Magneplanar -- I'm doing the SMG crossover upgrades

[squeegy200]

I've been reading the many posts chronicling Peter Gunn's adventures with his SMGs and their final resting place.

I would love to have him do the the Magnestand upgrade on my factory rebuilt SMGs. His website described a future DIY page with details.

However, there is one upgrade that I can perform and that is the crossover. He's experimented with different values and components.

I've ordered the essential parts including the Janzten Copper Foil Inductors and the Obbligatto Capacitors.

Funny thing, the capacitors are sourced from Hong Kong and the US postal service found it necessary to open my mail to inspect the contents. The Obbligattos are much larger than I anticipated. There are two wires at the top. I would imagine it looked much like an explosive device.

My mail was carefully repackaged into a nice sturdy express mail box and sent on its way.

I'm looking forward to the improvement. Judging from the numerous discussions on various blogs/websites, I have large expectations of what it will achieve.

[Feanor]

...

However, there is one upgrade that I can perform and that is the crossover. He's experimented with different values and components.

I've ordered the essential parts including the Janzten Copper Foil Inductors and the Obbligatto Capacitors.
...

I'm looking forward to the improvement. Judging from the numerous discussions on various blogs/websites, I have large expectations of what it will achieve.

Squeegy, let us know how it works out.

Of course, the supposedly mediocre stock crossovers on the Magneplanars have been a criticism for years.

[squeegy200]

Id heard that too often myself. When I finally opened up the socks and took a look for myself, I was astonished to find parts that were equivalent to crossovers in a pair of $50 book shelves speakers.

All the more impressive if one considers that Magneplanars sound as good as they do with such inferior components. Its a testament to the quality and design of the drivers.

If there is as much potential in these things as are stated, I'm going to attempt to release it.

[E-Stat]

If there is as much potential in these things as are stated, I'm going to attempt to release it.

I've heard quite a bit of good buzz about the Gunned Maggies. On the other hand, Magneplanar is getting a nice reaction to their new mini flavor at CES. There is a pic on Stereophile here. You have to scroll a bit to find them, but they really are small.

rw

[bubslewis]

[QUOTE=squeegy200]Id heard that too often myself. When I finally opened up the socks and took a look for myself, I was astonished to find parts that were equivalent to crossovers in a pair of $50 book shelves speakers.

All the more impressive if one considers that Magneplanars sound as good as they do with such inferior components. Its a testament to the quality and design of the drivers.

If there is as much potential in these things as are stated, I'm going to attempt to release it.[/QUOTE]


You'll probably have to mount the new capacitors and inductors externally since they'll be larger than the originals. I upgraded my 1.6's with Hovland capacitors and Alphacore foil inductors. I had two small black (they matched the black socks) speaker shelves that I used for the new stuff. I placed the inductors and capacitors on some bubble wrap and placed them on the shelves directly behind the speaker.

The shelves extend an inch or two past the ends of the speaker, but from a front or side view they simply look like part of the speaker mounts. I do spray down the exposed wire, inductor foil, etc. once in a while with Deoxit spray and once a year I apply a coating of Deoxall to the exposed parts. Otherwise they just sit there in the open on the speaker shelves.

Are you maintaining the same mfd capacitance and same milli henries inductance as the originals or are you experimenting some?

Good luck with your project,
Bill

[Feanor]

You'll probably have to mount the new capacitors and inductors externally since they'll be larger than the originals. I upgraded my 1.6's with Hovland capacitors and Alphacore foil inductors. I had two small black (they matched the black socks) speaker shelves that I used for the new stuff. I placed the inductors and capacitors on some bubble wrap and placed them on the shelves directly behind the speaker.
...
Bill

bubs, I'd like to upgrade my 1.6's too -- given I won't be trading up to 3.6's any time soon.

I'm a bit intimiated about (a) getting behind the socks, and (b) rewiring to the panel leads. Is there anything you can say to alay these trepidations of mine?

Also, any advice vis-a-vis changes other than quality of components? I see the standard X-over is a hybrid: 1st order high-pass / 2nd order low-pass. So (for example) ... say we lowered the high-pass a 100-200 Hz and made it 2nd order???

... standard x-over ...

[squeegy200]

You'll probably have to mount the new capacitors and inductors externally since they'll be larger than the originals. I upgraded my 1.6's with Hovland capacitors and Alphacore foil inductors. I had two small black (they matched the black socks) speaker shelves that I used for the new stuff. I placed the inductors and capacitors on some bubble wrap and placed them on the shelves directly behind the speaker.

The shelves extend an inch or two past the ends of the speaker, but from a front or side view they simply look like part of the speaker mounts. I do spray down the exposed wire, inductor foil, etc. once in a while with Deoxit spray and once a year I apply a coating of Deoxall to the exposed parts. Otherwise they just sit there in the open on the speaker shelves.

Are you maintaining the same mfd capacitance and same milli henries inductance as the originals or are you experimenting some?

Good luck with your project,
Bill

You are right.
I was surprised at the size of the capacitors when they finally arrived. The US Postal Service found it necessary to open my package from the supplier to inspect its contents. They didnt even bother resealing the original package. They just threw them into another box, taped it up, and sent it to me in their US Postal box. The Obbligato Capacitors are smaller than an aluminum soda can--but not by much. My intended project box turned out to be too small

So I ran out to the local electronics retailer and purchased larger project boxes to accommodate the larger capacitors.

Following Peter Gunn's example, he extended the wires from the mounting points at the base of the speakers. These extended wires will reach out to the project box which I will position behind the speaker. The speaker cable themselves will connect to the new project box which is presently on the floor. Eventually, i would like to have a platform like the Magnestands on which to place the box.

I used Janzten and Obbilgato since that is what Peter Gunn used on his SMGa. But I am experimenting with the values to retrace the path he took. I believe he finally rested at a capacitor value roughly 1/2 of the OEM value. I am starting with 22mf

[bubslewis]

bubs, I'd like to upgrade my 1.6's too -- given I won't be trading up to 3.6's any time soon.

I'm a bit intimiated about (a) getting behind the socks, and (b) rewiring to the panel leads. Is there anything you can say to alay these trepidations of mine?

Also, any advice vis-a-vis changes other than quality of components? I see the standard X-over is a hybrid: 1st order high-pass / 2nd order low-pass. So (for example) ... say we lowered the high-pass a 100-200 Hz and made it 2nd order???

... standard x-over ...

Feanor,
I'm sorry I can't answer your questions about lowering the high pass and making it 2nd order. I don't have that kind of technical knowledge (I'm pretty sure yours exceeds mine in this area).

Basically I simply "cut and pasted" the new capacitors and inductors into place and copied the original wiring schematic. I used a 7mfd + 7mfd + 8mfd capacitor array which replaced the 10 + 6.8 + 5.1 original array. Oddly, there was no 0.1mfd bypass cap that I could see. This bypass cap is on the schematic (I have the same one you do), but darned if I saw it anywhere. On the inductor side, I just snipped the wires from the original inductor and left it sitting in place. There's no way the new inductor was going to fit in the same space. I did not change 25.0 mfd bypass on the inductor side.

I gritted my teeth when I cut the sock, but it turned out to be very easy. I cut along the bottom edge of the crossover compartment and then up the sides to the top edge. I rolled up this "flap" (kinda like a rolled up awning) and pinned it along the top edge of the compartment.

My biggest fear was the soldering part. As feared, I did a messy job but everything seems secure, if not aesthetically first class. I will see if I can get a couple of pictures.

tx,
Bill

ps. Yes, I definitely think the effort was worth it.

[bubslewis]

[I used Janzten and Obbilgato since that is what Peter Gunn used on his SMGa. But I am experimenting with the values to retrace the path he took. I believe he finally rested at a capacitor value roughly 1/2 of the OEM value. I am starting with 22mf[/QUOTE]

Squeegy,

I am not even close to being an expert here, but a 50% reduction in capacitance seems like quite a bit. One would have thought that the Magnepan engineers would have been a little closer to the mark if indeed 11mfd is the "ideal" capacitance for the SMG's. I'll try to get a few pictures in my next post.

Bill

[squeegy200]

Squeegy,

I am not even close to being an expert here, but a 50% reduction in capacitance seems like quite a bit. One would have thought that the Magnepan engineers would have been a little closer to the mark if indeed 11mfd is the "ideal" capacitance for the SMG's. I'll try to get a few pictures in my next post.

Bill

Bill,
You are very observant!
Peter Gunn discusses this at length in his postings and one his webpage. On my SMGs the OEM spec is 50mf. He actually started experimenting with a value of 10mf and worked up in increments. He posts the theoretical response curves for each in the discussion. He settled at 20mf for subjective reasons on the SMGa.

I am going to start at his ending point and work my way to the OEM value.

He also discusses at length the soldering points and how he dealt with those. There is detail on how he removed the socks and rewired. His website has several pictures and how he actually applied some of the same modifications to the other models in the Magnapan line.

I did not pay attention to the discussions of the other models. However, I remember him stating that he adamantly preferred the simpler SMG crossover applied to the larger models vs the OEM configuration.

[bubslewis]

A few pics of my crossover upgrade effort. Note the size difference between caps and inductors, new vs. the oem's. I just left the original inductors (top right of compartment, the little red square things). Don't examine the soldering work too closely.

A cover would be nice but since we have no inquisitive pets and the kids/grandkids know it's death to go near the speakers, I haven't gotten around to it yet.







ps The Maggie 1.6's and lower have rather easily upgrade-able crossovers. On the other hand the 3.6 has 16 caps and 5 inductors. God knows what the 20.1's have.

tx,
Bill

[squeegy200]

Bill,
I'm interested in your impressions of your upgrades before and after.
Did you experiment with your values or did you replicate the original values?

That looks exactly like what I am doing. For my first values, i don't need to run parallel capacitors. However, if I do go to the oem value, Ill add another parallel cap to my existing.
My airfoil value is slightly higher, however, it was mentioned in the discussion that there is a 5% +/- value and that puts me within the OEM inductor value for my SMGa

You wire terminations verify what was suggested to me. There is no need to remove the older crossover because the audiophile grade components won't fit in the space provided anyways. I'm following your lead on placement and will extend some wires out to my crossover which Ill place in a generic electronics project box and position directly behind the speaker on the floor.

Question for you: What does the Deoxit do? Do you use it on the contacts too?

[squeegy200]

BTW: Ill be removing the fuse and the corresponding load resistor from the circuit altogether. That has been suggested by many folks on the boards. Clean amplifier power is mandatory!

[squeegy200]

I've so far wired the internals. There should be an attached JPG showing the components inside the project box. I placed regular padded drawer liner at the bottom.

Next up is to remove the post plate and extend the wires to reach this box.

[bubslewis]

I've heard quite a bit of good buzz about the Gunned Maggies. On the other hand, Magneplanar is getting a nice reaction to their new mini flavor at CES. There is a pic on Stereophile here. You have to scroll a bit to find them, but they really are small.

rw

Nice article/picture. That little panel was supplemented with (of all things) a Magnepan subwoofer module. I wonder how compatible this new subwoofer module might be with other Maggie speakers up to and including, maybe, the 1.6.

[Feanor]

[/i]
Feanor,
I'm sorry I can't answer your questions about lowering the high pass and making it 2nd order. I don't have that kind of technical knowledge (I'm pretty sure yours exceeds mine in this area).

Basically I simply "cut and pasted" the new capacitors and inductors into place and copied the original wiring schematic. I used a 7mfd + 7mfd + 8mfd capacitor array which replaced the 10 + 6.8 + 5.1 original array. Oddly, there was no 0.1mfd bypass cap that I could see. This bypass cap is on the schematic (I have the same one you do), but darned if I saw it anywhere. On the inductor side, I just snipped the wires from the original inductor and left it sitting in place. There's no way the new inductor was going to fit in the same space. I did not change 25.0 mfd bypass on the inductor side.

I gritted my teeth when I cut the sock, but it turned out to be very easy. I cut along the bottom edge of the crossover compartment and then up the sides to the top edge. I rolled up this "flap" (kinda like a rolled up awning) and pinned it along the top edge of the compartment.

My biggest fear was the soldering part. As feared, I did a messy job but everything seems secure, if not aesthetically first class. I will see if I can get a couple of pictures.

tx,
Bill

ps. Yes, I definitely think the effort was worth it.

Bill, your comments are very helpful indeed, and do encourage me to try this.

In particular I'm interested to hear the you cut the sock rather than trying to remove it as some people suggest; I think I'd user your approach. Your pictures are great, BTY, and answer the question whether you used the original cable connectors and fuse, (yes), unlike squeegy who is providing new connectors on account of his completely external crossover box.

From what I know, the most critical upgrades would the high-pass capacitors, but clearly a better low-pass inductor couldn't hurt. The 0.1µF capacitor would have almost no effect on the crossover point; I think the theory is that these low value capacitors in parallel with higher value ones provide better transient response. However I'm not sure whether this effect is proven or significant.

I doubt I'd try it, but as a matter of interest, a 2nd order high-pass would involve some increase in the capacitance (to lower the crossover point a bit) and inductor in parallel with the tweeter itself (to increase the crossover slope from 6 to 12 dB per octave). Note a 2nd order crossover induces 180 degress of phase shift so it would be necessary reverse the polarity of the tweeter -- from the diagram, it looks like the standard crossover has the tweeter opposite phase so you'd be bringing it back into phase.

Right now I user bi-wiring so I bypass the jumpers; also, I user a part of Hi-Fy Tuning fuse in place of the stock ones. These fuse are $60 a pair and make a slight, positive difference -- you'd no doubt get more improvement by simply bypassing the fuses.

[bubslewis]

Bill,
I'm interested in your impressions of your upgrades before and after.
Did you experiment with your values or did you replicate the original values?

That looks exactly like what I am doing. For my first values, i don't need to run parallel capacitors. However, if I do go to the oem value, Ill add another parallel cap to my existing.
My airfoil value is slightly higher, however, it was mentioned in the discussion that there is a 5% +/- value and that puts me within the OEM inductor value for my SMGa

You wire terminations verify what was suggested to me. There is no need to remove the older crossover because the audiophile grade components won't fit in the space provided anyways. I'm following your lead on placement and will extend some wires out to my crossover which Ill place in a generic electronics project box and position directly behind the speaker on the floor.

Onward and upward.
Bill

Question for you: What does the Deoxit do? Do you use it on the contacts too?

Deoxit is just a very good contact cleaner ( not the cheapest). Works great on copper and all electrical contacts, etc. Deoxall (also not cheap) comes in a little bottle with a small brush applicator. You brush it on and let it dry. It forms a thin layer and protects the wire, foil, etc. from moisture absorbtion and subsequent oxidation.

For me, the cost of the upgrade was worth it. Do NOT expect earthshaking, night-vs-day sonic improvements with this upgrade, since the Maggies sounded pretty good right out of the box. Before the upgrade, on certain songs on certain discs, I felt myself wishing that the vocals were a bit more up front, not as recessed as much. After the upgrade I didn't get that feeling when I listened to those certain discs. I didn't realize that immediately, but it just came to me that I no longer was wishing for more up frontness in the vocals.

To me, the lateral dispersion and holographic depth of the soundstage got even better. These are two of the strong points about Maggies to begin with and, for me at least, some of the main reasons I like them so much. I also feel that the bass response got a little better but I have no measurements or quantitative data to support this, so I'll just have to trust my very subjective ears on that one.

Also, I've also read that bypassing the fuse helps a bit (thinking of doing that myself). This should be safe enough unless you really blast your music at real high levels.

[bubslewis]

In particular I'm interested to hear the you cut the sock rather than trying to remove it as some people suggest; I think I'd user your approach.


I doubt I'd try it, but as a matter of interest, a 2nd order high-pass would involve some increase in the capacitance (to lower the crossover point a bit) and inductor in parallel with the tweeter itself (to increase the crossover slope from 6 to 12 dB per octave). Note a 2nd order crossover induces 180 degress of phase shift so it would be necessary reverse the polarity of the tweeter -- from the diagram, it looks like the standard crossover has the tweeter opposite phase so you'd be bringing it back into phase.

Right now I user bi-wiring so I bypass the jumpers; also, I user a part of Hi-Fy Tuning fuse in place of the stock ones. These fuse are $60 a pair and make a slight, positive difference -- you'd no doubt get more improvement by simply bypassing the fuses.

Feanor, If I had to take the entire sock off, I may well not have attempted the upgrade. I'm real glad I just cut what I did.

I believe most bipolar speakers reverse polarity in the crossover between high and low frequency as this has a major effect on the depth and rebound/reverb areas of the sound.

I may tinker with bypassing the fuse at some point in the future.

tx,
Bill

[squeegy200]

To my astonishment, upon closer inspection, my crossover wiring is unlike the SMGa of Peter Gunn. Its not even close to the SMG diagrams on the MUG user group site. I was perplexed at what to do last night.

Then I looked at Peter Gunns site once more and discovered my SMGs are wired similarly to the newer MMGs. As Bill noted, some of the tweeters on the newer panels have been reversed. The MMG solution worked.

[squeegy200]

I've finally completed the upgrade and after 12 hour break-in I decided to sit and give it a listen.
To clarify what happened earlier, I discovered that my SMGs did not match any of the MUG crossover diagrams that correspond to my model. Further, I found that the values of the capacitor inside my OEM crossover was 17mf not the 50mf of the posted articles. I could not determine the value of the inductor.

I was at a stand still not knowing what to do. I read further and found that Peter Gunn had used the SMGa configuration that he pioneered on his speakers with other 1.6 and MMGs. I dug a little deeper and discovered that the MMGs looked similar using the same crossover. So I decided to stay with the configuration I had and take the MMG approach.
The first diagram I took from one of the sites and mimics its wiring to the Xover. It was identical in almost every way. (See illustration below)

I also opted to remove the sock entirely off my first speaker. According to the previous owner, these were factory rebuilt prior to selling them to me. I can tell they were refurbished and for the most part, the socks looked fresh and the wood trim had a design I had not seen anywhere with two parallel trim pieces in the center. They all looked fairly new.
I was happy to have taken the socks off to discover underneath the fresh socks hid some impact damage to the frame. The main driver on the left side, although unaffected, was not attached at the top. This gave me an opportunity to correct this. I then proceeded to assemble everything back up

[squeegy200]

After the break-in I sat down and didn't like what I heard initially.

Gain was down noticeably on the right speaker. At first I thought that since the sock still remained it was the difference. But after adding some volume I noticed that the lower registers were weak and indeed the gain was far lower on the right than on the left.

1 I reversed the + and - from the right speaker to the crossover and suddenly there was this amazing filling of the bottom end. It seems the mid/bass panel was out of phase.

But sitting down, now the right was out of phase with the left speaker.

2. I then took the speaker wires from the amplifier and reversed the +/- main leads into the crossover.

Now that the speakers are in phase, the fullness that once was is now reappeared magically.

What I have to say about this upgrade is WOW!

SMGs have a reputation of needing a subwoofer. These do not. The bass response is unlike any Maggie I've heard previously. This modification really allows the true fullness of the mid/ bass panels to shine through.

Another observation I encountered is that the narrow focal point I had with the previous configuration has widened considerably. There is no longer this pinpoint sweet spot that I once had to restrict myself to listening in. The sweet spot has just opened itself up.

I'm going to still continue to experiment with the values. I'm currently using the Janzten .82 air inductor and 22mf Oblligatto Caps. I like the more delicate decay of instruments representing a larger sound stage and space between the instruments. Voices and strings have a smoother presentation. However, I do feel that there was a slight attenuation of the the upper registers such as the ring of a triangle or the shimmer of cymbals. Although there, the focus seems to have been pushed farther backstage.

The beauty of this configuration is that I can create another box with different values and swap them immediately allowing me to make A/B comparisons right away.

To my surprise, My wife really likes the Maggies with the socks off. Who would have imagined that?

[bubslewis]

Seems like you are having quite an adventuristic time!

I thought that, in one of your previous posts, that the 50 mfd capacitance on your HF seemed a bit strange. I believe 50mfd would produce a razor sharp zinginess to all the high frequency sounds (bells, cymbals, etc). Maybe the fact that your using 22mfd value vs the oem 17mfd value explains the slight sharpness at the upper end of your high frequency.

It's hard for me to tell what phase polarity you have ended up with, but your experimenting with reversing the wires was very interesting. I know that the high frequency and low frequency panels on my 1.6's have reversed polarity and I'll bet if I changed the polarity (which I'm not going to do) that I'd hear the same problems you described when you first listened.

Keep us posted on your progress.

thanks,
Bill

[Feanor]

Seems like you are having quite an adventuristic time!

I thought that, in one of your previous posts, that the 50 mfd capacitance on your HF seemed a bit strange. I believe 50mfd would produce a razor sharp zinginess to all the high frequency sounds (bells, cymbals, etc). Maybe the fact that your using 22mfd value vs the oem 17mfd value explains the slight sharpness at the upper end of your high frequency.

It's hard for me to tell what phase polarity you have ended up with, but your experimenting with reversing the wires was very interesting. I know that the high frequency and low frequency panels on my 1.6's have reversed polarity and I'll bet if I changed the polarity (which I'm not going to do) that I'd hear the same problems you described when you first listened.

Keep us posted on your progress.

thanks,
Bill

Squeegy & Bubs,

You guys probably know that the higher the capacitor value, the lower the high-pass crossover point. Speaking in general, a higher high-pass capacitor value without a correspond change to the low-pass values, will causing brighter midrange due to greater overlap of the drivers. Complicating this however is the polarity of the drivers: if the drivers are effectively out of phase with each other, their outputs will cancel each other where they overlap.

I'm not sure why in the MG 1.6 crossover the tweeter and woofer are wired with opposite phase except, probably, that the net effect of the respective 1st order and 2nd order crossovers more nearly approximates a 2nd/2nd order crossover than a 1st/1st order one. That being the case in reality, reversing the stock polarity would dull the presentation around the crossover point because the tweeter and woofer outputs would cancel each other.

But I also don't understand the 21 uF high-pass capacitor on the MG 1.6; theoreitically with a 4 ohm speaker this would yield -3dB around 1800-1900 Hz. Also, the low-pass values seem odd according to theory -- see here. However I'm told the effective crossover point for the 1.6 is 600-700 Hz -- very curious. No doubt the actual value are arrived at empirically by Magnepan -- and tweakers will need to do the same.

[squeegy200]

According to Peter Gunn's write up. My 1st order crossover using the 22uF capacitor would bring the crossover point to roughly 1100hz -4db. The theoretical ideal is supposedly somewhere around 800hz at -3db which theoretically could be achieved by increasing capacitance to 50uF and maybe even 60uF

I'm currently using this wiring diagram with slight changes. The right side speaker (Red & Blue) wires are reversed.

Ill leave it here for now and listen for awhile. These are beginning to sound better and better each time I sit down and give them a listen. I heard a background guitar that I had not heard before on Carole King's Tapestry on SACD. I can also hear ambient noises in the studio which were not apparent previously on Orion Weiss' paino recital on HDCD Yarlung Records. Even if my values and wiring assumptions are incorrect, the SMGs in their current state are far more detailed and revealing than they were previously.

In the meantime, I'm going to order those 50uF caps and Ill update you.

[squeegy200]

Nice article/picture. That little panel was supplemented with (of all things) a Magnepan subwoofer module. I wonder how compatible this new subwoofer module might be with other Maggie speakers up to and including, maybe, the 1.6.

A subwoofer! Is that the smaller boxes adjacent to the speakers? Is that something Magnepan once built? I noticed the wood finish matched the speaker panels but I didn't know what the purpose of those four boxes.

Are they dipole? I cannot imagine a traditionally designed subwoofer to fully integrate with Dipoles as they would be out of phase with either the front or rear portion of the sound energy emanating from a Maggie.