Couple measurements of the Maggie's

jr@mac

These are a 3-way stand mount using MCM Audio Select drivers - 6.5" KB, 3" KB, and a little tweeter. There has been something wrong with them, and I just stopped looking into it until last night when I re-measured.

To put things into perspective, the mid has a ripple in the response common to other drivers this size (think of the little Dayton PS95 - almost exact same response), and it bothered me from a measurement perspective so I notched it out. That drive impedance a little too low for my liking, so some added resistance came into play and that resulted in a bit of a smiley curve. They have a recessed quality, that finally showed up last night during some evaluation of the demo track from last DDIY. The cover of "Imagine" I put in there should sound very forward and even borderline harsh on a well-balanced speaker, and on these it sounded just fine.

Anyways...

Horizontals also show another issue with some bunching around 2-5k. 



On the other hand, the system distortion is very, very good. This was taken at 1M, SPL accurate:



You can see the smiley a little better in the distortion measurement, the horizontal overlays kind of obscure things.

So, since I have no job, tons of crossover components, and these drivers perform well enough to warrant a revisit of the crossover - I am going to rework them. 

I now return you to your regularly scheduled programming. It is icy as fuck outside here, by the way - plus snow coming. I am going to run and get beer so I have fuel for this project. 

Kornbread

mmm... beer... haven't had one on a long time.  Drink one for me. 

rjj45

That distortion measurement is great, no matter what the name on the driver is.

jr@mac

Yes it is, Don. Those KB drivers are incredible value if you are willing to tinker with them.

The ice turned into SoDak wind wrecked my deck table. RIP old friend.

Turn2

Sorry 'bout that table. I'd like to say those items need to be garaged for the winter in the brutal north, but then I looked at my own deck...oops.

Anyway, are you expecting this to be a minor incremental improvement, or is even more procrastination a good thing? Asking for an acquaintance   .

jr@mac

Looking like a new xo at this time. Simpler, so that's a positive.

jr@mac

So I spent considerable time the last week or so modeling and tweaking, and I have an entirely new crossover. Rich, I hope you did not already buy everything - if so, well lot of goodies to put in the spare parts bin :-)

This new crossover is an improvement across all areas. It is simpler, has much improved off-axis behavior, but I did discover the tweeter needs to run a bit "hot" to sound right. Not sure why that is, to be honest. Maybe some malingering electrons here and there.

Anywhoo, updated distortion measurement:



I don't care what anyone says, that is very, very good distortion levels for almost any price project. 

Updated on-axis:



I decided sins of omission etc is the better path here, so the weird hump in the mids response I decided to just let it ride. As a result, I have a minor dip just before the hump and another right after the hump - the dip after is actually somewhat beneficial, as will be seen in the off-axis. As you can see, the tweeter exhibits a rising response, and not sure but the need to run it slightly hot might be related to the off-axis, as well:



The LR "bloom" is now centered on the primary dip in the on-axis. Otherwise, the off-axis is quite good. 

Here is the modeled response, keeping in mind this modeling was done on measurements taken quite some time ago so it will not be an exact overlay, but close enough:



I see a dip! Must be a BW! 

Modeled impedance:



And finally, the revised crossover:



Since I bought a shitload of those TI 30uF capacitors, I just doubled up on them for the two large values. The woofer coil is a laminate core, the others are 18awg air core with the exception of the 1.2 which is 20awg. 

Anyways, serious listening improvement - stupid fake tweeter chamber notwithstanding. 

Wolf

I only have one question- why is the woofer impedance increasing to 12 ohms at 10Hz?

Otherwise- nice work.

Turn2

Nice work. The on hand inventory will withstand much of this, after the large coils. I came real close to wrapping these up over Labor Day, but have ignored them since. Glad it worked out that way now.

jr@mac

I think it was an artifact of a previous sound card, Ben.

Wolf

I guess I never would give that a thought.