I am in the planning stages for a 5.1 setup for my computer. Primary focus is on sound quality for music. I still haven't decided on a speaker design yet but will probably be of the 4.5" + tweeter bookshelf variety. Just planning on starting with the front L&R + sub (two 10" since I have them) for now.
Anyway, in looking through all the speaker designs the crossovers seem to be very complex. I'm OK with making a separate amp for the tweets and 4.5"ers but I'm wondering if that same level of complexity of xover can be carried over into doing it actively. Would that be an advantage or disadvantage?
Anyway, in looking through all the speaker designs the crossovers seem to be very complex. I'm OK with making a separate amp for the tweets and 4.5"ers but I'm wondering if that same level of complexity of xover can be carried over into doing it actively. Would that be an advantage or disadvantage?
Ryan,
There may be a third method worth considering if you have not done so--that of using software based digital XO's in conjunction with your soundcard. See here for discussion: http://www.diyaudio.com/forums/showthread/t-6356.html
First you would not have to build XO units, and much faster to find right frequency and levels. Also much better in some ways sonically as very steep, less phase distortion.
If this were expensive HT project, I would probably not mess around with computer method as it may not have enough flexibility, but for a small PC based system, this seems like a great place to start, and even free, assuming you will be building or buying amps anyway.
There may be a third method worth considering if you have not done so--that of using software based digital XO's in conjunction with your soundcard. See here for discussion: http://www.diyaudio.com/forums/showthread/t-6356.html
First you would not have to build XO units, and much faster to find right frequency and levels. Also much better in some ways sonically as very steep, less phase distortion.
If this were expensive HT project, I would probably not mess around with computer method as it may not have enough flexibility, but for a small PC based system, this seems like a great place to start, and even free, assuming you will be building or buying amps anyway.
Since you already have the schematics to building the passive crossovers, it should be much easier than you think. I can put one together in about 10 minutes.
If you insist on avoiding building the passive crossover, the transfer function is usually published and you can use that to program a digital crossover. I strongly beleive that passive is more transparent than digital crossovers. This does not apply to computer based digital crossovers though as you avoid the additional conversion stages.
The other option is analog active crossover, which you would have to build yourself as well. Since you are unwilling to put together a passive crossover, then I assume you would be unwilling to build an active crossover also.
If you insist on avoiding building the passive crossover, the transfer function is usually published and you can use that to program a digital crossover. I strongly beleive that passive is more transparent than digital crossovers. This does not apply to computer based digital crossovers though as you avoid the additional conversion stages.
The other option is analog active crossover, which you would have to build yourself as well. Since you are unwilling to put together a passive crossover, then I assume you would be unwilling to build an active crossover also.
@Dave,
Are you suggesting just using a full range 4.5" on each side? That would make supper simpler. I've read a bit about those and I know you guys know your stuff (certainly more than me) but there's that mental hurdle of no tweeter. Got a driver in mind? ....I bet you're thinking FR125S....
@Newmexiconoob,
That's not really what I was looking for but it looks interesting and I'll give it a read! Thanks.
Are you suggesting just using a full range 4.5" on each side? That would make supper simpler. I've read a bit about those and I know you guys know your stuff (certainly more than me) but there's that mental hurdle of no tweeter. Got a driver in mind? ....I bet you're thinking FR125S....
@Newmexiconoob,
That's not really what I was looking for but it looks interesting and I'll give it a read! Thanks.
Sure, no problem. There are those of us who are convinced that digital XO's are even better than none at all, and certainly superior to analog (whether high or low lovel signals). 
Certainly worth reading about--discovered a much more inclusive thread which is here:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=63078
Best of luck whichever road you take,
John
Certainly worth reading about--discovered a much more inclusive thread which is here:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=63078
Best of luck whichever road you take,
John
cotdt,
By active crossovers I meant at the line level (analog) using op-amps. I'm not unwilling to make the large active crossovers at the speakers themselves. My concerns were the cost of large high quality caps/inductors/resistors.... and then building an inventory of those to tune or play with different crossover points where I could get many of the same results in the active xovers by changing a couple/few little 1/4W resistors. A quick look into some prices using a fairly complex network for a 2-way that included higher order high/low pass plus additional notch filtering and such to smooth the frequency response, the cost of the xovers was quickly getting very expensive.
Also given effects like stray inductance, coil resistance, ESR in the caps, electrolytic caps having usually widely ranging tolerances, large resistors having an inductive content, most of the above being temperature dependent which I would assume can vary quite a bit given how hard the speakers are driven... blah blah blah.... I thought actives might be better for sound quality, not just easier on the wallet. I'm not suggesting active xovers don't have their own set of issues but I think they can be much easier minimized. At least I think, but that's why I'm asking
Thanks,
Ryan
By active crossovers I meant at the line level (analog) using op-amps. I'm not unwilling to make the large active crossovers at the speakers themselves. My concerns were the cost of large high quality caps/inductors/resistors.... and then building an inventory of those to tune or play with different crossover points where I could get many of the same results in the active xovers by changing a couple/few little 1/4W resistors. A quick look into some prices using a fairly complex network for a 2-way that included higher order high/low pass plus additional notch filtering and such to smooth the frequency response, the cost of the xovers was quickly getting very expensive.
Also given effects like stray inductance, coil resistance, ESR in the caps, electrolytic caps having usually widely ranging tolerances, large resistors having an inductive content, most of the above being temperature dependent which I would assume can vary quite a bit given how hard the speakers are driven... blah blah blah.... I thought actives might be better for sound quality, not just easier on the wallet. I'm not suggesting active xovers don't have their own set of issues but I think they can be much easier minimized. At least I think, but that's why I'm asking
Thanks,
Ryan
Ryan_Mc,
I see you are refering to analog active crossovers, and in that case assume you know how to design them. Both analog active crossovers and passive crossovers are equally transparent if decent parts are used. Active is even bigger than the passive if you include the power supply for the op-amps and additional amplifiers. For a stereo 3-way, you will need 6 channels of amplication. For me, buying a good aluminum case for my active crossover already costed more than the entire passive crossover, which are usually housed inside the speaker.
There is no difference in sound quality between analog and passive. In listening tests the passives do beat out the digital crossovers though. Most of the arguments against passives assume that the passives are generic designs not tailered to the actual drivers used, which is not a fair comparison for the passive. The extra things that active can do are not that useful IMHO. Active is indeed more efficient but you're using 6 amplifiers instead of 2 so they better be!
My recommendation, go passive for 2-way, and a combination of passive and active for 3-way to save money. If using a ribbon tweeter, going passive is a must. There needs to be a capacitor to protect the ribbon.
I see you are refering to analog active crossovers, and in that case assume you know how to design them. Both analog active crossovers and passive crossovers are equally transparent if decent parts are used. Active is even bigger than the passive if you include the power supply for the op-amps and additional amplifiers. For a stereo 3-way, you will need 6 channels of amplication. For me, buying a good aluminum case for my active crossover already costed more than the entire passive crossover, which are usually housed inside the speaker.
There is no difference in sound quality between analog and passive. In listening tests the passives do beat out the digital crossovers though. Most of the arguments against passives assume that the passives are generic designs not tailered to the actual drivers used, which is not a fair comparison for the passive. The extra things that active can do are not that useful IMHO. Active is indeed more efficient but you're using 6 amplifiers instead of 2 so they better be!
My recommendation, go passive for 2-way, and a combination of passive and active for 3-way to save money. If using a ribbon tweeter, going passive is a must. There needs to be a capacitor to protect the ribbon.
Ryan_Mc said:
I had a good laugh when I saw where that link lead to. Zaph's ZD5 uses a relatively simple crossover with few parts. I trust that you've not seen the ZDT3.5's at the same website (which is itself relatively simple design for a 3-way)?
Active crossovers have even more parts and take a lot longer to make. Building an active crossover is basically like building a complete amplifier, with power supply, casing, input/output jacks, etc.
I think that the important thing to remember is that no matter how you implment your XO, active or passive, to get the same sound, the XO transfer functions have to match. So for example, using the Zaph design pointed to earlier, you would have to match these transfer functions in your active XO plus imlement active delay to get the same frequency response that Zaph got.
http://www.zaphaudio.com/ZD5-modeled-transferfunction.gif
You will note, those are not textbook slopes.
Active and passive XOs each have their own set of benefits and drawbacks. It is up to the designer to decide which provides the most bang for the buck for each design.
Good luck
Dennis
http://www.zaphaudio.com/ZD5-modeled-transferfunction.gif
You will note, those are not textbook slopes.
Active and passive XOs each have their own set of benefits and drawbacks. It is up to the designer to decide which provides the most bang for the buck for each design.
Good luck
Dennis
I agree getting the same transfer function could be a bugger doing it all actively. Op-amp filters can perform all the functions of caps/resistors/inductors so yes it can be done but is it worth it. Usually you can kill two or more birds with one stone, for instance a notch filter can be done with one op-amp and high and low frequency crossover points as well as Q and gain can all be adjusted on the fly by pots (I believe that particular function is called a parametric EQ).
I'm not trying to push the actives BTW, eventhough it might sound like it. I like cotdt's idea (and it seems yours) to mix and match passive and active based on the particular situation.
It's looking like this build will require minimal xovers anyway as the more I think about the full range 4.5" the more I like it. I still have some time to decide until I build my amps though.
Thanks for all the info,
Ryan
I'm not trying to push the actives BTW, eventhough it might sound like it. I like cotdt's idea (and it seems yours) to mix and match passive and active based on the particular situation.
It's looking like this build will require minimal xovers anyway as the more I think about the full range 4.5" the more I like it. I still have some time to decide until I build my amps though.
Thanks for all the info,
Ryan
djarchow said:
I'd love to know if this is possible (getting the same sound). The passive XO interacts with the driver impedance, which actually changes a fair amount depending on the excursion and voice coil temperature. It would be interesting to find out how much of this variation is audible. With an active crossover, these driver variations are isolated by the amplifier, so changes in impedance with cone movement and voice coil heating shouldn't affect the output.
Some of the curves in the Klippel documentation show a substantial change in driver impedance with cone position and heating. I think it would be fun to set up a good A/B test that compared active and passive crossovers at different listening levels to see if the impedance change and the interaction with the other XO components is noticeable...and what does it sound like..?
At low crossover points under extreme temperature variations, passive crossover points would vary a bit. This is a moot point though, since the woofer's T/S parameters would change so much that performance would be degraded no matter whether you use active or passive.
In my experience, with decent parts passive and active sound the same. It's all about which is more convenient. Active crossovers are affected by choice of opamp and quality of its PSU while passive crossovers are affected by choice of capacitor, though neither makes that big of a difference. It's unclear whether the choice of capacitors used in active crossovers makes a sonic difference, I've always used good ones so I wouldn't know. Choice of inductor in passive crossovers make zero sonic difference.
Active crossovers are useful in dipoles and adding things like Linkwitz Transform. Those are not generally useful in 2-ways, so you probably won't be needing active.
In my experience, with decent parts passive and active sound the same. It's all about which is more convenient. Active crossovers are affected by choice of opamp and quality of its PSU while passive crossovers are affected by choice of capacitor, though neither makes that big of a difference. It's unclear whether the choice of capacitors used in active crossovers makes a sonic difference, I've always used good ones so I wouldn't know. Choice of inductor in passive crossovers make zero sonic difference.
Active crossovers are useful in dipoles and adding things like Linkwitz Transform. Those are not generally useful in 2-ways, so you probably won't be needing active.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.