils has been chemically etched to increase its surface area and permittivity. This gives a smaller sized capacitor than a plain foil type of equivalent value but has the disadvantage of not being able to withstand high a.c. currents compared to the plain type. Etched foil electrolytics are best used in coupling, d.c. blocking and by-pass circuits while plain foil types are better suited as smoothing capacitors in power supplies. Typical values of capacitance range from 1uF to 47000uF

...and some of the ones I've seen sold for crossover use (green IT brand) were sold as non-etched plus seemed larger than usual (less volumetric efficiency compared to etched) so seems to make sense.

[Chuck H]

ooops

Hey!?! Wait a minute...

The series/Xphase thing has a flaw. Any EMF cancelled in the surrounding area is also cancelled in the coils due to mutual inductance, no? I'm remembering an old Frankenstien's lab looking device I saw at a ham show once that had two inductors, one could be rotated inside the other to create a variable inductor. In phase the inductance would add and out of phase the inductance would null. I'll bet two beer that later today when I put a switch across my series inductor arrangement that I'll find it's doing very little or nothing... Sigh.

Well, I was happy for a day anyway. And that's worth something. And I still have options. Maybe a copper foil sandwich with an insulator in between the two inductors would keep them from talking directly to each other but still cancel most of the stray EMF. And there's still the iron powder toroid to try.

Chuck

Edit update: I put a switch across my series/Xphase inductor arrangement and found that it's doing virtually nothing. As if there were a much lower uH value inductor there. Back to the drawing board.

[Steve Conner]

Yep, that's the downside. The magnetic field is what gives rise to inductance, so attempts to cancel it will reduce the inductance. The good news is that the cancellation will be far from perfect, so you could just use bigger inductors to start with.

Or you may be able to find a geometry that just packs the field into a defined space instead of cancelling it, the toroid being the best example.

[Enzo]

It really works well and is very transperant.

SO if it works well and is very transparent, does it matter that your linked coils are not apparently doing anything? The results tend to indicate they are not needed then.

[Chuck H]

Well... The judgement of how it sounds was formed when I had the single inductor in place and it was misbehaving with my single coil pickups.

I was so exited that the problem had stopped with the new inductor arrangement that I didn't listen very hard to other aspects of performance yet. But after playing through it a bit I thought it sounded a little "off". That's when I instigated the test.

That's what happened.

But it does sound good without the second inductor circuit. A bit better at high attenuation levels actually. And I've thought about just leaving it out. But with just a small amount of attenuation it sounds better with the full circuit. This is why the more popular attenuators have high and low EQ circuits.

And after much consideration I've decided to damn it all and roll my own cheap inductors on mild steel rod. No matter what I use it will have to copied later anyhow. Even if it's a MFGd product. I think the main problem with the EMF could be my use of a very small case for this unit. The cap and both inductors are basically crammed in there, so layout could be improved. That and the air core inductor. I'm considering that speakers, after all, are iron core inductors. And yet my single coil pickups don't have a problem with speakers. No, I'm definitely over thinking this. I just need to idealize layout and use all iron core inductors. Betcha two beer it works fine after that. If not I'll go with toroidal inductors and factor in the extra cost.

Thanks again Enzo, Steve and everyone.

Chuck

[Chuck H]

Update...If anyone cares

Well I haven't ordered any iron core inductors yet due to budget constraints . But I did come up with an idea for the series/xphase thing I've already tried. I put an insulator between the inductors with a layer of copper foil on either side. So it's:

inductor/insulator/copper/insulator/copper/insulator/inductor (out of phase)

The results are very promising. I have about 60% effectiveness from the inductors and no whistle. I guess 60% because I put the original value single inductor in temporarily as a control sample. So if I up the value of the series inductors a little it should work fine. So if anyone still cares, this was a solution after all. It just needed the shields to prevent so much close proximity coupling

I'll still try some less expensive options for cost reasons. But this baby works

Chuck