I've been reading the ESP pages on bi-amping (www.sound.au.com). Rod (the author) is obviously a huge fan of it and backs his arguments up with explicit explanations. This has got me very interested in bi-amping indeed. But, I'm after other opinions from those who've tried it. Is it really the best thing you can do for your system without spending mega dollars?
I'm currently running Leach amps into some cheap three-ways and was thinking about bi-amping using some LM3876 IC amps for the tweeters. I'm on a budget (who isn't??), hence the LM3876s (ESP also seemed to rate the LM3876s well for some cheap ICs).
I look forward to reading your replies.
I'm currently running Leach amps into some cheap three-ways and was thinking about bi-amping using some LM3876 IC amps for the tweeters. I'm on a budget (who isn't??), hence the LM3876s (ESP also seemed to rate the LM3876s well for some cheap ICs).
I look forward to reading your replies.
I am thinking along the same lines, but have deferred it for a while until I get a bigger listening room.
My thinking was that the best amplifier (probably Class A) should be used for midrange, where most of the content is. Lower power but still clean LM3876 for tweeters (I think that's what Rod is going to use). Woofers would probably be ESP project 3A.
My thinking was that the best amplifier (probably Class A) should be used for midrange, where most of the content is. Lower power but still clean LM3876 for tweeters (I think that's what Rod is going to use). Woofers would probably be ESP project 3A.
I first went biamped somewhere in the late '70s; achieved triamp status in the early '80s. In '88 my personal world turned upside down and I was lucky to listen to music at all for several years. The first time I put the system back together with the intent to leave it going, I used a single amp. I couldn't figure out where all the music had gone. Surely, my memory of how good my system had sounded wasn't that far off...was it? I played around with this and that, and finally got around to biamping again. Poof! Some of the music came back.
To make a long story short, I'm now quad-amped and loving every minute of it. It is not cheap. It is not always technically easy. It is, however, essential in getting the absolute best sound that you can get from your system.
Grey
To make a long story short, I'm now quad-amped and loving every minute of it. It is not cheap. It is not always technically easy. It is, however, essential in getting the absolute best sound that you can get from your system.
Grey
Hi,
I have been using biamping for the last 8 years. Yes, it does raise the level of performance. But to obtain the full benefit, go active, i.e. use an active crossover and eliminate the passive ones in your speakers. You will be amazed with the improvements. Tighter bass, improved transparency, cleaner treble, etc.
N.S.
I have been using biamping for the last 8 years. Yes, it does raise the level of performance. But to obtain the full benefit, go active, i.e. use an active crossover and eliminate the passive ones in your speakers. You will be amazed with the improvements. Tighter bass, improved transparency, cleaner treble, etc.
N.S.
GRollins,
I've been thinking of biamping my speakers for a long time now (using 24dB/octave Linkwitz-Riley active x-over) but I still haven't.
The thing is that I don't like to have opamps in the signal-path (which 99,9% of the active x-over uses), do you use mosfets or tubes instead of opamps?
One more thing, do you compensate for the so called baffel step? Or shouldn't I be worried about that at all?
/Freddie
I've been thinking of biamping my speakers for a long time now (using 24dB/octave Linkwitz-Riley active x-over) but I still haven't.
The thing is that I don't like to have opamps in the signal-path (which 99,9% of the active x-over uses), do you use mosfets or tubes instead of opamps?
One more thing, do you compensate for the so called baffel step? Or shouldn't I be worried about that at all?
/Freddie
Freddie,
Once upon a time, I used LF353 opamps in a circuit similar to the one Linkwitz wrote about in one of his papers where you compensate for the rolloff of the driver (I was using KEF B139s for subs in those days) by boosting the signal in exact accordance to the slope of the rolloff. It takes a pretty healthy amplifier to do this, as you're burning power by the pound as your frequencies decrease. After a lot of R&D, I finally got flat response into the upper teens. Pretty good. But the problem was that the drivers were being driven very, very hard and were quite non-linear by the time decent listening volumes were attained. In short, it was flat, but not tight at all above a certain volume level. Pity, as it sounded so good at low volumes that you just wanted to crank it up a notch or two or three...
I have since heard several commercial offerings that use the same strategy and they all tend to have that same muddy sound when played at even moderate volumes. After noticing the pattern, and figuring out why it happened, I gave up entirely on the technique. My current subs will eventually use a feedback network (optical, if I can make my idea work out) that reads the cone motion and returns to the amp as a negative feedback signal to correct for both driver and amp distortions.
My current crossover (very much a work in progress) uses 2N5457s (FETs) and MPSA18s (bipolars) in follower configuration for the mids and highs. The original LF353 crossover boards are still in use for the subs, though, as I haven't taken the time to put together a discrete lowpass section, not that it's difficult, there just aren't enough hours in a day. Note that I've removed the board that compensates for the KEF B139 rolloff (I'm using twelve 12" Titanic drivers these days). This was simple, as I've always tried to build modular projects so as to be able to upgrade or reconfigure at a later date; I simply removed the boost board. (It's now on my video system upstairs, pushing KEFs again.)
My plan is to build a tube crossover later, at least for the mids and highs, once I've got some other things settled out. I've got a bunch of 6922s waiting to take flight--it's back to the time factor. Way too many ideas to implement in a day, or even a month, and I'm far behind on my writing, which does earn a penny or two, unlike audio, which costs money.
It's quite simple to use a gain device as a follower instead of an opamp. A follower has a very high input impedance, and low output impedance, just like an opamp. Use a Sallen-Key configuration, and you'll be just fine. Yes, there'll be some insertion loss, as followers are always less than unity gain. In practice, it's not enough loss to be a problem.
I'm not sure what you're talking about when you refer to a baffel step.
Grey
Once upon a time, I used LF353 opamps in a circuit similar to the one Linkwitz wrote about in one of his papers where you compensate for the rolloff of the driver (I was using KEF B139s for subs in those days) by boosting the signal in exact accordance to the slope of the rolloff. It takes a pretty healthy amplifier to do this, as you're burning power by the pound as your frequencies decrease. After a lot of R&D, I finally got flat response into the upper teens. Pretty good. But the problem was that the drivers were being driven very, very hard and were quite non-linear by the time decent listening volumes were attained. In short, it was flat, but not tight at all above a certain volume level. Pity, as it sounded so good at low volumes that you just wanted to crank it up a notch or two or three...
I have since heard several commercial offerings that use the same strategy and they all tend to have that same muddy sound when played at even moderate volumes. After noticing the pattern, and figuring out why it happened, I gave up entirely on the technique. My current subs will eventually use a feedback network (optical, if I can make my idea work out) that reads the cone motion and returns to the amp as a negative feedback signal to correct for both driver and amp distortions.
My current crossover (very much a work in progress) uses 2N5457s (FETs) and MPSA18s (bipolars) in follower configuration for the mids and highs. The original LF353 crossover boards are still in use for the subs, though, as I haven't taken the time to put together a discrete lowpass section, not that it's difficult, there just aren't enough hours in a day. Note that I've removed the board that compensates for the KEF B139 rolloff (I'm using twelve 12" Titanic drivers these days). This was simple, as I've always tried to build modular projects so as to be able to upgrade or reconfigure at a later date; I simply removed the boost board. (It's now on my video system upstairs, pushing KEFs again.)
My plan is to build a tube crossover later, at least for the mids and highs, once I've got some other things settled out. I've got a bunch of 6922s waiting to take flight--it's back to the time factor. Way too many ideas to implement in a day, or even a month, and I'm far behind on my writing, which does earn a penny or two, unlike audio, which costs money.
It's quite simple to use a gain device as a follower instead of an opamp. A follower has a very high input impedance, and low output impedance, just like an opamp. Use a Sallen-Key configuration, and you'll be just fine. Yes, there'll be some insertion loss, as followers are always less than unity gain. In practice, it's not enough loss to be a problem.
I'm not sure what you're talking about when you refer to a baffel step.
Grey
Another advantage to bi/tri/quad-amping is you never run out of things to build: "You're building ANOTHER amplifier?" becomes easier to answer. I always have lots of long-term ideas on the go, but I get around to building very few of them. Hmmm, sounds like somebody else in this thread...
For me, the journey really is the reward.
Edit: I also meant to add:
While you may want to avoid op-amps for the crossovers, my feeling is it's probably better to tri-amp with op-amps than single-amp with tubes (or whatever your preferred active devices are).
Grey, any comments about where to place the highest-quality amps, given that you're starting with legacy equipment and slowly evolving the system?
[Edited by paulb on 10-19-2001 at 02:37 PM]
For me, the journey really is the reward.
Edit: I also meant to add:
While you may want to avoid op-amps for the crossovers, my feeling is it's probably better to tri-amp with op-amps than single-amp with tubes (or whatever your preferred active devices are).
Grey, any comments about where to place the highest-quality amps, given that you're starting with legacy equipment and slowly evolving the system?
[Edited by paulb on 10-19-2001 at 02:37 PM]
I think that the placement of amps is subjective. For example, if you're combining tubes with SS amps, you may want to have the SS controlling the bass. In my situation, I have two SS, one of my amps had better highs, and the other a more defined midrange, so I used them accordingly. Sounds infinitely better than when I had them reversed. If you have big power differences however, I'd put the lesser powered amps up with the tweets and work my up as I go down the audio spectrum.
Yesss, know the parts about power, Rod Elliott summarizes it nicely in his article at his ESP site.
btw, I just received Linkwitz-Riley crossover PCBs from Rod; I don't plan to use them right away but I'll have them on hand for the future. They look very good, quality- and flexibility-wise. If one were to use high-quality op-amps, Rs and Cs I very much doubt it would be worthwhile to use a bunch of tubes instead. IMHO. But I'll never actually compare the two approaches anyway.
btw, I just received Linkwitz-Riley crossover PCBs from Rod; I don't plan to use them right away but I'll have them on hand for the future. They look very good, quality- and flexibility-wise. If one were to use high-quality op-amps, Rs and Cs I very much doubt it would be worthwhile to use a bunch of tubes instead. IMHO. But I'll never actually compare the two approaches anyway.
Grey,
When I talked about baffel step I meant this.
The baffle of the box produces half-space acoustic loading down to the frequency where one-quarter wavelength of sound equals the baffle size. A 20 cm wide baffle will start to "unload" from about 400 Hz, and the speaker's response will drop by 6 dB below 400 Hz. This phemonemon is often referred to as "the baffle step". The "step" is not absolute, but more like a 1st order function.
Good loudspeakers take the baffel step into account to make a more flat frequency response. And if I use a normal active crossover (Rod Elliot Linkwitz-Riley) then I will have a dip in the response below 400Hz (that can't be really good). So how do I take the baffel step into account when using active crossovers?
/Freddie
When I talked about baffel step I meant this.
The baffle of the box produces half-space acoustic loading down to the frequency where one-quarter wavelength of sound equals the baffle size. A 20 cm wide baffle will start to "unload" from about 400 Hz, and the speaker's response will drop by 6 dB below 400 Hz. This phemonemon is often referred to as "the baffle step". The "step" is not absolute, but more like a 1st order function.
Good loudspeakers take the baffel step into account to make a more flat frequency response. And if I use a normal active crossover (Rod Elliot Linkwitz-Riley) then I will have a dip in the response below 400Hz (that can't be really good). So how do I take the baffel step into account when using active crossovers?
/Freddie
Paul,
As for amps, you and Bryan have pretty well covered it. Make sure you've got something nice on the midrange. Generally speaking, solid state does a better job on bass due to better damping factor, although there are rare exceptions where a tube amp will do well. For mids and highs, it's up for grabs as to whether to use tubes or solid state--listen and use whichever sounds better.
On chips vs. discrete in a crossover, keep in mind that a follower is a single device compared to all the stuff in a chip. To the extent that many (myself included) feel that fewer devices tend to produce better sound, you'll have a difficult time getting less than one device.
At this point someone invariably comes up with the old chestnut about the sound having been through so many opamps in the recording studio that it won't hurt to use another one (or two, or three...). Muddy thinking. Use of an opamp isn't an all-or-nothing proposition; use of a single opamp, even an antique 741, will not kill the sound, it simply degrades it. Every additional opamp (or discrete device for the matter) degrades the sound a little more, like adding a little more grime to a window. You can still see outside, but the view deteriorates a bit with each speck of dirt. The objective is to keep the cumulative degradation to as low a level as possible. The studio used opamps? Well, rats. But don't punish the signal any more...it's not its fault. Treat it kindly and it will reward you.
Freddie,
I take it that the baffle has an unterminated end? Like a reflex cabinet using a shelf to create a rectangular port or a TL cabinet divider? I'd say it's a cabinet design problem. I tend not to use reflex cabinets at this point--nothing wrong with them, they just aren't appropriate for the drivers I'm using. Transmission lines are calling to me again, and I may build a pair or two in the near future. I've got some good drivers on hand and some scrap MDF to use up. The thing about TLs is that they're generally more heavily damped than you'd expect for a 'normal' cabinet. That and the taper (yes, I know some people don't taper their lines, but that's asking for trouble) tend to reduce or eliminate the problems you encounter in other cabinet designs.
That said, anything you can do with a passive crossover you can do with an active one, and generally do it better. Got a droop in your woofer response? Supply it with more signal to compensate. Got a pesky peak? Use a notch filter. Nothing to it.
Grey
As for amps, you and Bryan have pretty well covered it. Make sure you've got something nice on the midrange. Generally speaking, solid state does a better job on bass due to better damping factor, although there are rare exceptions where a tube amp will do well. For mids and highs, it's up for grabs as to whether to use tubes or solid state--listen and use whichever sounds better.
On chips vs. discrete in a crossover, keep in mind that a follower is a single device compared to all the stuff in a chip. To the extent that many (myself included) feel that fewer devices tend to produce better sound, you'll have a difficult time getting less than one device.
At this point someone invariably comes up with the old chestnut about the sound having been through so many opamps in the recording studio that it won't hurt to use another one (or two, or three...). Muddy thinking. Use of an opamp isn't an all-or-nothing proposition; use of a single opamp, even an antique 741, will not kill the sound, it simply degrades it. Every additional opamp (or discrete device for the matter) degrades the sound a little more, like adding a little more grime to a window. You can still see outside, but the view deteriorates a bit with each speck of dirt. The objective is to keep the cumulative degradation to as low a level as possible. The studio used opamps? Well, rats. But don't punish the signal any more...it's not its fault. Treat it kindly and it will reward you.
Freddie,
I take it that the baffle has an unterminated end? Like a reflex cabinet using a shelf to create a rectangular port or a TL cabinet divider? I'd say it's a cabinet design problem. I tend not to use reflex cabinets at this point--nothing wrong with them, they just aren't appropriate for the drivers I'm using. Transmission lines are calling to me again, and I may build a pair or two in the near future. I've got some good drivers on hand and some scrap MDF to use up. The thing about TLs is that they're generally more heavily damped than you'd expect for a 'normal' cabinet. That and the taper (yes, I know some people don't taper their lines, but that's asking for trouble) tend to reduce or eliminate the problems you encounter in other cabinet designs.
That said, anything you can do with a passive crossover you can do with an active one, and generally do it better. Got a droop in your woofer response? Supply it with more signal to compensate. Got a pesky peak? Use a notch filter. Nothing to it.
Grey
Well, I'm definitely going to bi-amp my system now. I'm keen to get it underway today but it's a project that'll have to wait a month or so (I've also got exams looming).
Tri-amping sounds like a good excuse to build another pair of amps too.
Any thoughts/advice about bi-amping two-ways (crossed-over at 5000kHz)? Everything I've ever seen on bi-amping refers to 3-ways.
Tri-amping sounds like a good excuse to build another pair of amps too.
Any thoughts/advice about bi-amping two-ways (crossed-over at 5000kHz)? Everything I've ever seen on bi-amping refers to 3-ways.
Yes, you do need to deal with the baffle step.
However, the standard theory, which assumes that there is a
6dB drop as the frequency drops past some frequency related
to the minimum dimension of the baffle, is too simple, as
it assumes that half the low-frequency signal disappears
through your front wall, never to reappear. In practice
some of it does come back, so to compensate for the acoustic
loss you will need to apply an electronic step somewhat
less than 6dB.
The standard theory and a passive compensation approach is
described on John Murphy's site:
http://www.trueaudio.com/st_diff1.htm
Many speaker designers use LF drivers with a response that
gently rises with frquency (as many of them do), so that
when they apply the canonical 6dB boost in the crossover to
compensate for the lost LF it is partially compensated by
the acoustic response of the driver and gives a more natural
sound.
In my own experience a simple active 6dB boost sounds wrong,
and is hard to correct elswhere, unless you're using just
the right woofer. I used a 3dB step in my active setup for a
while, and this sounded much better than either no
correction or a 6dB correction.
I'm working on a circuit that has the step height adjustable
in 1dB increments, and a variable centre frequency:
http://www.soton.ac.uk/~apm3/diyaudio/Diffraction.html
Alex
However, the standard theory, which assumes that there is a
6dB drop as the frequency drops past some frequency related
to the minimum dimension of the baffle, is too simple, as
it assumes that half the low-frequency signal disappears
through your front wall, never to reappear. In practice
some of it does come back, so to compensate for the acoustic
loss you will need to apply an electronic step somewhat
less than 6dB.
The standard theory and a passive compensation approach is
described on John Murphy's site:
http://www.trueaudio.com/st_diff1.htm
Many speaker designers use LF drivers with a response that
gently rises with frquency (as many of them do), so that
when they apply the canonical 6dB boost in the crossover to
compensate for the lost LF it is partially compensated by
the acoustic response of the driver and gives a more natural
sound.
In my own experience a simple active 6dB boost sounds wrong,
and is hard to correct elswhere, unless you're using just
the right woofer. I used a 3dB step in my active setup for a
while, and this sounded much better than either no
correction or a 6dB correction.
I'm working on a circuit that has the step height adjustable
in 1dB increments, and a variable centre frequency:
http://www.soton.ac.uk/~apm3/diyaudio/Diffraction.html
Alex
Was just thinking that ppl biamping could use the parametric equalizer in the Soundblaster Live/Audigy cards to get the crossover frequencies / dB boost that you like. It'll probably make the active crossover easier to design, and you can try a few crossover points.
The idea is to connect the front output to the midrange while connecting the rear output to the tweeter. Then just edit till you get the sound you want.
The idea is to connect the front output to the midrange while connecting the rear output to the tweeter. Then just edit till you get the sound you want.
bm_mode,
The ESP site mentioned earlier focuses more on biamping. Rod recommends 300 Hz crossover. See the article at http://www.sound.au.com.
The ESP site mentioned earlier focuses more on biamping. Rod recommends 300 Hz crossover. See the article at http://www.sound.au.com.
Too much amps loading in the house?
Hi,
I have just completed two of the four JLH for ESL mono blocks now, the older pair is bias at 4A each, the new pairs are 3A each. If I go bi-amp I will have
4 + 4 + 3 + 3 = 14 amp on the house line. In Canada we normally have 15A max house hold wiring , correct? I think I have a problem here. Since all other things like Rotel 5 channel A/V, 61" TV, one 300W subwoofer, one 150W subwoofer, pre-amp, DVD player, MSB LINK DAC, VHS player and DIP upsampler all connect to the same house line.
Now with 2 more new amps with a active cross over too.
Any suggestions?
Thanks,
Chris
Hi,
I have just completed two of the four JLH for ESL mono blocks now, the older pair is bias at 4A each, the new pairs are 3A each. If I go bi-amp I will have
4 + 4 + 3 + 3 = 14 amp on the house line. In Canada we normally have 15A max house hold wiring , correct? I think I have a problem here. Since all other things like Rotel 5 channel A/V, 61" TV, one 300W subwoofer, one 150W subwoofer, pre-amp, DVD player, MSB LINK DAC, VHS player and DIP upsampler all connect to the same house line.
Now with 2 more new amps with a active cross over too.
Any suggestions?
Thanks,
Chris
As you should know in every home on Canada there is 240V AC available and you can split your amps requirements between two separate AC lines which will carry only half of the total load. That's exactly what I'm doing and monoblocks in ea. channel are powered by separate, dedicated 30A AC lines taken from separate phases.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.