This is my first post, so...
I'm just starting building speakers myself, but enjoying it very much. I have built some 4 different loudspeakers (not my own design), and i read a lot about it. I am beginning to understand the basics of designing speakers.
I was wondering if it would be possible to build a two-way box with very good sound and roll-off-waterfall etc. with a very simple 6dB/oct XO?
And a reasonably linear frequency-response between say, 100Hz - 12kHz, using just one capacitor and one inductor?
Or does anybody know of an existing design? I like the idea very much of using a very simple XO, because it seems closest to an active crossover, without the need for biamping.
I'm just starting building speakers myself, but enjoying it very much. I have built some 4 different loudspeakers (not my own design), and i read a lot about it. I am beginning to understand the basics of designing speakers.
I was wondering if it would be possible to build a two-way box with very good sound and roll-off-waterfall etc. with a very simple 6dB/oct XO?
And a reasonably linear frequency-response between say, 100Hz - 12kHz, using just one capacitor and one inductor?
Or does anybody know of an existing design? I like the idea very much of using a very simple XO, because it seems closest to an active crossover, without the need for biamping.
I guess one would need high quality drivers. But I was just sort of dreaming about the idea. It will take me months to figure it all out. Understand me right, i don't expect people to 'talk me through' or anything.
Most designs i come across, for example from visaton, or int'l, use many many components in the xo. While to me the idea of as little as possible seems so much better.
Is it possible to find a matching partner for the Neo3 PDR tweeter?
Neo3 at PartsExpress
The specs seem to me to be very good, the lowest XO-point is at 2kHZ.
Most designs i come across, for example from visaton, or int'l, use many many components in the xo. While to me the idea of as little as possible seems so much better.
Is it possible to find a matching partner for the Neo3 PDR tweeter?
Neo3 at PartsExpress
The specs seem to me to be very good, the lowest XO-point is at 2kHZ.
Yes sure!
If you choose carefully your drivers, you can be surprised.
Drivers need a better flat impedance. Also, choose your crossover point far from the resonance (at least 3 octaves).
I don't think you should consider a 1st order passive crossover to be closer to an active crossover than any other type of passive crossover. This is very different.
If you choose carefully your drivers, you can be surprised.
Drivers need a better flat impedance. Also, choose your crossover point far from the resonance (at least 3 octaves).
I don't think you should consider a 1st order passive crossover to be closer to an active crossover than any other type of passive crossover. This is very different.
Robin Marshall of Epos knew all about it
Hello,
Read this old review and take notice of the thorough graphs.
There is also a splendid interview with Robin Marshall.
http://www.stereophile.com/showarchives.cgi?566
and the Saga continues...
http://www.onhifi.com/product/epos_m15.htm
Hello,
Read this old review and take notice of the thorough graphs.
There is also a splendid interview with Robin Marshall.
http://www.stereophile.com/showarchives.cgi?566
and the Saga continues...
http://www.onhifi.com/product/epos_m15.htm
X-OVER
Hi guys,
Max Townsend of the famous Rock Elite TT once presented a two-way speaker made out of a black piano lacquer finished plaster mixture (sort of a plaster of Paris thing) it use a Lowther PM6 and a hornless Kelly ribbon (I think) crossed at ~10K with a simple cap.
Really impressive demo.
I believe in simple x-overs too,the less components the better your chance for good dynamics,stereo image and microdetail.
Bear in mind that some amps will still like to see a Zobel network on the woofers to even out impedance irregularities.
Check this out too:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=6060
Cheers,
Hi guys,
Max Townsend of the famous Rock Elite TT once presented a two-way speaker made out of a black piano lacquer finished plaster mixture (sort of a plaster of Paris thing) it use a Lowther PM6 and a hornless Kelly ribbon (I think) crossed at ~10K with a simple cap.
Really impressive demo.
I believe in simple x-overs too,the less components the better your chance for good dynamics,stereo image and microdetail.
Bear in mind that some amps will still like to see a Zobel network on the woofers to even out impedance irregularities.
Check this out too:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=6060
Cheers,
massive,
For 8" 2 way, the woofer directly connected, and a single series tweeter capacitor works great provided you do two things.
That is put a zobel network across each driver and make the impedence of each driver measure flat out to 40 kHz or so.
This will require bypassing of larger capacitors with smaller values down to ceramic disc capacitors.
The series tweeter cap should be around 1.1111 uF or so.
This will give a resistive overall load that any amplifier is perfectly happy with, and the Zobel networks will cause the bandwidth to extend low and up past hearing limits.
One product of this arrangement is that the phase response is about as flat as you are going to get, and therefore is utterly revealing of source music AP.
Another product is much better power handling and transient response.
I had a pair of 8" 2 ways, fed perfectly clean 150W+150W that resonated the block of 6 apartments I lived in on the right bass lines.
These speakers were rated 40W rms and 80W music and I had no failures despite cranking the hell out of them.
One proviso of running at this power level is that all windows and external and internal doors need to be closed to avoid unloading/overdriving.
Eric.
For 8" 2 way, the woofer directly connected, and a single series tweeter capacitor works great provided you do two things.
That is put a zobel network across each driver and make the impedence of each driver measure flat out to 40 kHz or so.
This will require bypassing of larger capacitors with smaller values down to ceramic disc capacitors.
The series tweeter cap should be around 1.1111 uF or so.
This will give a resistive overall load that any amplifier is perfectly happy with, and the Zobel networks will cause the bandwidth to extend low and up past hearing limits.
One product of this arrangement is that the phase response is about as flat as you are going to get, and therefore is utterly revealing of source music AP.
Another product is much better power handling and transient response.
I had a pair of 8" 2 ways, fed perfectly clean 150W+150W that resonated the block of 6 apartments I lived in on the right bass lines.
These speakers were rated 40W rms and 80W music and I had no failures despite cranking the hell out of them.
One proviso of running at this power level is that all windows and external and internal doors need to be closed to avoid unloading/overdriving.
Eric.
More & more both commercial speakers & DIY are returning to very simple XOs.
The series XO is one of the premeire examples of 1 cap 1 coil. Andy Graddon's stuff is a good example.
More common yet is the use of a full range (ie my BD-pipes) or a smooth midbass (ie the TLb) run flat out with a well behaved tweeter XOed to fill in above their natural roll-off.
The Reference 3a speakers are an example of a very well respected commercial offering.
Well behaved drivers with extended bandwidths are required to carry this off. There are more and more of them at all sorts of price points. Jordan, Tagband, Fostex, the Vifa P13 are just a small sampling...
dave
The series XO is one of the premeire examples of 1 cap 1 coil. Andy Graddon's stuff is a good example.
More common yet is the use of a full range (ie my BD-pipes) or a smooth midbass (ie the TLb) run flat out with a well behaved tweeter XOed to fill in above their natural roll-off.
The Reference 3a speakers are an example of a very well respected commercial offering.
Well behaved drivers with extended bandwidths are required to carry this off. There are more and more of them at all sorts of price points. Jordan, Tagband, Fostex, the Vifa P13 are just a small sampling...
dave
yes dave the KISS principle wins again.
i remember the 70s when manuftueres were falling over themselves to make 4 way and 5 ways systems. bose's 901 was one the defied this.
if you ever work you way down to my side of town drop in to listen to my 6" 2way which hopefully will be complete by then. you bring the salmon, i'll get the barbie or the local version of one(tandoor) and beer.
BTW if i remember right there is atleast one good indian restraunt near the steam powered clock on the mainland (vancouver). I did not say it was cheap but i think they might have a tandoor.
talking about full range drivers there are some 2-3" versions out there too.
i remember the 70s when manuftueres were falling over themselves to make 4 way and 5 ways systems. bose's 901 was one the defied this.
if you ever work you way down to my side of town drop in to listen to my 6" 2way which hopefully will be complete by then. you bring the salmon, i'll get the barbie or the local version of one(tandoor) and beer.
BTW if i remember right there is atleast one good indian restraunt near the steam powered clock on the mainland (vancouver). I did not say it was cheap but i think they might have a tandoor.
talking about full range drivers there are some 2-3" versions out there too.
1st order on a tweeter.
most tweeters have an FS in theorder or 1500Hz how much tpower can a tweeter take with a 1st order Xo even if it at 6k or 8k.
my fear is that 1st order does not limit power to tweeter enough.
more importantly 1st order does not limit low frequencies (read as high excursions) to tweters enough.
ihad a 1st order 3k XO ont a pair of Morel MDT33 and blew them with a 250W amp.
after that pardon me for eing a bit paranoid.
also I live in India so getting replacemetns is not as easy as the US.
most tweeters have an FS in theorder or 1500Hz how much tpower can a tweeter take with a 1st order Xo even if it at 6k or 8k.
my fear is that 1st order does not limit power to tweeter enough.
more importantly 1st order does not limit low frequencies (read as high excursions) to tweters enough.
ihad a 1st order 3k XO ont a pair of Morel MDT33 and blew them with a 250W amp.
after that pardon me for eing a bit paranoid.
also I live in India so getting replacemetns is not as easy as the US.
Destructive Resonances....
Navin, in my experience if you feed tweeters dead clean you can feed very high powers without burning voicecoils.
The main failure mode in tweeters is the flex leads failing, mainly due to excessive excursion at resonance, and indeed resonance of the voicecoil connecting wires.
This can be cured by fitting a parallel series resonant netwok to damp/shunt power at the tweeter mechanical resonance frequency.
Eric.
Navin, in my experience if you feed tweeters dead clean you can feed very high powers without burning voicecoils.
The main failure mode in tweeters is the flex leads failing, mainly due to excessive excursion at resonance, and indeed resonance of the voicecoil connecting wires.
This can be cured by fitting a parallel series resonant netwok to damp/shunt power at the tweeter mechanical resonance frequency.
Eric.
Low count cross-overs
The basic idea of minimising the component count is one I try to stick to where-ever I can. If you look at the general design principle of any crossover, you will see that most of the complexity comes from either wanting sharp slopes or having to correct problems with the drivers.
The main thing is that everything is a compromise. To get the very best detail through the mids, usually requires a hard compound cone, which usually leads to breakup points above the cross-over freq. hence the need for steep slopes or correction.
To minimise the x-o component count one has to select the drivers very carefully. The woofer must have a beautifully smooth roll-off, the tweeter response must be flat as and must not have any resonance peak problems or must have a very low res freq so you can avoid those problems. If you want to get rid of L-pad resistors, you also need to choose drivers where the sensitivities are basically the same (although I have found that to dispense with the resistors I generally need the tweeter to be 1-2dB LESS than the woofer in a series x-o, that might just be a listening preference though)
This generally limits low-x-o counts to uses with poly cone woofers, and some paper cones ! (carbons, kevlar, aluminium etc generally pose way to many problems, but will generally be more detailed through the mids). Any tweeter with a res fr greater than say 1000 is going to be problematic in a lo-count 2-way (except maybe a small bookshelf crossed to a 5" or 4"). Ferro tweeters are also more likely to be able to be used because of their higher power handleing and lower res peak.
So, If you want lo-count cross-overs, start researching the drivers !!!
Good Luck !!
and you other guys, there are a LOT of general statements here, please be gentle !!!
The basic idea of minimising the component count is one I try to stick to where-ever I can. If you look at the general design principle of any crossover, you will see that most of the complexity comes from either wanting sharp slopes or having to correct problems with the drivers.
The main thing is that everything is a compromise. To get the very best detail through the mids, usually requires a hard compound cone, which usually leads to breakup points above the cross-over freq. hence the need for steep slopes or correction.
To minimise the x-o component count one has to select the drivers very carefully. The woofer must have a beautifully smooth roll-off, the tweeter response must be flat as and must not have any resonance peak problems or must have a very low res freq so you can avoid those problems. If you want to get rid of L-pad resistors, you also need to choose drivers where the sensitivities are basically the same (although I have found that to dispense with the resistors I generally need the tweeter to be 1-2dB LESS than the woofer in a series x-o, that might just be a listening preference though)
This generally limits low-x-o counts to uses with poly cone woofers, and some paper cones ! (carbons, kevlar, aluminium etc generally pose way to many problems, but will generally be more detailed through the mids). Any tweeter with a res fr greater than say 1000 is going to be problematic in a lo-count 2-way (except maybe a small bookshelf crossed to a 5" or 4"). Ferro tweeters are also more likely to be able to be used because of their higher power handleing and lower res peak.
So, If you want lo-count cross-overs, start researching the drivers !!!
Good Luck !!
and you other guys, there are a LOT of general statements here, please be gentle !!!
navin said:
Any tweeter can be used with a 1st order XO if the XO is high enuff. The more extended its range down low, and the flatter its impedance curve the better. With my speakers i'm XOing the T in at about 10k and i'll try almost anything. Best one yet thou was a little 2" isophon alnico cone tweeter.
dave
general rule of thumb for tweets for 1st order x-os......
with minimal correction and component count ::::::
for ferro-fluid tweeters cross at 2.5+ x Fs
for non ferro fluid, cross at 3+ x Fs
It all depends on the woofer, its not just a tweeter thing, you have to consider things as a whole combination!
For instance, an 8" woofer is generally best crossed below about 2200, hence requires a beefy tweeter with a very low Fs, say below 700 eg Seas 27TFFC,
A 6.5" can maybe go up to say 3000, so a tweeter with Fs below about 1000 is required for a 1st order x-o
5" maybe up to 4000 etc etc
these are all very much generalisation, there are always exceptations.
with minimal correction and component count ::::::
for ferro-fluid tweeters cross at 2.5+ x Fs
for non ferro fluid, cross at 3+ x Fs
It all depends on the woofer, its not just a tweeter thing, you have to consider things as a whole combination!
For instance, an 8" woofer is generally best crossed below about 2200, hence requires a beefy tweeter with a very low Fs, say below 700 eg Seas 27TFFC,
A 6.5" can maybe go up to say 3000, so a tweeter with Fs below about 1000 is required for a 1st order x-o
5" maybe up to 4000 etc etc
these are all very much generalisation, there are always exceptations.
Re: general rule of thumb for tweets for 1st order x-os......
Maybe a bit irrelevant but just found a kit design by Aurum Cantus using their infamous G2 ribbon tweeter and some 5" OEM midbass. One inductor + zobel for woofer and one capacitor for ribbon. I guessed the value to be 6.6uF, which should bring the crossover point strait to the tweeter’s resonant frequency - at 1700Hz. Of course tweeter is padded down by L-pad but still makes me wonder about tweeter (and ribbon as such) protection needs.
argo
Maybe a bit irrelevant but just found a kit design by Aurum Cantus using their infamous G2 ribbon tweeter and some 5" OEM midbass. One inductor + zobel for woofer and one capacitor for ribbon. I guessed the value to be 6.6uF, which should bring the crossover point strait to the tweeter’s resonant frequency - at 1700Hz. Of course tweeter is padded down by L-pad but still makes me wonder about tweeter (and ribbon as such) protection needs.
argo
An externally hosted image should be here but it was not working when we last tested it.
plenty of speakers use such an xover
one cap, one inductor. my old Tannoy 605 monitor (6.5" poly woofer, 1" aluminum dome) was one such. ok it had a resistor thrown in too but only to pad the tweeter. that design is from about 10 years ago and sounded very good for a low-cost speaker - excellent imaging, the hallmark of a good 1st order design.
many many of the good simple 2-way high end speakers use such a crossover as well. can't name them all off the top of my head but Meadowlark and Audio Concepts comes to mind. NSM Audio as well, again with maybe just a resistor on the tweeter to match the output level to the woofer. i'm currently reviewing the Windjammer speaker by Role Audio:
http://www.roleaudio.com/windjammer.html
one resistor, one inductor, and one cap. maybe two caps actually, but the 2nd one is just a small bypass for the larger one. not sure where it's crossed over but i think around 3khz. i don't recommend a xover point much higher than 3khz with any woofer larger than 4" as you'll get into the cone breakup range. the 4.5" woven carbon fiber driver and compact 1" soft dome (really tiny neodynium magnet) in this speaker sound excellent, and they are both shielded. very nice speaker.
one cap, one inductor. my old Tannoy 605 monitor (6.5" poly woofer, 1" aluminum dome) was one such. ok it had a resistor thrown in too but only to pad the tweeter. that design is from about 10 years ago and sounded very good for a low-cost speaker - excellent imaging, the hallmark of a good 1st order design.
many many of the good simple 2-way high end speakers use such a crossover as well. can't name them all off the top of my head but Meadowlark and Audio Concepts comes to mind. NSM Audio as well, again with maybe just a resistor on the tweeter to match the output level to the woofer. i'm currently reviewing the Windjammer speaker by Role Audio:
http://www.roleaudio.com/windjammer.html
one resistor, one inductor, and one cap. maybe two caps actually, but the 2nd one is just a small bypass for the larger one. not sure where it's crossed over but i think around 3khz. i don't recommend a xover point much higher than 3khz with any woofer larger than 4" as you'll get into the cone breakup range. the 4.5" woven carbon fiber driver and compact 1" soft dome (really tiny neodynium magnet) in this speaker sound excellent, and they are both shielded. very nice speaker.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.