To Biamp or not to Biamp, that is the question???????

[Apogee]

Hello all,

Advice needed...

I'm about to start building my Alephs (possibly XA's but haven't decided yet). This weekend, I was thinking through the design and the idea came about to build a pair of stereo amps rather than a pair of large monoblocks... Essentially instead of a pair of 1.2's, I'm considering a pair of stereo 2's...

Yes, the heat - but I believe I've got that handled... Water-cooled is the current direction...

So my question is which is the better solution sonically? Biamp using a very high quality active crossover or a pair of large monoblocks biwired to the speakers???? I've always figured that eliminating the passive crossover components in the speakers would be a good idea but so far the electronic crossovers that I've heard have a far worse signature... I've tried most of the crossovers that are available and have currently settled on a pair of Bryston 10B L/R's as the best option that I have found from a sonic standpoint... They are fully discrete, balanced and 24 db/octave using very high quality components. Unfortunately, I can still hear them so I keep going back to the no active crossover option... The next attempt is to try high quality passive components between the amp and the preamp matched to the input impedance of the amp...

System presently consists of Apogee Studio Grands, Threshold Ta-300 (converted to a SA 3.9), Melos SHA Gold preamp, Pioneer DVL-91 Laserdisc player as CD transport and California Audio Labs Alpha 24/96 tube DAC, and lastly Goertz MI2 speaker wire and with home made interconnects using Canary star-quad wire.

Yes Nelson, I would love to try one of your new crossovers but it will take a while...

So, if you were considering building a new pair of monoblocks and the cost of implementing them as a pair if stereo amps vs. monoblocks wasn't the primary issue, what would you do????

Lastly, the Apogee's resolve to an extent that I can easily hear an interconnect change so please keep this in mind when thinking of adding another component to the signal chain...

Thoughts????

Thanks very much!

Steve

PS - Just so you also know the upgrade plan - amps first, preamp (Aleph P or X), new TT with Pearl or Ono, then new CD player and/or DAC...

 

[Peter Daniel]

I think that biamping is a better way to go. That's what I'm doing in my system. I'm using an active crossover at 100 Hz for subwoofers (from Threshold). I would never use active crossover for mids. In my setup I have open baffle Triangle midranges and they roll-off naturaly around 120Hz, 6dB/octave. So for quite a long time I was not using any crossover for midrange drivers. Just recently I placed 0.022uF cap in series with the input of my amp and I think that the sound is slightly better, it feels like midrange breathes more easily.
The active crossover in the mids/highs section will always have its signature so it's better to be avoided. My system has such resolution that I can easily hear the reversing of the cables, not to mention taking top covers off. That's why I also don't use a preamp.
I would build 2 stereo amps and tried to build active crossovers around front stage of the amps. One channel would be for subs, the other for mids and highs.

 

[Nelson Pass]

For guys like us, biamping is the minimum.

All the other usual arguments aside,

I like ultra-efficient drivers, and you can realistically
get 98-105 dB from upper bass all the way to 50 KHz
without much effort, but the woofer, even with a 100 dB
rating, will not deliver until higher frequencies. If you
want to play this way, you must biamp to mate the
bass with the rest of it.

Of course, I am currently quint-amping

 

[audionut]

Harry...

What making the "active crossover" out of passive components such as inductors resistor's and capacitors. This is what I have done, and I think sounds pretty good.

Also, any ideas on how you would build the filter into the front end of an Aleph 3 amp.

Thanx

 

[Peter Daniel]

In the following thread http://www.diyaudio.com/forums/showthread.php?s=&threadid=4098&pagenumber=2 is the schemtic of my current crossover. As you see ea. filter consists of caps and resistors at the input and feedback of ea. discreet op amp. If you are going to use your Alephs permanently as low pass and high pass amps you could probably build those type of filters around the input differencials. Maybe Mr. Pass could share some opinions?
I am minimalist kind of builder but in that case I might even go into tri-amping.

 

[Nelson Pass]

My experience from doing tons of these is partly distilled
into the xvr1 manual, but here are a couple of salient
point:

1) You will have to fool around with every value of the
filter parts to get what you want. This takes a lot of time
and taste, and often you end up very far from where you
started.

2) Driver makers often have a very different taste and agenda
than you do, so don't take their recommendations too seriously.

3) Same for textbook crossover designs.

4) A lot of the time you can use single pole (6 dB/oct) for
all the drivers except the woofers. There you will often
find yourself with 12 or 18 dB/oct.

5) I have never ended up using 24 dB/oct or higher slopes.
Just never liked the sound, don't know why.

 

[Apogee]

Clarification....

Thanks Nelson, we were typing at the same time...

Thanks to all for their comments thus far... Please keep them coming!

I wanted to clarify how the amps would be used...

My Apogees have a tweeter/midrange and a bass ribbon. You can see them here if you're interested...

http://208.51.252.167/studiogrand.htm

The passive crossover from tweet/mid to woofer panel is at ~ 1K starting at 6db and progressing to 12db. What I was considering is using one channel from each of the proposed chassis for the tweet/mid and one channel for the woofer panel - both wired directly to the ribbons... I will then use another crossover point and amp to drive the subs... 3 channels per speaker (tri-amped)...

The line of thinking was that perhaps it would sound better if I bypassed the passive crossovers and drove the ribbons directly with a dedicated channel for each... My assumption being that there would be a benefit because the drivers wouldn't be interacting with each other electrically...

Yes, I realize that this is a lot of power for the tweeter ribbons. I simply wanted to make certain that I had plenty of power for the woofers while at the same time having the same sonic signature from the amps between the tweet/mid and the woofer panels...

So, the dilemma is what is the best solution at this 1 k crossover point since it's smack in the midrange??? Do I just build a pair of monoblocks and biwire or do I attempt to build/find a really really good active crossover????? Which is better in these freqs since I really have no choice - active or passive???

Thoughts....

Thanks,

Steve

 

[GRollins]

I don't know that a vote is the best way to resolve this, but at any rate, I say: biamp.
As folks are no doubt tired of hearing, I'm in a gradual process of rehabilitating my system after years of being mothballed. All well and good, but I also try to have music along the way, even if it's not quite the quality that I'm aiming for in the long run. My first step in the process was to bring out the Magneplanars and put a single amp (Threshold S-500) on them.
Yuk!
I know there are people who claim that you can't remember sound (in which case, how do they recognize their mother's voice, but I digress...), but I <i>knew</i> my system had sounded better than that previously. Lots better. So I scrambled madly for parts and slapped together a crude active crossover. Presto! Serious steps forward in sound quality. Not up to what my system had been capable of in its prime, but a fantastic leap forward.
I've made it up to quad amp territory, now. I still lag behind the glory days in some ways, but have finally managed to exceed the old benchmarks in others. Progress is painfully slow, but I'm getting further along.
It's difficult to beat a simple follower with a cap for simplicity and clean sound.
Don't use opamps; many commercial crossovers do, with predictable effects on the sound quality. Discrete is the only way to go.

Grey

 

[roddyama]

I've been tri-ampping off and on for years depending on the number of working amplifiers on hand. I can't say I have the "golden ears" I may have once had before I worked in the bottling plant of the Stroh Brewery, but I can hear the improvement from a single amp to bi-amp to tri-amp. The improvment comes in all aspects of the sound. The 3 areas that are the most dramatic are the detail, and the depth, but most of all dynamics. Somehow, when you direct couple the driver to the amp, all the compression vanishes.

Rodd Yamashita

 

[macka]

To Biamp or not, that is the question???

Steve,

Sorry I'm a bit slow to pick up on this thread.

I've done a fair bit with active crossovers, both in pro sound and hi fi and I think is more about an evolution of your system rather than a revolution (active crossovers).


So here's my vote:

1. Your situation is interesting but I am curious about the 3 ohm load, this will need hefty amounts of class A bias for the benefits of Class A to be fully realised?

Perhaps Mr Pass and offer you advice on this matter.

2. Its probably fair to say that your 1.0K crossover is of good quality given the resolution your are getting and would be quite senstive any any amplitude variations with other crossover types (in other words if it ain't broke, don't fix it).

So to replace this with an active device would be subjectively quite difficult to improve on over the passive crossover.
(assuming you use Class A power)

3. I expect if you used a big enough Aleph then an active crossover for the 1K crossover would be not justified and there would be little gain from load sharing as 80% of the power is being taken by the mid/bass panel and being Class A, there can't be much harmonic distortion reaching the tweeter ribbons.

4. As to the sub crossover , this can be an absolute jungle and a bastard to get right.

A minimalise approach to the hi pass filter for the bass/mid panels will fair better than any active filter for purist taste.

Assuming you have a reasonable low drive impediance from your pre amp, just use a nice series capacitor/shunt resistor network for 6 db slope here and the panels will continue to roll of naturally being bi polar.

5. An A/B amp with plenty oh headroom for the subs should suffice and you can then tinker with the low pass network for best integration with the panels.

In a nut shell, built your Aleph 1st, then see what happens, you'll be surprised at the improvement.

PS Nelson has a great article on crossovers which you will find useful.

best regards

macka

 

[Bill Fitzpatrick]

You might be interested in the information about bi-amping on our web site at:

http://kbacoustics.com/ygbsm/2waycrossovers/index.html

Call me biased if you wish.

 

[Peter Daniel]

I know there are people who claim that you can't remember sound (in which case, how do they recognize their mother's voice, but I digress...), but I <i>knew</i> my system had sounded better than that previously. Lots better.

Grey

Those people might be right, might be not. I usually don't do A/B tests because they are boring and not consistant. However when everything is right in your system you just know it.
I've been playing that game for some time recently and that feeling (about right setup) comes and goes, depending on changes you do.
But when everything locks in place, even the bad recordings sound good, there is magic in the air, you discover every album again, each instrument and sound exists in its own 3-dimentional space and suddenly you realise that everything they write about in Stereophile is true. But what's important, you don't get tired listening and you feel good about the time you wasted to achieve it. Isn't DIY all about it?

It's kind a like adjusting VTA. Usually, there is only one "perfect" position. I decide it to right about it, because just today I achieved that bloody state,.... after a week of tweaking.

 

[Apogee]

Thanks all for the responses!

Ok - Biamping will be the direction so one more question if I may...

I have possibly found a source for some nice heatsinks so I may not have to go water cooled after all...

If I plan on total dissipation of 600w per chassis (one chassis per side with two channels per chassis) should I go for a lower output wattage with higher biasing or higher output wattage with the bias reduced a bit???

I'm thinking that I can either put a pair of Volksamp 60's with the bias turned up, or perhaps set the rail voltages for approximately 80 watts and use lower bias...

I know typically that things sound better with higher bias but I'm not sure how the two stage Aleph will handle the low impedance loads that the ribbons create (currently 3 ohms). I've only heard the Os and it wasn't through my speakers.

Is there a preference or performance gain either way when driving low impedance loads?

Thanks again for all of your opinions/comments!

Steve

 

[Nelson Pass]

High bias does a good job of curing the problems
driving low impedances. I think it will work fine.

 

[FullThrottleRic]

Of course, I am currently quint-amping

Is it just me?!

I'd really like to see some pics of NP's system

 

[GRollins]

If you're saying that the Apogees are 3 ohms, you're going to need to jack up the bias a bit or face current limiting. The Volksamp versions of the Aleph circuit had even lower bias than the equivalent Pass Labs models, so proportionately they will require even more boost on the current. Heat sinks will need to be larger in consequence.
Your circumstances aren't all that different from what I'm facing in biasing an Aleph-X for the Magneplanar ribbon tweeters (about 2.5 ohms). Although I posted the Aleph-X showing 15V rails, mine will be more along the lines of 12V, possibly even less, with the bias current about 50% higher. If I decide to go with an Aleph-X for the Bohlender-Graebener RD-75s (about 5.5 ohms), I won't have to go with such high bias, although the rails will need to be higher. All in good time...got other things on my plate at the moment.
As for pictures of Nelson's system, I imagine that they would need a time stamp, as the pictures would be obsolete within a few weeks.

Grey

 

[PedroPO]

biamp question

Hey Guys!

Need some help pls.

I have two amps, one SOZ and one Audio Note Kit One.

I am tring to biamp using SOZ for the low frequencies and the AN for the High ones.

I connect the amps with separate wires to the terminal of the speakers (infinity delta 50) and use the speaker crossover.

From what understand you build your own crossovers. Are the ones in your speakers not good enough for you or you speakers do not have separate inputs (or do you build your own speakers and crosovers? ;-)

Another issue is how to calibrate both amps to deliver the same "amount" of sound.

I use my ear, but can't we use a microphone, a sinusoidal constant sound and a voltmeter to find the correct adjustment?

Thanks,
Pedro Oliveira

 

[Peter Daniel]

The whole idea of biamping is to use crossover before the amps. That way ea. amp is working in it own frequency range. The system is far less subject to intermodulation distortion effects and you can achive increase in dynamic range which is about twice that of conventional passive system. Additional, there is direct and intimate coupling between the amplifier and the loudspeaker driver that is not possible with a passive system. This allows control of the loudspeaker driver by the full damping factor avalable from the amplifier.
I always use test disc and a mike/meter to adjust the system.

 

[GRollins]

Why not try the SOZ for the high frequencies and the Audio Note for the low frequencies?

Grey

 

[Bill Fitzpatrick]

Lest anyone forget.

You can't have a 1st order slope on a satellite speaker, for example, if the crossover frequency is meant to be somewhere around the F3 of the satellite. The low frequency roll-off of the satellite has its own order - 2nd order for sealed and 4th order for vented. Adding a 1st order transfer function simply results in a 3rd or 5th order.

Even if the crossover frequency is 2 x F3 of the satellite, it still won't happen.

All this talk of first order crossovers in a bi-amping situation needs to be squelched.

In my experience, the best thing to do with a sealed satellite is to use a 2nd order high pass filter whose F3 matches the F3 of the satellite. Then use a 4th order low pass for the woofer where F3 is -6db. You wind up with the classical LR filter.