[Maritan]

• I know this is a hot button issue.
• I'm not jumping head first into this.
• I know there's science involved in this and I'm willing to understand it.
• While there's science involved, what pleases our ears is what matters most.
• "Upgrading" crossovers WILL CHANGE sound, not necessarily for the better.

Do I have most (if not all) of the disclaimers out of the way? 

I'm calling on all Maggie owners that have upgraded their crossovers to chime in. I'm looking to figure out how parts where matched before and after. I personally have the 1.7s and I've had some idle curiosity regarding this.

With the Maggie's presenting a 4 ohm load, how much does each individual section present? For example, in the 1.7, how much load does the tweeter and super tweeter present? How about the mid bass? Are there numbers available or calculations that can be done

When capacitors are changed to "better" ones, ESR will probably be lowered or at least changed. With the speaker element resistances being quite low, a change in ESR will affect crossover point a little. Will it be audible? I don't know what the ESR of the caps in the 1.7 are, so I don't know the parameters of my simulation.

Inductors are also an issue. Maggie 1.7s uses an iron core, which allows for small size and low DCR. Great. If I wanted to go to an air core, size becomes an issue (unless building an outboard crossover) and DCR change one way or another will again affect crossover and mid bass level. A quick simulation showed that going to from a "stock" DCR of 0.2 ohms to 0.4 ohms will lower the mid bass level by 1dB. My understanding is 1dB is imperceptible to most (if not all) humans.

How do folks that upgrade their crossovers account for these variables?

[Davey]

Nobody's posted a 1.7 schematic yet, so we can't determine the frequency allocation of each individual section.  However, if some adventurous sole takes a look inside their speakers I'd be glad to draw up a schematic and analyze it fully.

For me, a 1db relative change between tweeter and woofer on my MMG's is easily audible.  I think the DCR of inductors should not be discounted.

The rest of your query regards subjective evaluation of "upgraded" components.  I don't think anyone can answer that meaningfully but you.......if you try said components.

Cheers,

Dave.

[Maritan]

Actually, like him or not, and agree with him or not, Peter Gunn has posted a circuit. I'm not a fan of how it's laid out, but here it is:



When I simulated the circuit, I divided the 4 ohm load in thirds - 1.33 Ohms each for Super tweeter, Tweeter, Mid bass. Probably an incorrect assumption. I'm guessing ideally I should dig into the speaker, disconnect the crossover and measure the DC resistance with a handheld multimeter.

[Davey]

Yep, I've seen that schematic.  I have little confidence it's correct. 
Even if it was, you'd still need exact impedance of the drivers to fully simulate.

Cheers,

Dave.

[kevin360]

When I simulated the circuit, I divided the 4 ohm load in thirds - 1.33 Ohms each for Super tweeter, Tweeter, Mid bass.

Yes, that's definitely an incorrect assumption. In terms of the resistance presented by each voice coil on the Mylar, it will depend upon length. The super-tweeter is a single loop and I'll WAG that the tweeter is three loops, which would make the resistance three times as great for the tweeter (assuming they're both comprised of the same foil). Of course, being much longer than either, the bass/mid driver will thus present the bulk of the resistance.

I've done a number of crossover changes with my MMGs, but in every case was it apples to oranges. Therefore, I can make no meaningful qualitative comments on the component changes. I can only state that, of the three I tried (2nd gen stock (as delivered), 1st gen with expensive parts and PG's highly touted series XO (components as specified by PG)), my preference is Magnepan's original design with better parts. I also played around with line level passive (with Dave's kind assistance), which is definitely the way to go, if feasible (not currently an option for me). I haven't tried line level active with the MMGs.

[Maritan]

Davey and Kevin360 - You guys are right. I would need exact impedances to simulate the circuit. And as Kevin360 pointed out, if I ever went down this rabbit hole, I would stick with the stock topology, but just change the stock caps for "better" ones.

That said, with matching DCR of the one inductor and then figuring out ESRs for the capacitors is going to take some very dedicated effort and time...

[satie]

The schematic has to be wrong since it leaves no filters to do midbass/tweeter - it's missing at least one inductor.

[pongagt]

The schematic works for me. It is the same as the crossover in my 1.7s. I didn't have my speakers too long before i updated the crossovers but i know they sound awesome now and have plenty of bass.

[Davey]

Assuming that schematic is correct, I wonder if you fellas have noted the puzzling configuration and that this is not a series crossover? 
The woofer is connected in a normal, parallel configuration.  Only the tweeter/supertweeter are connected in a quasi-series configuration.

(edit)  I've redrawn the schematic with normal convention so it's easier to understand.  Actually, it is a series connection for the woofer, but the high-pass configuration for the supertweeter leaves me puzzled.
I still wouldn't be surprised if that schematic is not correct.

Oh well.

Dave.

[Roger Gustavsson]

The bass seems to have a second order low pass filter. There has been a lot of talk about first order filters in the .7 models....

[Davey]

The bass seems to have a second order low pass filter. There has been a lot of talk about first order filters in the .7 models....

No, it's actually a first-order network.  Only the addition of the 10uF capacitor alters the configuration from the classic two-way series network.
The interesting thing about it is the supertweeter has the same high-pass cutoff as the tweeter.  It might function as a supertweeter but it's not driven that way. 
The 10uF capacitor rolls off the highest frequencies applied to the tweeter (slightly) relative to the supertweeter.

It's not the way I would have implemented this crossover (because of the huge capacitor required.) 

My standard caveat......assuming the schematic is correct. 

Dave.

[Roger Gustavsson]

I do not get it. There is a coil in series and capacitor in parallel to the bass driver. Is that not a second order filter?

[krikor]

I do not get it. There is a coil in series and capacitor in parallel to the bass driver. Is that not a second order filter?

Took me awhile to get my head wrapped around how series crossovers work. Magnepan used these many years ago and now they've returned. I've been thinking of trying one out on my 10.1 pair. This helped clarify it for me:

http://en.wikipedia.org/wiki/Audio_crossover#Series

[Davey]

I do not get it. There is a coil in series and capacitor in parallel to the bass driver. Is that not a second order filter?

Yeah, but the coil is not really in series with the woofer.  The tweeters shunt the coil, and likewise the woofer shunts the capacitor.
This is the standard configuration of a first-order series crossover.

Cheers,

Dave.