-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
Sometimes a passive crossover is just stupid...
- Thread starter Waxx
- Start date
terryforsythe
Major Contributor
Here is an article that discusses the issue: https://sound-au.com/biamp-vs-passive.htm
terryforsythe
Major Contributor
I'm done with passive crossovers.
But, many people do not like the idea of active speakers. A common perspective is that if a plate amplifier gives out some time down the road, a replacement that works in that speaker may be hard to find. Also, people like to choose their DACs, amplifiers, etc.
I pulled the passive filters out of my speakers and kept the electronics separate. This addresses the issues in that system. I do have a pair of KEF active speakers, though, in another room. Hopefully they last a long time.
OP
Waxx
Major Contributor
- Thread Starter
- #24
I know all the disadvantages of passive crossovers, but reality is, most speakers still have them because (good) dsp is still to expensive for the average customer. And the hassle to have several amp, a lot of cables and so is too much for them.
And when it's prebuild, the amps go down to fast... and you end up with a box that is trash because of the dead amp. That's a very important reason for a lot of people to not go active. They sound of passive crossovers (even crappy ones) is good enough and the convenience wins mostly. Even for me, i like them when done right and the convenience is also good. I can't afford 6 MiniDSP flex's with DIRAC (i have 6 setups in my house). This guy actually can, he's a bussiness man with a well running construction company that gives him a lot of profit.
And let's be honest, tons of good speakers are build with passive crossovers, they measure good (here or with Erin) and sound good. But yeah, dsp can do a lot that passive can't, and avoid a whole leap of issues yo need to watch when designing.
He wants a garden system that is high power (woofers are double 18" reflex cabinets) but insist on all analog. Even the crossover to the woofers will be analog, active analog that is. He says (HE SAYS) that digital messes up the sound, and no arguments can convince him that is the past (the distance past even for top level systems). He came to me because of an other build that i designed that he saw and heared (that uses a Lab Gruppen/Lake amp/dsp setup btw) at a private garden party last summer...
The question is not if dsp is better or not for this purpose. Off course it's better. The question is how to convince such a guy that his id is stupid.
And when it's prebuild, the amps go down to fast... and you end up with a box that is trash because of the dead amp. That's a very important reason for a lot of people to not go active. They sound of passive crossovers (even crappy ones) is good enough and the convenience wins mostly. Even for me, i like them when done right and the convenience is also good. I can't afford 6 MiniDSP flex's with DIRAC (i have 6 setups in my house). This guy actually can, he's a bussiness man with a well running construction company that gives him a lot of profit.
And let's be honest, tons of good speakers are build with passive crossovers, they measure good (here or with Erin) and sound good. But yeah, dsp can do a lot that passive can't, and avoid a whole leap of issues yo need to watch when designing.
He wants a garden system that is high power (woofers are double 18" reflex cabinets) but insist on all analog. Even the crossover to the woofers will be analog, active analog that is. He says (HE SAYS) that digital messes up the sound, and no arguments can convince him that is the past (the distance past even for top level systems). He came to me because of an other build that i designed that he saw and heared (that uses a Lab Gruppen/Lake amp/dsp setup btw) at a private garden party last summer...
The question is not if dsp is better or not for this purpose. Off course it's better. The question is how to convince such a guy that his id is stupid.
terryforsythe
Major Contributor
You could try educating him, but that may not be successful, depending on how much he already is brainwashed.
Since the system is going to be in a garden, it seems unlikely that it will be used for critical listening. If he insists on staying analog, passive filters probably will be sufficient if you use heavy gauge air core inductors in the crossovers. Also, metalized film capacitors tend to last longer than electrolytics, so I probably would stick with those as well.
Chris A
Active Member
So it's too hard and too expensive to have better and more stable sound quality even in music transients?
(That was a rhetorical question, if you didn't catch it.)
I've always noted that the audio world seems to be driven by cost...above all other considerations. Structural engineers and mission critical equipment and software designers have a totally different gestalt, it seems. (I spent a good portion of my engineering career in other similar domains.) The phrase "at that cost", is ever present in all conversations on audio (except those individuals temporarily taken in by "high end" discussions, who seemingly have lost all perspective about what they're trying to achieve).
When I see buyers who buy their "the right price" loudspeakers, and then later whine about the performance that they're getting, I have to suppress a grin.
It's actually a lot less expensive to do it better the first time around.
Chris
Last edited:
OP
Waxx
Major Contributor
- Thread Starter
- #27
That i could explain to him (thanks also for the cap test of Amir to steer him away from expensive snake oil caps). He will get Dayton PMPC film caps and Dayton DPR10 non inductive resistors (those are 1% tollerance) and big custom made (precision) aircoils of 15AWG OFC. But even then, those aircoils costs about 400€ a side, and other parts still add about 100€, that is a +1K crossover... You can buy a lot of dsp for that.
terryforsythe
Major Contributor
I fully agree.
Inductors have become very expensive. I use to wind my own, even when inductors were a fraction of the price they are currently.
terryforsythe
Major Contributor
FYI, the use of non-inductive vs sand-cast resistors in the woofer filter definitely will not be audible - you already have moderately high series inductance in the circuit, and the amount of inductance added by the resistors will be negligible. You can model this in your circuit if you know the inductance of the resistors.
Also, I doubt the difference will be audible on the tweeter considering the manner in which the speakers likely will be used. The difference will be extremely subtle. To hear the difference one probably would need to have very good hearing (i.e., young) and listen to the speakers in a controlled environment. Even then, I'm not confident the vast majority of people who do have very good hearing would detect the difference in a proper A/B comparison.
I might try saying that since the complexity and cost of the passive is high, I'd like to try to be as sure as possible the passive sims xover points, etc, are going to measure well.
And that I want to replicate the passive sim via active/DSP before building it..
Although, I don't know how I'd actually go about truly replicating a passive sim.
Real passives are easy enough, just grab an electrical transfer function, ..........but don't know where to start against a sim.
Anyway, I'd set up the rig in the garden with DSP. See how he likes it, which hopefully he does,
.....and then say that's what we HOPE to replicate via passive. Still want to do it ?
FWIW, when I'm doing my most critical listening tests, it's outdoors on a calm day. (wind is worse than noise)
If you finally end up having both active and passive to compare, that's where I'd do it.
Keith_W
Major Contributor
A long time ago, an older and wiser doctor told me that "you can prescribe the drug, but you can't make the patient take the medication". I put it more bluntly, it is not my job to save people from their own stupidity. I give them advice, they can choose whether to take it.
Waxx, you can also sell him the end all amp. Usually called a "big assed amp". He needs that to go with his analog filters. I'm thinking 500 to 800 watts per channel? Plus you can take the name off the amp and say you custom built the amp just for him! I see money to be made.........
I wonder how many active speakers are "pure" DSP though? Say the DSP can only handle 4 IIR filters, then they might decide to use a passive crossover to get it in the ballpark, then 1 to improve the crossover, and then 3 to improve the response. In that scenario you don't get the benefit of "maximum control", but how does the consumer know?
Chris A
Active Member
I don't use passive crossovers in my 5.1 setup or on other temporary setups that I've used.
I've found that the fast limiters in the DSP crossovers do a much better job at protecting small tweeters with very limited power handling. And I don't use amplifiers that have a tendency to fail often, especially with DC on the outputs (in other words, I don't use tube amplification).
All the DSP crossovers that I use have at least 8 output channel biquads and the same number of input channel biquads (upstream of the output channels). I can add more upstream if I use a PC or Raspberry Pi instead of a DSP hardware box. I'm pretty sure that I won't use passive crossovers in the future. Presently I never plan on using passive filter components at the loudspeaker.
Chris
There are legitimate technical reasons to use passive filters in some cases. See this Purifi application note for an example. Note that the effect shown can be much greater for drivers that have simpler, higher distortion motors. It would be possible to achieve a similar result using a purpose-built power amplifier with high output impedance in the right frequency range, but a conventional voltage source amp plus a few passive components is generally a much simpler solution.
Low driving impedance (high damping factor) does not offer control except at low frequencies. Higher than an octave or so above the system resonance, most woofers produce significantly less odd-order distortion when current driven (i.e. near infinite driving impedance).
Low driving impedance (high damping factor) does not offer control except at low frequencies. Higher than an octave or so above the system resonance, most woofers produce significantly less odd-order distortion when current driven (i.e. near infinite driving impedance).
My current (DIY) setup is. However, I'm actually planning to try adding some passive parts soon for two reasons: 1) to make the setup less demanding regarding DAC/amp noise (the sensitivity of the compression driver + waveguide is ~110dB/2.83V@1m), and 2) to see if I can reduce further reduce distortion in the midrange.
Chris A
Active Member
The only time that I would consider using passive components between the amplifier output terminals and the individual driver(s) is for those applications where current drive (transconductance) amplifiers cannot be found at a reasonable price, and the user wants to experiment with current drive using a typical voltage control amplifier. This operation using in-series resistors has definite upsides and downsides. I would only do this for extremely high sensitivity/low power applications (like bmc0 just mentioned) using very high sensitivity compression drivers/horns, where all you have access to are amplifiers with audible noise floors.
Once such application that I (unfortunately) found was Hypex FusionAmps (FA122) that apparently had small-signal thresholding problems (Hypex never acknowledged the problem). In that particular case, however, I simply sold the brand new FA122s to someone else that wanted to try them out on an effectively much lower sensitivity passive crossover setup using very large and very expensive passive networks (mounted on a 1m x 1m flat board beside each loudspeaker on the wall as wall art).
Personally, I'd find better amplifiers. I recommend First Watt or other Nelson Pass single-stage FET designs instead of adding passive components at the driver to put a band aid (a.k.a., "plaster") on use of poorly performing multistage amplifiers that cannot perform cleanly at extremely low voltage levels without experiencing audible noise floor issues (i.e., a bad choice for the task).
The entire concept of using passive components at the amplifier output terminals to see full power cycles in monoamping mode, especially for lower efficiency direct radiating drivers where the passive components may be experiencing 100s of watts of amplifier output power, is historic in balancing network practice, but one that clearly arose during a time when amplifiers and amplifier power were at a premium economically (1930s-1950s), and very high efficiency loudspeakers were employed to minimize amplifier costs (commercial theaters).
Those days are gone, however, and the entire concept of using passive networks that see full amplifier output power is now a concept that seems quite odd...a poorer choice nowadays. Much better to move around to the input terminals of the amplifiers for crossover operations and use one amplifier channel per acoustic driver, even for current drive applications. There are several DIY transconductance amplifier designs that can be used instead of heating resistors in series with acoustic drivers.
Chris
Once such application that I (unfortunately) found was Hypex FusionAmps (FA122) that apparently had small-signal thresholding problems (Hypex never acknowledged the problem). In that particular case, however, I simply sold the brand new FA122s to someone else that wanted to try them out on an effectively much lower sensitivity passive crossover setup using very large and very expensive passive networks (mounted on a 1m x 1m flat board beside each loudspeaker on the wall as wall art).
Personally, I'd find better amplifiers. I recommend First Watt or other Nelson Pass single-stage FET designs instead of adding passive components at the driver to put a band aid (a.k.a., "plaster") on use of poorly performing multistage amplifiers that cannot perform cleanly at extremely low voltage levels without experiencing audible noise floor issues (i.e., a bad choice for the task).
The entire concept of using passive components at the amplifier output terminals to see full power cycles in monoamping mode, especially for lower efficiency direct radiating drivers where the passive components may be experiencing 100s of watts of amplifier output power, is historic in balancing network practice, but one that clearly arose during a time when amplifiers and amplifier power were at a premium economically (1930s-1950s), and very high efficiency loudspeakers were employed to minimize amplifier costs (commercial theaters).
Those days are gone, however, and the entire concept of using passive networks that see full amplifier output power is now a concept that seems quite odd...a poorer choice nowadays. Much better to move around to the input terminals of the amplifiers for crossover operations and use one amplifier channel per acoustic driver, even for current drive applications. There are several DIY transconductance amplifier designs that can be used instead of heating resistors in series with acoustic drivers.
Chris
manisandher
Addicted to Fun and Learning
I can certainly attest to this, having converted 3 passive speaker systems to active. But the improvement in SQ in each case has been more than worth it.
Mani.
I envision a day when each loudspeaker has only one cable attached to it -- a power cable to charge the batteries.
An all-digital signal is sent wirelessly to each loudspeaker, which is processed by an onboard DSP and sent to individual battery-powered Class-D amplifiers, one per driver.
An all-digital signal is sent wirelessly to each loudspeaker, which is processed by an onboard DSP and sent to individual battery-powered Class-D amplifiers, one per driver.
That would be awfully convenient and practical indeed, even without battery (what's that good for, except portability?).
One immediate problem I see with current technology is latency. Of course for pure listening, even the usually horrible blutoof latency isn't a real concern - but for any timing-critical realtime application such as gaming and music production, even a WiFi latency of, say, 15ms can be too high. It just sucks. Even 25 years ago, that was already a solved problem with wired PCI audio - stable performance at 1ms latency was state of the art.
I don't see that performance happening in wireless form anytime soon. Perhaps I'm a bit too nitpicky about it, but believe me - the human mind and ear has a much higher time resolution than it does in anything else. The difference between 1 and 10ms is already very noticeable.
