[Chris Adams]

Haven't seen this here so I thought I would post it.

I can't say how well this works for every Maggie in every situation but on the AA it was a big hit. That's where I learned about it.

On your speakers where you would normally have a jumper to the tweeter, remove it and insert a Bourns 5522-RC or 5502-RC. Here's the basic info on the choke. http://www.bourns.com/data/global/pdfs/5500_series.pdf

If you have a resistor in that location, just remove it and attach one end to one end of the choke. Then put the free ends of the choke and the resistor in each jumper hole. So now the resistor and the choke are in series. If you have modded your speakers and have access to both woofer and tweeter, put one on each in series with the positive terminal. Actually, I don't think it makes a difference whether it's on the + or -. I have the 5502 on my tweet - and 5522 on the woofer +. The tweeter made the biggest difference, but noticed some additional improvement with the one on the woofer as well.

On my MMGs, I heard much better focus, more air around instruments and voices, smoother highs, better layering of instruments and improved detail. Cost me about $18 including shipping.

Newark has them on sale right now and they are cheep to begin with. A no brainer to try. http://www.newark.com/

[Emil]

I remember reading something about this over at AA.

What exactly does these chokes do?

[Elizabeth]

I have a set of chokes in position in my 3.6s.
High frequencies a little smoother with choke in place.

[*Scotty*]

This is a fix for SS amplifiers that do not have an adequate zobel network on their speaker outputs to stop RFI from getting back into their feedback loop. The the Maggies "voice-coils" behave like RF antennas and RFI can enter through the back-door of the feedback loop of the SS amplifier if it is not protected from this by a filter network. The 10 uH choke can help stop the RFI from getting back to the amplifiers outputs and entering the feedback loop.
 These chokes could be put directly in-line at the speaker terminals on the positive and negative speaker posts or built into the speaker wires in the same position.
 Generally the SS amplifier has a low pass network across the speaker terminals consisting of a capacitor in series with a resistor. This network is in parallel with the speaker load and is not directly in the signal path but is in a shunt position. If high quality parts are used in this application there is no sonic penalty from the filter.
 I would rather put a filter network on the output of the amplifier or increase the size of the capacitor in the existing network to move the filter pole downward rather than listen to the series inductor at the loudspeaker.
 A secondary zobel network could be placed directly across the speaker terminals and it could accomplish the same reduction in RFI as the series inductors.
 Scotty

[Davey]

Totally agree with Scotty....he posted while I was typing. 

I don't particularly like the choke method myself.  I believe if this is an issue, a better scheme is an RC shunt network (at the speaker end) optimized above the audio band.  This also has the advantage of terminating the speaker wires themselves (at the far end) and keeping them from possibly acting as antennas.

Cheers,

Dave.

[Chris Adams]

Scotty and Dave, thanks for the info.

[kevin360]

I tried it with my 3.7s and really didn't notice a difference. When I decided that my pico fuse experiment failed (too fast - nuisance opening once every few weeks), I also removed the chokes. That did make a small improvement - hard to describe, but it's enough that I think I've permanently dumped the safety net of fuses (that may not even protect the ribbons anyway).

[Emsquare]

Personally, I prefer some strategically placed ferrite cores where it makes sense. No connections are altered. The only parameter changed is some mild inductance added to the circuit. Far less invasive than the 'choke mod'. What worries me about the popularity of the chokes is how they skew the crossover values. Make no mistake, they do. If you love it then so be it but realize that they are might in series with another high value inductor. Maybe not such a great idea after all.

I tried them and it was far too aggressive of a mod with my speakers. I can see where it might sound like they clean up the high treble. So are they doing that or just rolling off some of the treble?

[Davey]

"Skew the crossover values"?  I don't think so.  You must have been experiencing some other variable aspect.

A 10uH inductor with a 4 ohm load creates a low-pass Fc of nearly 70khz.  Well outside of the speakers crossover range.

Some parasitic component of the inductor you're referring to?

Cheers,

Dave.

[Emsquare]

And where you have two inductors in series with each other? Do they not have an electrical equivalency? Some of the bass inductors are not all that large in value. 10uH isn't much but the bass inductor in the 2.6 is only 390uH I believe (.39mH). Negligible? Within the tolerance but precise? How about in series with the 2uf cap for the tweeter leg? Let's say I was actually using the internal crossover... I don't like it.

[Davey]

Have you done some actual measurements to confirm your theory?  Would a simulation convince you?

ONE of the bass inductors in a 2.6 is 0.7mH (700uH.)  (That's a good deal larger than the 10uH inductor mentioned here.)
The series capacitor in the 2.6 is 35uF. 

http://www.integracoustics.com/MUG/MUG/tweaks/mg2.6xo.gif

Cheers,

Dave.

[Emsquare]

You're right. I was being myopic in the way I was looking at it. I can write the whole explaination why I thought that but the short answer is I was having a senior moment.

That being said... I need to teach myself to use one of the circuit simulator software tools to get a better picture of things like this. National Semi has a number of free software downloads that look promising. And I have downloaded those, 5Spice and LTSpice from Linear Technologies. Any recommendations? I kind of loathe the thought of trying to figure those out but it should be worth it. Perhaps even find a place to do it with a few other people to cut the learning curve down.

[Davey]

If you're primarily interested in speaker-related simulations, I'd use the (free) demo copy of LspCAD.  It's limited, but you can still do quite a bit.

http://www.ijdata.com/LspCAD_demo.html

For more general usage, Circuitlab is a neat browser-based simulation tool that's fairly new.

https://www.circuitlab.com/

There are plenty of others.

Cheers,

Dave.

[kevin360]

Cool stuff 

I'm preparing to install new flooring, so I’m moving – essentially (a bunch of stuff is in the garage (it’s a big garage) for the next week, or so). I boxed the books yesterday – found my 2nd, 3rd & 4th editions of the Electronic Designer’s Handbook (along with several other ‘cookbooks’ and a few textbooks) from when I was more in touch with this stuff (feels like a lifetime ago). I felt like such a schlubb as I looked over the math that was once practically 2nd nature. Thank goodness for computers – and the kind assistance of generous folks like you, Dave.
 
It’s funny how large our living room looks now – without half the furniture and all of the accoutrements. The garage, on the other hand, is getting a bit tight.

[Emsquare]

Well isn't that useful? Thanks Davey.

[dawnrazor]

This is a fix for SS amplifiers that do not have an adequate zobel network on their speaker outputs to stop RFI from getting back into their feedback loop. The the Maggies "voice-coils" behave like RF antennas and RFI can enter through the back-door of the feedback loop of the SS amplifier if it is not protected from this by a filter network. The 10 uH choke can help stop the RFI from getting back to the amplifiers outputs and entering the feedback loop.
 These chokes could be put directly in-line at the speaker terminals on the positive and negative speaker posts or built into the speaker wires in the same position.
 Generally the SS amplifier has a low pass network across the speaker terminals consisting of a capacitor in series with a resistor. This network is in parallel with the speaker load and is not directly in the signal path but is in a shunt position. If high quality parts are used in this application there is no sonic penalty from the filter.
 I would rather put a filter network on the output of the amplifier or increase the size of the capacitor in the existing network to move the filter pole downward rather than listen to the series inductor at the loudspeaker.
 A secondary zobel network could be placed directly across the speaker terminals and it could accomplish the same reduction in RFI as the series inductors.
 Scotty

Hey Scotty and Davey too.

I love the improvements the chokes brought on my mmgs with the pass x150.  Weird thing is that I already had some Zobels in place and they did far less than the choke did.  The chokes were a big wow and the RCs not so much.  Why would that be the case?

[Davey]

Component values of the R and C in your setup?

Dave.

[dawnrazor]

Component values of the R and C in your setup?

Dave.

Hey Dave,

I got them from Cory M.  and just put them in the system.  Here is a post where he details what they are made with .  I am pretty sure these are the ones he sent:

http://db.audioasylum.com/mhtml/m.html?forum=mug&n=135771&highlight=networks+cory+m.&r=&search_url=%2Fcgi%2Fsearch.mpl%3Fsearchtext%3Dnetworks%26b%3DAND%26topic%3D%26topics_only%3DN%26author%3Dcory%2Bm.%26date1%3D%26date2%3D%26slowmessage%3D%26sort%3Dscore%26sortOrder%3DDESC%26forum%3Dmug

[Davey]

I suspect the difference is because of the low capacitor value.

A 0.01uf capacitor and 10 ohm resistor provide a cutoff much higher than the recommended 10uH series inductor.  0.01uF might be too high to effectively shunt some of the RFI (or other garbage) you might be picking up.

So many variables here....both environmental and in the equipment, to really yield a solid conclusion. 

I would always recommend a much higher value capacitor for an RC shunting network.....at least a 0.1uF.  That would get the cutoff down into the sub 100khz range and more effectively "terminate" the speaker wires and shunt the transducers.

Cheers,

Dave.

[kevin360]

Ah, Davey - we do live on a noisy planet in noisy space. Much of that noise is in the VLF range. A VLF/LF radio receiver can be entertaining, or, at least, revealing of the possible sources of negative influence.