Magnepan MG3.6R - Modifications
It strikes me as odd that a speaker with potential of the Magnepan 3.6R hasn’t been offered by the manufacturer in a ‘hot-rod’ or SE version, at least not here in the USA. Why? Well anyone who has looked closely at the internals of the 3.6R will know exactly what I mean. Let me sum it up succinctly: the internal parts/wiring and the standard of workmanship are below par and not conducive to extracting the best performance from the planar design. Coming from a manufacturing background I understand the concept of building for a specific market and controlling costs to deliver a competitive product and at the same time remain profitable. But why oh why must the 3.6R look like a cheap transistor radio internally?
My estimate would be that spending an additional $750 or so at the manufacturing stage could make the 3.6R competitive in the $15,000 market sector, maybe higher. As it stands, the 3.6R with its list price of under $5000 presents a real dichotomy.
The elements of its performance that it does well, it does better than most any other speaker out there. Its natural tone through the midrange and seamless, cohesive transition through the entire frequency range is hard to find in any comparably priced box speaker, at least any that I’ve heard. But its weaknesses are almost debilitating to many potential owners. It has the reputation of not doing bass particularly well and being poor in the imaging department. It doesn’t have the dynamic impact in the lower registers that a good box design has, and in the new era of thunderous HT sound effects, well, 60% of the potential Magnepan market just evaporated.
While many people refer to the Magnepan’s limitations as being of a physical nature, I’ve never been really convinced of that. Having heard many times the comments about the need to move air to deliver lower frequency dynamic impact, well that’s quite a large bass and midrange panel that the 3.6R has.
Anyway, if you hadn’t noticed, a new craze has swept though the audio world: ‘modifications’. Take a component of reasonable quality, shake it around, add some better parts and voila, watch it compete with the best.
And so driven by a sense of frustration and the urge to prove to myself that I could still use a soldering iron, I embarked upon the task of destroying a perfectly good set of speakers. Well, not quite.
First, let me cover some basics for people who are not familiar with the stock Magnepan 3.6R.
Design / Construction:
The 3.6 is a 3-way speaker, with a planer/magnetic midrange/bass and a true ribbon tweeter.
It’s specifications on paper are 85db/500hz/2.83v sensitivity and a 4 ohm impedance, but believe me when I talk from experience and say that the 3.6R needs a strong amplifier to make it sing. Select an amp with the ability to deliver high current and clean power without clipping and you’ll be rewarded with a deep/wide soundstage and vastly improved dynamics, even in stock form.
Measuring 24" x 71"x 1.625" the 3.6R is a large speaker, but not particularly heavy, so unpacking, assembly and preliminary positioning is not a difficult task for one person.
Magnepan, for reasons known only to them, are not proponents of the ‘rigid construction is best’ theory. The stock stands are flimsy at best, obviously victim of the ‘design to a cost’ policy.
Disappointing also is the external crossover box, and its method of coupling to the rear of the speaker. As mentioned previously, the 3.6R is a 3-way design, the crossover components for the tweeter/mids are mounted inside of the speaker, but the bass crossover components are mounted within two small boxes which attach to the rear of the speakers using metal pins. These pins also act as electrical conductors, making the connection between the box and the speaker terminals. Small brackets are supplied which attach the crossover box to the speaker, removing support stresses from the pins. All of this is barely adequate in my opinion.
How can a speaker gain so much critical acclaim, yet completely fly in the face of everything we have come to know about the importance of construction integrity and rigidity?
Sound of the stock 3.6R:
I’ve tried a dozen different amps with these speakers, including powerful tube Monoblocks, and bi-amped SS configurations, both analog and digital. I won’t describe the sonic attributes of each here, suffice to say that my SimAudio W10 Monoblocks are the best partner that I’ve found for the 3.6R so far.
The first thing that strikes you about the Magnepan sound is how open and natural it is. You immediately realize that a good percentage of what you hear in a typical box speaker, is the box.
The next thing that strikes you is how differently these speakers express themselves when compared to more conventional box designs. The sound has tremendous scale and clarity. Sound flows from these speakers in a big open wave, it doesn’t jump at you with the same dynamic snap and attack that typical box speakers provide.
The soundstage from the 3.6R’s can be huge, with excellent center fill energy and a good cohesive stage presentation. The stage width easily extends beyond the outer edges of the speaker when required to, and given sufficient driving force. Stage depth was quite good, though not as deep as other speakers, particularly the Quad ESL 57’s or the Apogee Caliper Sig. II’s that I’ve owned.
As mentioned earlier, bass is somewhat lacking in the 3.6R, both low bass extension and mid/upper bass fullness, which gives the sound a slightly lean presentation.
I also initially found the tweeters to be a little too bright for my tastes, with just too much treble energy. Brass instruments such as solo trumpets and even solo woodwind instruments came across with just too much energy and force. Installing a resistor in series with the ribbons, as facilitated easily on the speaker back-plate, robs the speaker of some air and transparency and is not recommended as a solution for reducing treble energy.
I was able to tame this slightly forward presentation somewhat by moving the speakers closer to the front wall, but with an unwanted loss of soundstage depth as a trade-off.
I was able to achieve something of a compromise between tonal presentation and stage depth, but it always left me wanting more of what the compromise had removed. In a nutshell, I wasn’t really satisfied with the sound of the stock 3.6R to the point where I could live with and enjoy the speakers in the long term.
Modifications and a new Perspective.
. . . the moon gazed on my midnight labors, while, with unrelaxed and breathless eagerness, I pursued nature to her hiding-places….Mary Shelley, Frankenstein.
There are several stages of ‘improvement’ available to the stock 3.6R and each will provide the inspiration to move to the next. For those of you suitably skilled and motivated to take on the ultimate upgrade, skip points 1 and 2. For those of a more nervous disposition, or perhaps with more limited funds, try points 1 and 2, in conjunction with 3 and 4 and you will have yourself a nice improvement for under $1000
The following is a summary of improvements and their impact on the sound:
Ditch the pins that connect the external XO box to the speaker and replace them with custom Cardas ‘Magnepan’ wires, available from www.dedicatedaudio.com . This is a worthwhile modification and adds a little clarity and transparency to the sound. The cost of the cable kit is around $200 and all that is required is an Allen key (and $200!).
Remove the fuses and the tweeter attenuation circuitry on the inside of the back-plate. Re-crimp the wiring inside the cavity behind the back-plate with better quality gold-plated crimps. This is very worthwhile indeed. The sound improves in many ways, almost like a removing the cloth grills from a typical box speaker. More detail, more air and even slightly improved bass extension. Only bypass the fuses if you are certain that your amp can deliver sufficient power to the speakers at higher SPL’s without clipping. Otherwise you may end up with an expensive panel/ribbon replacement.
Replace the stock stands with custom stands that add mass and rigidity to the frame. MYE stands, available at www.myesound.com seem to be the preferred stands amongst Magnepan users and offer very worthwhile improvements for under $400. I’m cheap, so I decided to build my own stands based around the MYE design and a set of SoundOrg custom 3.6R stands, and of course I spent much more than $400.
The principal of the MYE design and ‘my’ design is to mass-load the stand for rigidity and support the speaker further up the frame to prevent flexing. Then, couple the stand to the floor with vibration dissipating spikes or brass cones, and there you have it.
Improvements were again noted in transparency and also a little increase in stage depth and layering.
Try Magwire! – yes, I owe this to Paul Speltz, but magwire really works for me. I’ve tried many different speaker wires including Purist Audio, Sonoran, AudioQuest, AlphaCore Goertz and Virtual Dynamics Nite II’s. Prior to discovering magwire, the VD Nite II’s had proven far and away the best choice in my setup, but the magwire provides the detail and clarity that even the $2200 VD’s couldn’t touch.
Crossover modifications: by far the most complex and irreversible of the modifications made. I’ll cover this in bullet form, but would be happy to provide more details to anyone wishing to embark on similar madness.
Organize your parts carefully and plan on replacing all internal XO components and external XO components. This way you can have the new XO fully wired before your speakers undergo the knife.
Choose a XO board of suitable proportions. I used two 24" X 12" X 2" solid Mahogany boards, but would have appreciated another 2 or 3" on both length and width to allow better spacing of the components. This sounds like a large board, but your new XO components and going to be much larger than stock, so you will need it.
Plan on removing the back-plate from the circuit completely. Rewire the speakers internally and take leads from the 3 drivers through the back-plate cavity and terminate the leads with good quality spades. I used Nordost flatline cable throughout and AlphaCore solid silver spades. I left 18" on the leads from the point of leaving the back-plate. I didn’t use spades inside the speaker, I drilled small holes in the ends of the Nordost wire and used the actual speaker wire as the crimp.
The leads will now terminate on one side of your new XO and I used good quality Cardas binding posts available from www.dedicatedaudio.com You will need a total of 10 sets of binding posts, around $400 if you use the Cardas design.
I built the new XO using the same value components as stock but selected better quality items. I used Solen fast-caps for larger values, Kimber for the mid values and Jensen oil/paper caps for the smaller values on the tweeters. I used Solen iron-core inductors for the bass, and AlphaCore air-core inductors on the mids/highs. (www.madisound.com)
I used Cardas silver solder and flux, and believe me, it makes life much easier than the Ratshack stuff that doesn’t stick to anything other than the carpet. (www.dedicatedaudio.com)
I used DH Labs wiring through the XO www.madisound.com
Building the XO’s is relatively easy, though some of the soldering maneuvers are a little tricky, particularly where 8 caps need to be wired in parallel.
I experimented with isolation under the actual components with great results, so it’s a good idea to leave a little slack on the point-to-point wiring to facilitate this. I tried Vibrapods under components and also foam. I found the best results were attained with packaging air-bags, but they are a real pain to work with.
I used brass cones on the underside of the XO bases and spiked them into the carpet (www.supercellaudio.com)
The end result is worth the effort. After around 2-300hrs of break-in, the sound is better in every possible respect. The deficiencies associated with the stock 3.6R speaker are almost fully addressed. Dynamic impact in the lower registers is greatly improved. Bass is more extended, tighter and more tuneful. The highs are effortless and extended, without sounding as shrill.
Transparency is greatly improved, as is stage presentation, particularly depth. There is a much-improved sense of layering in the soundstage with air around instruments/performers that just wasn’t there to the same extent previously.
The total cost for the XO upgrade and speaker re-wire was in excess of $2500, so it falls outside the scope of being a ‘tweak’. But it performs beyond its price, at least when taking a typical audiophile perspective on things.
There is something very special about the Magnepan 3.6R, but with modification it can be elevated into a different league altogether.
The 3.6R is clearly capable of stellar performance, given the removal of certain constraints and partnering with appropriate equipment, in particular, a good quality and high powered amp.
If you have the time, patience and inclination to pursue the goal of achieving a big improvement to the sound of the stock speakers, then I highly recommend trying one or more of the listed upgrades to your Maggie’s.
For a sound that is open, full scale, articulate and free from the sonic coloration’s of boxes, you can’t go wrong with the Magnepan 3.6R.
Attention: Please ensure that you have the necessary skills to perform any modification to your speakers. Some of the modifications are difficult to reverse, so you will need to plan the process carefully. If you have doubts over your ability to complete any modification satisfactorily, then it is better left to one of the many companies out there that can perform the modifications professionally on your behalf.
Rooze - April 2005