You've obviously never seen the schematics of modern AVRs. There are very few opamps in there as they typically go straight to a custom input IC that includes an ADC. Signals are processed digitally then sent to the DACs and the power amps.
Bollocks.
It's clear your opinions are based upon ideology not engineering.
That's true. Some rigid cones require complex compensation circuit to tame the breakups. This is hard to do and in some cases not possible to do with active. Imagine you need to create a deep (high Q) notch filter. You cannot do it with a simple one opamp. A gyrator is not deep for such purpose.
You could use a bootstrapped twin-tee filter, capable of a very narrow notch with two opamps. Its harder to do this with inductors because high Q off the shelf inductors are jolly hard to find. Then if you find them their tolerance is going to be worse than for capacitors and resistors as used in the twin-tee.
More bollocks. It's as easy as pie with a DSP, and it's repeatable and exact which you can't do with passive especially as the target Z moves around with driver position in the gap.
Plus an easier way to design with drivers with lots of breakup is to not run the driver until close to the breakup, rather use drivers with lots of range overlap, filter steeply and add in an extra notch (much smaller now) for the breakup mode. I've also done this exact same ting with analog actives.
You can do anything with digital. An old bad sounding analog tape can be remastered into high resolution audio file.
If we had that luxury.
I believe you know that not all passive crossovers are equal and not all active crossovers are equal. It is how good you are with any or both of them. Then at certain point and situation, you may have to combine both for best results.
IMHO, after 300+ messages, THIS response is STILL the most sensible and truthful one.
Lots of ideological gut-reactions on this thread with little to do with actual experimental evidence.
Also, I think that many of those who claim that "active is superior" are not capable of competently designing a passive crossover, and then sure, using DSP tools it's easier to arrive at something acceptable by sheer trial and error.
My 2 cents.
Their incompetence is probably because they haven't got the necessary parts at hand and means to swap quickly. One has to plan this ahead.
I disagree. There are many knowledgeable and experienced builders here but most do not see the big picture. It is so sad that there is no better way than trusting others who seems to 'get it'. But better not assume that big names must get it. Better use logic.
No: not use logic. Measure, simulate and optimize, build, measure again and again. Only then you are you allowed to listen. Marco, Lojzek and Earl Geddes are right.
It's all in the transfer functions.
It's all in the transfer functions.
When it comes to passive vs active there are certain laws of Physics that dictate the behavior of each. One thing you can state categorically is that unless the source amplifier has zero output impedance there will be crosstalk between the drivers in a passive system. Add to that the impedance of the speaker wire between amp and speaker terminals, shared grounds, stray fields resulting from the layout of the passive components, etc, and all these things make the picture worse. Back emf generated by a woofer will find its way to the mids and tweeters etc. You can argue as to the audibility of such but you can not deny it.
You can test this easily. Disconnect you speaker cable from the amp. Connected a 0.1 ohm resistor across the amp end of the cable (nominal damping factor of 80). The connect a scope across the speaker's binding posts. Then tap on the woofer cone. Adjust the scope voltage sensitivity and you will see a voltage across the speaker's input terminals. If it's at the terminals, its going to find it's way to the other drivers as well.
These things aren't a factor in active crossovers. And, as a designer of my NAO speakers at one time I offered a fully active version of my NaO II. The active version duplicated the transfer functions of the passive crossover for the main panel. There were certainly differences in the sound, perhaps small, but a definitely an increase in clarity and definition. Today I only offer fully active versions with DSP crossovers with performance superior to anything I could design and build using passive or active analog components.
You can test this easily. Disconnect you speaker cable from the amp. Connected a 0.1 ohm resistor across the amp end of the cable (nominal damping factor of 80). The connect a scope across the speaker's binding posts. Then tap on the woofer cone. Adjust the scope voltage sensitivity and you will see a voltage across the speaker's input terminals. If it's at the terminals, its going to find it's way to the other drivers as well.
These things aren't a factor in active crossovers. And, as a designer of my NAO speakers at one time I offered a fully active version of my NaO II. The active version duplicated the transfer functions of the passive crossover for the main panel. There were certainly differences in the sound, perhaps small, but a definitely an increase in clarity and definition. Today I only offer fully active versions with DSP crossovers with performance superior to anything I could design and build using passive or active analog components.
I have been using a balanced topolgy for the Woofer branch of my crossovers to minimize this (see e.g. attached diagram for a 6th order LP).
Have you ever tried it? I welcome your thoughts on this.
Marco
Attachments
Hi John,
Sure, its going to happen, but I simply cannot see this as an audible factor. The same filters that filter the incoming signal will filter the crosstalk signal. For example, the back EMF of a woofer is going to be highly concentrated at resonance (that's why there is an impedance peak there.) This signal will be pushed down in level by many orders of magnitude by the tweeters passive filter - the back EMF is no different than the incoming signal to the tweeter.
Done it.
My experiences mirror John K.'s: Active the bass is audibly cleaner and more defined. I suspect that the usual series inductor works a bit like the audio equivalent of a soft-focus lens or it is the series resistance or a bit of both.
Dunno.. I have actives and long (decades) fiddled with crossovers for My tannoys and frankly the Active gives (arguably) no audible advantages.
True.. as I've long tried to hear such and anyone who visits gets press-ganged into listening for differences 😱.
Yess fooling with the LF inductors ESR does alter bass ... some.
Risky though, as balance is affected.
But, IF that's all you notice ? Inductors are Cheapish (even Hi zoot ones) and a helluvalot simpler to integrate than a fully setup active.
Frankly a 'Better' Amp gives equal or better results, in my experiences.
That said; this is a hobby and everyone has differing views 😉
True.. as I've long tried to hear such and anyone who visits gets press-ganged into listening for differences 😱.
Yess fooling with the LF inductors ESR does alter bass ... some.
Risky though, as balance is affected.
But, IF that's all you notice ? Inductors are Cheapish (even Hi zoot ones) and a helluvalot simpler to integrate than a fully setup active.
Frankly a 'Better' Amp gives equal or better results, in my experiences.
That said; this is a hobby and everyone has differing views 😉
I have built, auditioned and measured identical speakers in both active and passive for decades and my conclusion is "Yea there is a small difference, but not great." Ten years ago when I studied this situation in depth I concluded that the costs were not justified by performance changes either through measurements or auditions. But back then DSP units were exceedingly expensive. Today my system is active and the DSP is built right into the amps, so its not so expensive, almost a wash cost wise with passive. It was much harder to implement the active than the passive.
But I have never found a substantial audible or measurable improvement from active. In fact, it is a small enough audible change that I cannot say that it is better or worse, or if I am even imagining the difference (knowing that there is a small measurable advantage.) So I am very suspicious of claims of substantial improvements. (of course the passive designs might have been flawed and the active was not, that's possible, but when optimized passive is compared to optimized active, the differences are just not that great.)
But I have never found a substantial audible or measurable improvement from active. In fact, it is a small enough audible change that I cannot say that it is better or worse, or if I am even imagining the difference (knowing that there is a small measurable advantage.) So I am very suspicious of claims of substantial improvements. (of course the passive designs might have been flawed and the active was not, that's possible, but when optimized passive is compared to optimized active, the differences are just not that great.)
It doesn't necessarily work that way Earl. If you look at the woofer as the generator it sees a very different load than the amp. Passive components in the tweeter network can interact with passive components in the woofer network (simple 2-way) creating strange looking band pass filters. For example, in my original NaO II design simulation with a realistic cable impedance and amplifier output Z showed at any emf generated by the woofer could potentially reach the tweeter through what looked like a very high Q band pass filter with center frequency somewhere around the nominal crossover point. The amplitude of the filter at the center frequency was only about 10dB below the nominal tweeter level at the frequency.
Take a simple 1 st crossover for example. The amp sees the tweeter in series with its cap, in parallel with the woofer in series with its inductor, with series cable impedance in between. But from the woofer's perspective, woofer sees a series LC between it and the tweeter, with the amp output Z in series with the cable Z, in parallel to the tweeter and its cap.
Audibility is a reasonable question. I imaging there are situation where it isn't a factor, other situations where it is. I think this may be one reasons why two speaker with similar HP and LP response that sum to a similar over all response can sound different, just because the crossover topology is different. I guess, it's part of the art of passive design. You know, it doesn't matter until it does. 🙂
Possibly the limiting factor with active crossovers is the DAC. The DAC can make a huge difference to the final sound with DSP. I use an active analogue crossover, that way I can afford a good DAC but have limited control on the final result. Not a bad thing sometimes.
- Status
- Not open for further replies.