In case you did not notice, I use a transformer-based stepped attenuator on the compression driver. This is comprised of an actual autoformer, which "multiplies" the impedance "seen" by the filter by a very large factor, and an appropriate "shunt" resistor in parallel to the autoformer, which brings the impedance back to a very stable and "flat" ~8 Ohm (actually, about 7.7 Ohm +/- 0.1 Ohm as measured).
So whatever impedance variation the "raw" compression driver may have does not matter at all anymore.
Marco
Exactly.
Blanket statements like "active is better than passive" are just oversimplification. There may be pros and cons to either, depending on how they are implemented and on what needs to be done to get the response to match the intended target.
Marco
9
I also use autotransformers with a swamping resistor in parallel. But I still got more low frequency signal to the mid driver with my attempt of a passive crossover than when using the same 2nd order crossover on minidsp. And the gritty and dull sound of the large inductor was gone. And I get the strongest possible signal to the mid and tweeter horn without too much attenuation. I only use minidsp and a protection cap now and I like it. I don't need to worry if the passive crossover is designed properly in accordance to the impedance variation of the driver. How do you measure impedance of the driver when everything is hooked up?
Just the fact that I would need to attenuate a driver 15dB or more to match with the woofer, is a good reason enough not to go passive. L-pad or autotrafos, the amp and the driver works best with the strongest possible signal. Large coils is another good reason. And with my limited knowledge of passive crossovers, I also risk ruining the driver with too much bass. I may not be an expert but I have ears and I no likey big coils in my midrange. Btw, Marco. Your horn is too small. Sorry, that was below the belt.
I also use autotransformers with a swamping resistor in parallel. But I still got more low frequency signal to the mid driver with my attempt of a passive crossover than when using the same 2nd order crossover on minidsp. And the gritty and dull sound of the large inductor was gone. And I get the strongest possible signal to the mid and tweeter horn without too much attenuation. I only use minidsp and a protection cap now and I like it. I don't need to worry if the passive crossover is designed properly in accordance to the impedance variation of the driver. How do you measure impedance of the driver when everything is hooked up?
Just the fact that I would need to attenuate a driver 15dB or more to match with the woofer, is a good reason enough not to go passive. L-pad or autotrafos, the amp and the driver works best with the strongest possible signal. Large coils is another good reason. And with my limited knowledge of passive crossovers, I also risk ruining the driver with too much bass. I may not be an expert but I have ears and I no likey big coils in my midrange. Btw, Marco. Your horn is too small. Sorry, that was below the belt.
Last edited:
The "strength" of the signal that reaches the driver is exactly the same regardless of where the attenuation takes place. If it weren't, it would sound louder, and hence out of balance with the woofer, in the active setup.
Now, that is a very good reason (for you) to go active
Sure, do as you like best. I remain unconvinced that the reason you don't like the passive setup is due to the "big coil" (I think it is due to an improperly designed passive crossover).
Btw, Marco. Your horn is too small. Sorry, that was below the belt.
Sure, let's keep on with the blanket statements
A larger horn would indeed be required if I wanted to cross over to the Woofer at a lower frequency (which may be a good idea in some respects, but not such a great idea in others... but this is getting out of topic, so I'll just leave it at that.)
Marco
I only know what I hear. An L-pad or autotrafo with around 15dB attenuation does suck the life out of the music more than with little or no attenuation. Your amp might sound exactly the same 80% or at 5% volume but mine does not. The sound is much cleaner with less attenuation on mids and tweeters.
At least you don't need to worry about blowing up the driver with such a high crossover point at 750 hz. Have you really measured the impedance at about 200-500hz with the swamping resistor?
At least you don't need to worry about blowing up the driver with such a high crossover point at 750 hz. Have you really measured the impedance at about 200-500hz with the swamping resistor?
Last edited:
Once again: if the end volume (SPL) is the same, the actual voltage driving the compression driver must be the same also.
If your amp sounds noticeably worse at 80% (when driving the compression driver with the attenuator in place) than at 5% (when driving the compression driver directly), then it probably just means that it is underpowered for the first application, and that it therefore distorts (clips) during transients. In that case, of course, using the same amp in the active setup is a much better idea.
It still has nothing to do with the "sound of big coils" or "of attenuators", however.
For the record, I cross at 850 Hz.
Yes, and it is flat.
Marco
Ah!
So you are comparing apples to oranges in terms of the individual levels of the Woofer and Mid, and therefore of overall frequency response?
If that's the case, then all bets are off - and that explains a lot, actually!
I don't mean to come across as patronizing, but I think that designing a good crossover (passive or active) requires quite a bit of study, and not just a few "hints".
Marco
I've been fiddling with a DSP for crossing 2 and 3-way systems for a couple of years now and believe me, the smallest of adjustment make a huge deal in phasing (sometimes it a game of opposites). I have to hand it to those who can implement a seamless linear passive cross-over. The best speaker I've had with such good luck with crossovers was Volent.
After two years of up and down and all over the place I now run my 10" 94db efficient mid-woofer out of the DSP, 1st order with a Zobel (6db down at 2200hz). The CD runs through the DSP and gets a little EQ and time correction, but more importantly Q matching to the woofer. The P-Audio SD44-BN CD is then crossed passively with a second order at 3240hz but is down 6db at 2000hz (no DSP crossing).
It takes a tremendous amount of time getting phasing correct, even with DSP. I've had phasing correct probably fifty times but always look for a little change hear or there and loos it again.
Now that I'm running passive and have it bang on I'm finally listening to performances with pass,,, and not fiddling.
What do you do when you correct phase in a passive crossover? I think passive is great for upper mid and treble, but it becomes to unmanageable (at least to me) when crossing the woofer and the mid with large coils. I never found a good enough sound when crossing around 500hz passively and 2nd order or higher.
It sounds like you have come to conclusions without the requisite effort to really explore passives. Some of the best systems made are passive. A nice thing about passive is the reliability and durability.
As an example: amp or preamp pops or anything of the sort can readily fry tweeters in active systems. The solution often used is to use a protection cap- but the cap, unless big enough to be near-zero impedance in the passband, is going to add it's own frequency shaping, and interact with the impedance peaks common in compression drivers (domes tend to be more benign from an impedance perspective). If it's big enough to make that a non-issue, it's not going to be doing a great job protecting.....
and so on.
Interesting. I usually find that it sucks out the amplifier noise and impedance dependent frequency response variations.
As Frank sings - "That's life, I can't deny it."
That's mostly my take as well. I have strongly opposed the idea that active is inherently better as I don't think that is the case. I will say that now that I make speakers either active or passive, the active offers some degree of added flexibility that can be very useful at times. But audibly the differences are minimal. Active can be a real PITA to implement in a home theater using an AVR. May not be worth the trouble.
I am going to finally replace my aging speakers with the newer ones and I plan on doing that active. If I find a big difference I will post it somewhere, so time will tell.
PS. Badman - even in my active systems there is always some passive parts for protection. They are simply accounted for in the active settings.
Last edited:
I don't have a crossing at or near 500hz. I run the 10" mid-base full out at the bottom and 1st order at around 1500hz, 6db down at 2200, but it has a zobel circuit. The tweeter matches nicely with wires inverted/out of phase. When a sub is added it goes through the DSP. My work has been mainly directed at not needing a sub and feel my Transmission line with this wonderful woofer does it, in a room close to 100 cubic meters.
Earl - a quick observation re AVRs & "active" speakers - depending on how far up the model line you go, it may not be as much of a pain in the a$$ as one might think - well aside from the number of components / channels of amplification and interconnecting cables required- but that'd be the case as well in a surround system with separate processor / amps.
At some point in most brands' model ranges you'll find line outputs for all of the main front row, surround, and LFE channels. Going active with any of those channels would be a relatively simple matter. At the same point where that feature is found, will almost certainly include the ability to assign up to 4 amp channels to any of the surround channel locations, to passively bi-amp the front mains, or in certain models to use the onboard DSP for digital XO and active amplification of the front L&R mains. My Onkyo TX NR818 has that feature, and while it'd be nice if the centre could be similarly handled, it's certainly proven handy for testing and demo-ing FAST-type 2-way systems. XO points (but not slopes), levels, and woofer/tweeter offset distances can be fairly easily adjusted via menus displayed on the TV screen.
At some point in most brands' model ranges you'll find line outputs for all of the main front row, surround, and LFE channels. Going active with any of those channels would be a relatively simple matter. At the same point where that feature is found, will almost certainly include the ability to assign up to 4 amp channels to any of the surround channel locations, to passively bi-amp the front mains, or in certain models to use the onboard DSP for digital XO and active amplification of the front L&R mains. My Onkyo TX NR818 has that feature, and while it'd be nice if the centre could be similarly handled, it's certainly proven handy for testing and demo-ing FAST-type 2-way systems. XO points (but not slopes), levels, and woofer/tweeter offset distances can be fairly easily adjusted via menus displayed on the TV screen.
Lynn said "No DSP"
If Badman had bothered looking, the reasons why active crossovers, and particularly DSP based active crossovers, were discarded were clearly stated by Lynn about three years ago.
Nothing has changed since then.
Most DACs still suck and IT coupled PP300B amps are still big, hot and expensive.
That's before you get into the now-additional minefield of digital filter design & construction (and yes, signal processing was one of my majors, followed by a decade implementing complex embedded systems )
* the exception was analog active for the subwoofer crossover.
If Badman had bothered looking, the reasons why active crossovers, and particularly DSP based active crossovers, were discarded were clearly stated by Lynn about three years ago.
Nothing has changed since then.
Most DACs still suck and IT coupled PP300B amps are still big, hot and expensive.
That's before you get into the now-additional minefield of digital filter design & construction (and yes, signal processing was one of my majors, followed by a decade implementing complex embedded systems )
* the exception was analog active for the subwoofer crossover.
True active is not always better, but if you can handle the problematic areas while minimising higher power components, it usually has more potential of being better. There are too many things involved and need to be tackled one by one as you gradually reduce system hardware. In many cases, taking the passive crossover out of the system actually reveal deficiency upstream which can present an unpleasant experience which is often construed as being worse where in fact the problem was always there, just masked by the passive crossover.
The interfacing challenge is one place where generally less tackled. Each individual component can measure well, but when you connect them together, the interaction is generally ignored. Additionally, the methods of testing to get to the true problem is also not widely used. People use general acceptance/qualification test standards, these are generally not fine enough for diagnosis.
Yes, how dare I fail to look back through years and nearly 14,000 posts. Did I say anything that was arguing with Lynn's decision? Chill.
