George have you worked out the transfer function of the original passive crossover?
That should really be your starting point when contemplating active operation.
Could be that the drivers are well behaved for an octave beyond their operating window in which case things would be easy but if there is some eq and/or notch filters in the passive one things are more complex.
In the first, rare, case you can get good results with a generic active crossover, in the second I'd try DSP for convenience, speed and flexibility.
I say that running my heavily eq'd Tannoy pepperpots analogue active.
That should really be your starting point when contemplating active operation.
Could be that the drivers are well behaved for an octave beyond their operating window in which case things would be easy but if there is some eq and/or notch filters in the passive one things are more complex.
In the first, rare, case you can get good results with a generic active crossover, in the second I'd try DSP for convenience, speed and flexibility.
I say that running my heavily eq'd Tannoy pepperpots analogue active.
Last edited:
Most of the people think about "active filters" and their counter part passive ones, but by going active you have several tools to manipulate speakers FR that is imposible to do in the "passive world". For example, you have the Linkwitz Transform to move the driver's zeros and poles up and down in the s plane, and change the drivers' Q and fs to match your own needs.
You can also use all-pass filters to time-align the in-box drivers' response in order to avoid the loudspeaker lobe wander in your room.
Well... you must go active if you want maximum quality and performance from your loudspeakers... but you will also need a lot of knowledge to achieve it...
You can also use all-pass filters to time-align the in-box drivers' response in order to avoid the loudspeaker lobe wander in your room.
Well... you must go active if you want maximum quality and performance from your loudspeakers... but you will also need a lot of knowledge to achieve it...
Last edited:
I will try better next time sir!
At first glance both the original and Troels upgrade appear fairly simple 2nd and 3rd order networks with nothing much in terms of eq only level controls which are not needed when active. The drivers must be fairly well behaved all in all.
I suppose you'd get quite similar results with a generic 4th order L-R using the appropriate xover points although you'd have to un-invert the tweeter.
Also the woofer appears to be losing 4dB through the original passive (88dB/1W/1m with xover vs 92dB without) which you would get back going active.
At first glance both the original and Troels upgrade appear fairly simple 2nd and 3rd order networks with nothing much in terms of eq only level controls which are not needed when active. The drivers must be fairly well behaved all in all.
I suppose you'd get quite similar results with a generic 4th order L-R using the appropriate xover points although you'd have to un-invert the tweeter.
Also the woofer appears to be losing 4dB through the original passive (88dB/1W/1m with xover vs 92dB without) which you would get back going active.
What started me on this question was a project being started by a local Loudspeaker suppler... see How to Build DIY Speakers and Use Component Parts
Basically I was considering whether I could get similar or better results by using my Yamaha speaker units in the dipole designs that he describes. It turns out that he built his own active crossover circuit based on Linkwitz-Riley and spent a lot of time in measuring and tuning. I have still not given up on the active idea and I like the plan of updating one pair of Yamahas with optimised passive crossovers and the other pair with an active system. This of course is subject to WAF...and I already failed on that with my Klipsches, my JBL 300, my JBL 4343, my Stacked ESL .......
Basically I was considering whether I could get similar or better results by using my Yamaha speaker units in the dipole designs that he describes. It turns out that he built his own active crossover circuit based on Linkwitz-Riley and spent a lot of time in measuring and tuning. I have still not given up on the active idea and I like the plan of updating one pair of Yamahas with optimised passive crossovers and the other pair with an active system. This of course is subject to WAF...and I already failed on that with my Klipsches, my JBL 300, my JBL 4343, my Stacked ESL .......
Other than some slightly boutique component choices I cannot see anything wrong with Troels' analysis and the 1000's look to be pretty good for 40 year old speakers so certainly worth experimenting with.
You sure that's not BSC? sounds about right OOM, and one area where its def worth doing the eq outside the speaker as a shelving function is easy at line level.
I would say 'can' be simple. It can also be complex. You might be well served picking up a second hand DCX2496 to experiment with as they hold their value well and have huge tweakability to deal with audiophilia nervosa. DSP is unmatched for speed of trying things even if you end up analogue or passive.
This guy Connectors, Passive items in KMTech Design store on eBay! sells pcb's, kits and fully assembled ones in the UK.
Never bought anything from him so can't say anything about the quality of his design but he does have pretty decent feedback.
Either way it might be quicker and cheaper than ordering from Oz.
Never bought anything from him so can't say anything about the quality of his design but he does have pretty decent feedback.
Either way it might be quicker and cheaper than ordering from Oz.
There does not seem to be any BSC implemented in xover, it is just a simple 2nd order with the normal inductor DC resistance.
My Tannoys woofer part lost 3dB through the passive crossover which also was just the simplest of 2nd order jobs with a series cored inductor and parallel cap.
+1. Second hand P.A. market you'll find plenty of them. I wouldn't stay long term with them but for tweak and first approach of principle, nothing come close imho (quality/price/tweakability).
By the way i've got a full page from P. Newell in 'recording studio acoustic design' about advantage of multi amped active filtered solutions.
If you want i could list them and report back. In proaudio world (P.A. and studios) this for a long time the 'workhorse' principle. This is without saying that marvellous loudspeakers is absolutely possible using the passive way (in fine the finest i've heard to date ARE passive) but for demanding applications AND not genius 'tweakers' this is convenient and you can achieve quite close results to those finest speakers.
Last edited:
Personally I used to be a huge active crossover fanboy, before I ever built any. Now not so sure.
The areas you gain most in with active crossovers are related to power. First, you no longer need power resistors to pad drivers to match, like you do on most tweeters. Less V = less power = less waste. This is something you deal with in the active stages.
With careful matching, you can also get more dynamic range with smaller amps than with a large amp.
Power is a function of the square of the voltage. P = (V*V)/R so for instance, in pure theory, a 100 Watt amplifiers can produce 28Vrms each into 8 Ohms. To produce say, 50 watts into the tweeter and 100 Watts into the woofer you would need 48 Vrms. If you do this with a single amp this needs a 290 Watt amplifier, but with active crossovers you would get the same SPL with a 100 Watt + 50 Watt amplifier.
Of course, it's always more complicated than that. How much power you ever need in a tweeter due to the musical spectrum changes the results and therefore the overall value.
Old school pro crossovers only did simple slope and knee selection and would rely on either having speakers that didn't need EQ, or expected an EQ elsewhere. To truly create an active crossover that mimics the passive version you need to take this into account. If the passive design has a notch filter for the mid-woofer, you'll need one in the active section, or you'll need to keep a passive version. This is why digital solutions like the miniDSP are so attractive.
In addition to the padding resistors another thing you don't have to concern yourself with in an active stage are any impedance matching networks such as a Zobel. No longer necessary.
Another benefit if you are really really tweaky is being able to choose different amps. For instance, a solid state or digital for the woofer and some small, sweet sounding amp for the tweeter. For instance, I could see the F6 kit at 25 watts being a fabulous choice for the upper frequencies, or a number of tube amps.
The major disadvantage is noise.
You might want to consider the Hypex AS2.100 or similar plate amp with built in DSP. It will work as a 2 way active crossover AND do a number of EQ functions and delays. When you add up the amplifier cost, crossover components, etc. it's a really good deal.
The digital solutions however may really impact the resolution. The most cost effective one's rely on ADC. At low volumes, the resolution just isn't there. However, using a digital crossover that takes digital in and spits out analog out or digital out eliminates this issue, but can get really expensive. Imagine a 2 way crossover with digital I/O. It would require 2 stereo DAC's if you wanted to pick your own. Ouch!
Having said all of this, I've still never built an active stage except on a bench.
Best,
Erik
The areas you gain most in with active crossovers are related to power. First, you no longer need power resistors to pad drivers to match, like you do on most tweeters. Less V = less power = less waste. This is something you deal with in the active stages.
With careful matching, you can also get more dynamic range with smaller amps than with a large amp.
Power is a function of the square of the voltage. P = (V*V)/R so for instance, in pure theory, a 100 Watt amplifiers can produce 28Vrms each into 8 Ohms. To produce say, 50 watts into the tweeter and 100 Watts into the woofer you would need 48 Vrms. If you do this with a single amp this needs a 290 Watt amplifier, but with active crossovers you would get the same SPL with a 100 Watt + 50 Watt amplifier.
Of course, it's always more complicated than that. How much power you ever need in a tweeter due to the musical spectrum changes the results and therefore the overall value.
Old school pro crossovers only did simple slope and knee selection and would rely on either having speakers that didn't need EQ, or expected an EQ elsewhere. To truly create an active crossover that mimics the passive version you need to take this into account. If the passive design has a notch filter for the mid-woofer, you'll need one in the active section, or you'll need to keep a passive version. This is why digital solutions like the miniDSP are so attractive.
In addition to the padding resistors another thing you don't have to concern yourself with in an active stage are any impedance matching networks such as a Zobel. No longer necessary.
Another benefit if you are really really tweaky is being able to choose different amps. For instance, a solid state or digital for the woofer and some small, sweet sounding amp for the tweeter. For instance, I could see the F6 kit at 25 watts being a fabulous choice for the upper frequencies, or a number of tube amps.
The major disadvantage is noise.
You might want to consider the Hypex AS2.100 or similar plate amp with built in DSP. It will work as a 2 way active crossover AND do a number of EQ functions and delays. When you add up the amplifier cost, crossover components, etc. it's a really good deal.
The digital solutions however may really impact the resolution. The most cost effective one's rely on ADC. At low volumes, the resolution just isn't there. However, using a digital crossover that takes digital in and spits out analog out or digital out eliminates this issue, but can get really expensive. Imagine a 2 way crossover with digital I/O. It would require 2 stereo DAC's if you wanted to pick your own. Ouch!
Having said all of this, I've still never built an active stage except on a bench.
Best,
Erik
Last edited:
So how will that work for a 3-way speaker such as George is looking at?
Why does nobody read threads before posting?
Erik, i suppose the noise related issue is related to entry level ADC/DAC?
I use and have used some pro dsp hardware available on the market in EU. As soon as you get a little serious (the middle range in either Dbx,Xilica,Dolby,...), ADC and DAC are goods to very goods on this pro units! You can have the possibility to sync to wordclock, multiple inputs and outputs, real time operation, sometimes a practical gui for relatively cheap, and on some units dsp and algorythms are quite good!
If you can't afford a pro hardware unit, there's some lovely solution using Pc: there's a relatively wide offer of software (from the mastering grade to more affordable-even ASIO plug ins) and small (read relatively cheap) but good quality multiple inputs/outputs soundcards are widely availlable second hand.
If you use only digital source you can have everything in a small package (player and treatments), it can even be noiseless.
And do other things too.
Last edited:
As long as you can do a gain calculation and are not running some pathologically sensitive speakers with huge high gain amplifiers noise should not be an issue. It just requires you to look at your levels and think a bit. If you have amps sized to cause deafness at 2V RMS in and listen with 40dB digital attenuation all the time you might need to rethink.
Or get this DCX2496 active output mod & 6-channel vol control | Linear Audio NL . It's not cheap as a kit but moves volume control back in the analog domain to give a warm and fuzzy to those afraid of digital volume controls.
You have choices, choice is good
Or get this DCX2496 active output mod & 6-channel vol control | Linear Audio NL . It's not cheap as a kit but moves volume control back in the analog domain to give a warm and fuzzy to those afraid of digital volume controls.
You have choices, choice is good
You are right billshurv! But... for the price of a fullkit moded dcx you can now score a used Dbx 4800 ( you'll need to search i agree, it'll probably not be in a mint cosmetic condition (ah... those PA engineer!
) but it's feasible)! Between the two no need to question yourself! For noise i agree with Erik about entry level DAC/ADC: my experience with them is that they're still not to the level of pro dedicated gear.
Last edited:
The dbx is a bit too lit up for me. Confession time, I have a miniDSP 4x10 AND a linkwitz ASP but whichever I end up using will have a a multichannel volume control. I need it for level matching (miniDSP is +14dBu out on balanced) but also removes any residual audiophilia nervosa. But all because I can rather than because I need to.
Don't worry, once dialed up you have the option to tame it down to full dark. This is pro gear not a 'car tuning show'
Confession time too: i ended using an analog pad with 2 settings using constant Z input
for the same reason (1 setting for day -20db, the other for night -35db) for each output. But i still use the digital trim at the input of the filter for fine triming... Not ideal but i'll have to wait for the amb delta1 solution.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.