Crossover upgrading

[Carlman]

It's amazing to me how quickly Rick figured out so many specific things that took me a long time to discover through listening.  There IS a science to this stuff after all!   Rick has pulled back the curtain and let me see into the workings which has been a fun learning experience.  He explains why something is happening which is helpful when I marry the explanation to my experiences.  And a greater understanding is reached!
Yes, it is very Zen-like...

[richidoo]

Looking forward to hearing the updates.

Hopefully you can hear the finished product on your visit....   

[stereofool]

I feel my pocetbook starting to pucker-up 

[Carlman]

I haven't heard them yet, nor do I expect to for a while.  I'm out of town Sat-Wed so I'm hoping to hear them when I get back.. I'm curious and excited and trying not to let my hopes blossom into unrealistic expectations.  I think it's probably like being in the waiting room during your wife's breast enlargement surgery.
-C

[Carlman]

I'm posting a timeline and emails from Rick (with his permission):

This is the initial assessment after Rick had some time with the speakers:

--- On Mon, 8/31/09, Selah Audio wrote:

Carl,

I spent some time listening to the Piega today. Here's a
few notes I made on the sound.

1) Driver integration needs some work. I could hear
shifting of sounds vertically when there was a transition in
frequency from the lower to upper mids. The mid/tweeter
panel at times really seemed to overwhelm the output from
the woofers depending on the source material.

2) Overall balance on some recordings was good but on
others the top two octaves were too prominent. Everything I
played is mixed fairly well so I could see where poorly
recorded stuff could drive you out of the room with this
speaker.

3) Hardness / edginess / congestion was evident at times in
the upper mids / lower treble. Again, the speaker isn't very
tolerant in this area but I had suspected that would be the
case given the mid and tweeter design.

4) Bass at times was bloated and lacked pitch definition.
There seemed to be a resonant mode - not sure the reason for
that but the measurements will help reveal the source of the
problem.

5) Integration of the mid to the tweeter was good and the
coincident layout is working well. Vertical dispersion for
angles above the tweeter axis was pretty good.

6) Lower mids were lacking "body". Some of this may have
been intentional in the voicing the designer prefers. I also
suspect the crossover to be a problem in that area as well.

7) Horizontal off-axis response was not as good as what I'm
used to. In the top octave the tweeter's width affects that
so there's not much you can do with that (not that it's bad
but I've been listening to the Raal ribbons quite a bit
lately). Once I measure I'll see if the crossover is an
issue as well.

8) Sensitivity is low and this speaker really sucks power
from the amp. No way it's close to being 89db. Like we
discussed I think the crossover is sucking some of the life
from the speaker with inductor losses and phase angles.

Overall I think there's definitely room for improvement and
redesigning the crossover should help improve most of these
areas.

[Carlman]

Another email from Rick with technical measurements and values:
Lots of testing today.  Explanations below:

PWFREE  - free-air test of the woofer. Fs=41.5hz which indicates their 28hz
claim for bass extension is a little "optimistic"


PUWNEAR : Uppermost woofer measured nearfield. The port contribution won't
show here so the extension is only 55hz/-3db and 41hz/-10db


PLWNEAR: The bottom woofer summed with the port. 48hz/-3db and 31hz/-10db
which is normal given the Fs measurement of the woofer. Not bad extension
but falls short of what I would expect for a speaker at this price.


PONAXIS: On-Axis measurement, mic at tweeter height. The elevated upper
midrange and rising response in the top octaves that I noted in listening is
evident here.


POFFAXH: Response is pretty flat a 30 degrees horizontal off-axis. Typically
this would be tilted downward a little as you move up in frequency response;
however, since the on-axis has a rising response the off-axis has more
energy. In a room with little or no treatment and / or walls that are close
on each side the speaker will sound too forward / bright.


POFFAXV: Vertical response on-axis at 1M. Measured from the top woofer to
about 6" above the top of the cabinet. Very little change at all angles and
shows that the concentric mounting of the tweeter works well.


PIEIM: As expected from looking at the crossover the impedance shows a
difficult load with large phase angles. Given the low sensitivity this is a
pretty difficult load and high current solid state amps will fare much
better than tube amps.

[richidoo]

Even just biamping the ribbon array my 100W tube amps did not fare well.

It gonna be much better than a boob job. Because you're gonna be able to play with them whenever you want.   

[Selah Audio]

When you modify a speaker you never know what to expect. I've had a few unusual speakers come in from two other manufacturers that reside at Audio Circle. Sometimes I tell the owner it's not worth it as was the case with one where the driver combination and design execution sucked really bad. I felt for the owner but basically told him what the problems would be before I even measured it.

 The second one was made by a company who gets quite a bit of hype on the net. Upon removing the drivers I found out that the woofers didn't match. From the outside they look the same but apparently they never bothered to do incoming inspection or test them before shipping. After measuring the response of both I found that the wrong woofer had been used for the crossover design and so the crossover wasn't optimal for the cabinet (sealed cabinet but woofer intended for a ported cabinet). I told the owner that he should be compensated because he paid for expensive crossovers that were unusable with his specified drivers. When confronted with his error the owner of the company refused to give him a refund - so much for great customer service!

[Selah Audio]

The replacement parts arrived today

[Carlman]

Yay!  Really looking forward to hearing them...  I'm all excited like a little kid!

[Selah Audio]

Yay!  Really looking forward to hearing them...  I'm all excited like a little kid!

The initial listening test started today. Originally the crossover had 63 components (19 inductors, 24 caps, and 20 resistors). Right now the new crossover has 12! The first change is a huge increase in sensitivity. Only one resistor! Overall response is more linear, greater dynamics, and a cleaner presentation. Now on to fine-tuning the voicing.

[Hantra]

I always knew in my mind that the Piega crossover was the real culprit in how it sucks power down like nobody's business. 

I've thought about upgrading mine for a long time, but I have always been terrified of how it would turn out.  I just assumed the extreme complexity in the crossover was necessary to match the wildly different drivers.  I will be anxious to hear the result.   

[stereofool]

I'm with you guys...if a substantial improvement, then this may be in my future, as well .

[Rob S.]

Carl,  think you'll have these by Saturday or Sunday?

Also- did you go with Rick's recommendations on parts for the crossover?

Rob S.

[bpape]

Yay!  Really looking forward to hearing them...  I'm all excited like a little kid!

The initial listening test started today. Originally the crossover had 63 components (19 inductors, 24 caps, and 20 resistors). Right now the new crossover has 12! The first change is a huge increase in sensitivity. Only one resistor! Overall response is more linear, greater dynamics, and a cleaner presentation. Now on to fine-tuning the voicing.

63 to 12 - that's quite a difference Rick.  If it can be made balanced, I'm all for the KISS principle.

Bryan