Fully active setup, worth it?

[Jussi]

Hi,

I have a pretty complicated and crazy 4-way system under construction (finally!). Naturally fully activating such system requires quite a stack of electronics.

Is fully active setup really worth it?

Better control over each diaphgram, less distorsion(?), better use of efficiency...

Perhaps I should tell a bit more of the system. It's pretty much Audio Artistry inspired 4-way open baffle system. First I'll finish the main panels and then add woofer colums later. Original Beethoven used passive crossovers with its mainpanel. Later Elite and Grand models, Phoenix kit and Orion are fully active designs. In both Orion and Phoenix cases I understand why. Crossover frequences are pretty critical and drivers are pretty much extended to their limits (1400Hz cross to tweeter) and low crossover points (100Hz 2nd in Phoenix, 4th order 120Hz in Orion) would require huge passive components.

Then there are systems like NaO that uses half-active approach using passive upper cross in the panel and has active cross for the lower crossing.

My system has Dynaudio Esotar tweeters in waveguided setup, twin W18E001 Excels for midrange duty and twin 25W/8565-01 Scan Speaks for lower mids. Woofers determined later. With these aspects few things are different to Orion/Beethoven/Phoenix systems. Uppermids are smaller and crossover point to tweeter could be somewhere around 2-2,5Khz. Lower cross can be done with gentle 1st to 2nd order loop around 250-300Hz. Lowest cross to separated woofers is going to be active. And active highpass for the mainpanel would allow some active response shaping where it's required. Mainly 100-500Hz dipole equalizations comes to my mind. Therefore most passive components would be just for the crossing, not that much needed to eq the response and therefore most of the efficiency is also saved.

What do you think? Half active with two stereoamps or fully active with a whole lot more stuff. Is fully active worth it in this case?

Thanks!

Jussi

[BobEllis]

I'm a fan of fully active systems. This is largely because I am more likely to carry the design through to the end if I only have to change a few cents worth of resistors than more expensive passive XO parts. I already own a bunch of amps so that isn't a consideration for me.

Take a look at the FR plot of the W18E001. The main breakup mode starts rising just above your proposed crossover point. It will take a lot of effort to tame that, if you can tame it at all crossing above 2K. I'd look to cross to a suitable tweeter as close to 1KHz as I could.
Nothing but money says that you cannot cross passively at very high slopes. Search for the Modula MTM on the HTguide forum for an example of how to make a Cauer-Elliptic filter that approximates an 8th order L-R in the crossover region. It suppresses breakups by 50 dB or so, and then reverts to a 4th order roll off.

For the price of crossover like that, you could build a multi channel gainclone and an active crossover. Do you see where I am going

Another issue with a high crossover and MTM is that you'll get a big change in directivity and lobing issues at the crossover point. Your center to center distance will be greater than a wavelength, and you are starting to get some beaming with a 6" driver at 2kHz.

Seems like S-L had something there - The more I try to come up with something like Orion the more I end up being driven back to the very similar ideas. Maybe try an updated Phoenix - like an MTM version of the Orion? If you need more output add another set of W22s crossed in to counter the dipole roll-off before crossing to a quartet of XLS12s. That's what I would do if I had the money and room.

[john k...]

(JPK) Hi Jussi. You will probably never get a truly unbiased answer to this question. There are positives and negatives to both fully passive and fully active approaches.

And then there is my take on hybrid design.

http://www.musicanddesign.com/HybridDesign.html

I see little point or advantage to driving a tweeter from a separate amplifier provided the LP filter for the midrange driver can be suitable designed. If the mid is a metal cone with nasty breakup, the breakup may be easier to deal with in an active circuit, but if it can't be handled passively perhaps the better solution is to look for a driver with more satisfactory response.

On the other hand, the NaO AEP used the W18E001 with a 2.0k Hz crossover point and a passive mid to tweeter x-o with no problems.

The all active approach appeals to some because you can get out your EE filter text book and cascade stages of HP, LP notches, shelves, delays, etc, to shape the response as you like in a brute force engineering approach. Passives, or all actives designed to minimize the number of active stages, are a little more imaginative.

I would think your main 3-way system would be ideally suited for a hybrid approach. If you want to add an additional sub later, then that would probably be easiest to do with a power sub and active x-o to the 3-way.

[BobEllis]

John is much more skilled at this stuff than I. Idon't doubt that the NAO sounds good, but I don't think I could come up with a design that crosses that high and sounds good.

As for cascading textbook filter sections, guilty as charged. At my level of (in)expertise I find it easier to get good results that way than trying to be clever and make each section serve multiple functions. I've played with simulations and think I can get something that will work passively for many situations, assuming that the impedance curves are accurately measured.

I already have plenty of amplification and a room where nobody objects to lots of equipment, so I can indulge my bias/fantasy. Staying passive for Mid-Tweet crossover is a good compromise for many, though.

[Jussi]

Bob,

I know W18E problems. But they are also implemented pretty well with 2-2,5Khz crosses in Ellis 1801, Seas kits Thor and Odin and so on. Unfortunately I don't have measurements from any of them but I'll make some from my own cadget to find best possible xover point when it comes to distorsion and ringing.

John,

I know this is one of the infinite questions and there are numerous different aspects to think it through.

Here couple points I can think of:

1. How big advantage active system really has so it compensates poorer overall electronics. Bigger pile, less quality for each instrument.

2. Top quality active crossover isn't easy task. There are still some advantages in passive components.

Jussi

[Jussi]

John,

Did you actually try NaO with all active crossover?

If different frequency ranges are played with similar sensitivity drivers (not that much resistors in passive crossover), responses are matched and phasing works, is there really that much difference between active and passive crossovers? Difference in transparency? Integration?

Normal 2-2,5Khz cross doesn't even need that large components and there are plenty of low loss components available. That 250-300Hz 1st to 2nd order is a bit different thought...

Jussi

[john k...]

(JPK) I designed a fully active version of the NaO II as an option. The fully active design matches the transfer functions of the hybrid system as closely as possible. I have played both systems at length, even playing one channel fully active and one channel hybrid. Driven by the same electronics I can not reliably identify which system is which.

[Cloth Ears]

Agree with John K, regarding the tweeter, although I tend to look at 4-ways unless my mid has quite a wide range.

Current set-up is a non-DIY 2-way actively crossed to bass bins. Working on a subwoofer (also to be active). Also working on (with same bass bins) an active 4-way (but mids and tweeters to be passively X-ed) and an active 3-way (probably also with mids and tweeters passive).

The cost of good passive components in Australia starts at around $10-20, the next step up being $60-100 (and then stratosphere). Even a fairly simple passive can cost you the same as a cheap eBay sourced amp and a Behringer CX-2310. The sound may not be as good, but it's very easy to experiment with.

[AJinFLA]

John is correct, although he forgot to mention one very important (IMHO) other possible advantage to the all active approach. Adaptablity.
There is no reason why one would want to change (for example) the XO frequency from mid to high after the final design stage. So passive vs active there is a wash.
But what happens if you decide (or the SAF ) to rearrange the room. Or move the speakers around the room. Or move to a different room or house. Maybe an environment that is a bit more lively and reflective. Where you want to adjust the voicing on the treble or even midrange. Something like my DCX can easily adapt. Doing so passively at the xo stage would not be trivial. One can try passive correction to the room itself, although I find this method to be a bit too non-linear and somewhat bothersome to not do it actively.
Or at least in conjunction with active correction. My 2c's.

cheers,

AJ

[john k...]

(JPK) Hi AJ. Voicing is an area where the hybrid approach can be useful since it is possible to control the balance of mid to tweeter in the acitve stage. On the other hand, I always design my passive sections to have a couple of dB adjustment on the tweeter level built in.