Welcome! Need support, you got it. Or share your ideas and experiences.
Return to Planar Speaker Asylum
84.132.61.220
In Reply to: RE: active crossover install posted by tberd on October 12, 2016 at 06:44:50
Hey tberd
Constructing a crossover (active or passive) is no trivial pursuit.
Itīs actually very challenging to get phase, impulse and amplitude of each of the drivers to work properly together.
-and in this regard, the Bryston crossover is simply to limited in itīs fixed 24dB/oct. setting.
There is good reason that the stock crossover is asymmetrical and that the mid-driver is inversed, and I must report that I had my best audible results in "close-to" factory settings.
Digital crossovers have often the (huge!) advantage that they can time-align the drivers (especially delaying the bass) and improve coherence.
I tried hundreds of settings on my former MG3.3Rīs via digital crossover, and I can testify, Iīll never ever have it again without proper measurement assistance.
If youīre not able to exclude analog or digital streaming playback, there is only customised active crossovers to help.
However : the sheer dynamics, increased focus and breathtaking glory that (well setup) active Maggies can produce, is worth the extra hazzle in getting a well-configured X-over.
Kind regards L.
cMP2 Computer // Prism Sound Lyra // Spectron Musician III // Analysis Audio Omega
Follow Ups:
Even my "Simple MMG biamp project would have been a nightmare without measurement - a passive crossover often is also an equalization timing and phase adjustment. When you reconfigure a passive loudspeaker to active, you need to measure before and after - just like the manufacturer does when he develops the loudspeaker.
Many manufactures and designers start with an active EQ - and then implement something similar with the passive components.
"The hardest thing of all is to find a black cat in a dark room, especially if there is no cat" - Confucius
"Even my "Simple MMG biamp project would have been a nightmare without measurement - a passive crossover often is also an equalization timing and phase adjustment. When you reconfigure a passive loudspeaker to active, you need to measure before and after - just like the manufacturer does when he develops the loudspeaker."
Actually, you don't. In fact, it's better if you don't.
Passive to active conversion (assuming your intention is to leave the existing transfer functions intact) is a very simple approach, and requires no acoustic testing of any sort. Neo has provided much of the required conversions in his spreadsheet and I provided a 3.6 working example many years ago on this very forum. (In fact, it's in use successfully by a number of PA lurkers.)
http://www.integracoustics.com/MUG/MUG/tweaks/davey/mg3.6xo.htm
If your intention is to re-engineer and/or improve the existing design(s), you better know what the hell you are doing regards acoustic measurements. :)
Cheers,
Dave.
Well my experience is otherwise. even if you want to mimic the existig transfer function you have to measure whet the transfer funtions are (not just the component values, because the driver has to be measured as pare of the transfer function. Even level matching LF to HF is better done with measurements - or at the very least customized band limited pink noise signals and the ability to quickly compare levels. but even you mind is terrible at distiquishing tone difference as your memory gets biased. This requires more rigor tobe employed in your set up. You have to listen to the HF noise signal first and then climb the LF noise signal to match, the repeat with the LF Fixed and adjust the HF upward to match and see how the end results compare - they were different by a couple dB each time I tried it... Using measurements before and after is far easier, and you likely can spend less than you cable budget to rent a very high quality measurements device... See link - Full disclosure I work for B&KOf course, the great advantage of a Active crossover is veer away from the manufacturers design and see if you can make it better! The mananufacture always applies some limitations in the crossover design - related to cost, physical layout of the transduser - and let listening room may also place some restrictions on the performance with the standard passive design. With my MMG's I tried steeper slopes and higher and lower crossover frequency. I ended up with a bit higher crossover (1100Hz) and 24dB/Oct slopes and added a EQ and HP filter to better integrate with my subwoofers.
"The hardest thing of all is to find a black cat in a dark room, especially if there is no cat" - Confucius
Edits: 10/13/16
Sorry, I couldn't disagree more (see my post below). Granted some manufacturers often make compromises in the design or, more likely, in the implementation (quality of parts). However it is just as easy to use a 17.3 mH inductor as a standard 18 mH inductor when ordering in quantity. Magnepan was not cutting costs when they used a 4th order LP filter for the 3.6 when all their previous designs used less costly (less parts) 1st, 2nd or 3rd order filters. They may have save money by using lower quality parts but I don't feel that they compromised the design. BTW when I measured the values of the caps and inductors in my 3.7i, they were spot on with the design - close tolerances suggests that corners were not cut.
Obviously I can't argue with you that your configuration for the MMG sounds better to you than the original design but this is proof of nothing. It's anecdotal and may, as you noted, be particular to your room acoustics and/or your preferences. The bottom line is that you have arrived at a happy solution and that is great but I would not generalize from that experience.
As far as measurements, as Davey notes, my spreadsheet is based solely on mathematics - no measurements and no verification. Using the appropriate math, a one-to-one translation of a speaker-level crossover to an ALLXO will produce the same electrical response and, assuming no modification to the actual drivers, will give the same fs. There are other benefits from active biamping but changing the fs and phase is not one of them.
That said, I do use acoustic measurements to help set up my speakers and I find it beneficial. But this is done to position speakers and tune the room, not to modify the crossover.
I married the perfect woman. The downside is everything that goes wrong is my fault.
If you read again, I said "requires no acoustic testing."
Electrical-only testing (at each transducers terminals) may be required, and it has the distinct advantage of not containing any of the pitfalls acoustic testing (or listening) might. It also incorporates all parasitic characteristics of the crossover components.However, computer simulation will achieve nearly the same level of accuracy electrical testing will......assuming the components in the simulation are fairly good. Simulation of Magnepan transducer and crossovers are more straightforward in this aspect since the driver impedance's are essentially resistive. Neo's spreadsheet contains electrical transfer functions of Magnepan speakers he's never measured (either electrically or acoustically), yet I believe them to be very accurate depictions.
Cheers,
Dave.
Edits: 10/13/16 10/13/16
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: