Tri-amp my Klipschorn speakers?

leedal · October 24, 2005

Got a pair of 1987 Klipschorn's with AK-2 crossovers. They sound fine attached to my Marantz 2325 and other amps I have, but I'd like to try tri-amping if there is any benefit. Recently I picked up a quantity of BGW SPA-3 tri-amps. These are apparently 225 watt/channel amps and under the front cover have adjustable gain/digital delay, etc and built-in electronic crossover ahead of the amps - 3 amps in each box. They were intended for large venues where various speakers can be placed in different locations in the room and and various delays and cut-offs set. Bottom line - can I make my Klipschorns sound better by splitting each of the drivers and using these amps, instead of using my existing factory crossovers and full range amps? Is the difference worth the hassle?

DrWho · October 24, 2005

Well it's always worth trying, especially if you already have the equipment!

I really like the idea of time-delay correction...how small are the time increments? You'll need to be somewhere around 7ms delay for the tweeter and 6ms on the squaker. And how steep are the slopes in the unit?

pauln · October 24, 2005

Intended for large venues, 200W+? My first impression is that the first watt will sound bad.

artto · October 25, 2005

Ive pondered doing this for years and eventually came to the conclusion that its not worth it. One of the beautiful things about Klipschorns is that you can use lower power extremely high quality amplification that the low efficiency speaker crowd cannot. Bi-amping and tri-amping are typically used in large sound reinforcement systems, primarily to drive multiple drivers (in the same horn for instance) and achieve higher sound pressure levels/power output capacity. Some audiophiles have attempted to adopt this strategy in an attempt to supply specialized amplification to the drivers used in each frequency range. Usually, these are the same people who are using low sensitivity speaker systems that consume phenomenal amounts of power, and often, actually need more amplifier power to achieve realistic sound pressure levels even in modest residential environments.

<?xml:namespace prefix = o ns = "urn:schemas-microsoft-com:office:office" />

The reality of the situation is that every component you add to your system adds more noise and distortion in one form or another. So unless you are truly in need of increasing the sound coverage over a much larger area than typically found in residential rooms, have an extremely large listening space which requires higher speaker output levels, or are trying to optimize the amplification for each driver (3 BGW amplifiers of any kind aint gonna do it), IMHO, less is more. There are other things you can do to improve system performance and increase your enjoyment of music much more than hanging three 225 watt stoves on each driver, especially when that speaker system is a Klipschorn, because each of those particular amplifiers highest distortion is exactly in the power output range where they will be working 99.99% of the time with a Klipschorn. There are also several papers on the subject in the Klipsch Dope From Hope dealer newsletters and the Klipsch Audio Papers.

DrWho · October 25, 2005

The reality of the situation is that every component you add to your system adds more noise and distortion in one form or another.

The implementation of an active crossover includes getting rid of the passive crossovers, so you actually end up with the same signal chain length (In fact, a digital active EQ will have a shorter signal path!). And it just happens to be the case that active digital equalization introduces much less distortion and has a longer more consistent life than the stock passive crossovers. So if you're going to update the crossover, you might as well go active and implement bi-amping. From what I'm learning in school there are also some reactive aspects to diaphragm motion that benefit by having a single amp powering it (versus sharing an amp amongst 3 drivers). It's really not about achieving more SPL, but rather less distortion.

Anyways, the difference is subtle but there is definetly an improvement. The added benefit to tri-amping is that you can also introduce time-delay correction. Roy Delgado claims that time-alignment is vital and that's good enough in my books. I don't think he would go through all the trouble to demo the speakers with time-alignment if he didn't feel it resulted in marked improvement.

formica · October 25, 2005

The added benefit to tri-amping is that you can

also introduce time-delay correction. Roy Delgado claims that

time-alignment is vital and that's good enough in my books. I don't

think he would go through all the trouble to demo the speakers with

time-alignment if he didn't feel it resulted in marked

improvement.

OT; What did he demo (at the pilgrimage) with the time delay? Do you know what Roy is personally using?

jrt · October 25, 2005

I asked a similar question a while back, and didn't get much response. I'll be interested in what you think about the sound if you do it- and also give me another 4 months or so and I'll tell you what I think (grin). I'm finishing up a Dynaco ST70 clone that I think will be the 'third amp' in my system- currently I'm running a 300B for the squawker/tweeter and an SS device for the bass horns. I really like the improved bass from the SS addition- that's a no-brainer.

Bottom line for me is- if you have the cables, the time and the amps, it's worth the experiment. If you don't like it, you can always put it back.

The biggest drawback is that it is now damn complicated to turn on the system (8 switches). My lovely wife is not too happy with that aspect, but she humors me.

D-MAN · October 25, 2005

I would also agree with experimenting if you have the equipment.

This brings up the thought, though - putting all of the money ahead of the problem, so to speak.

The Khorn with stock drivers and horns can only be taken so far before you hit the "point of diminishing returns".

IMO, perhaps the money could be spent in other areas to greater effect.

DM

PrestonTom · October 25, 2005

Some comments were made about inroducing distortion with such scheme. I would also be worried about that.

However DrWho (who seems knowlegible) stated that the signal path would still have as many elements in the chain (passive crossover vs active) and that active crossover would have less distortion.

Well, this is shock to me. First, there is absolutely no need for big amps (and are the big amps necessarily going to be quiet enough for these efficient speakers). Second, there would be many new pieces in the signal chain: ADC & DAC conversion, various op amps etc. Third, I cannot believe that running a signal through some passive devices (inductors & capacitors) would produce relatively more distortion (these are "quiet" components).

Clue me in guys....

Good luck,
-Tom

sfogg · October 25, 2005

Tom,

"there would be many new pieces in the signal chain: ADC & DAC

conversion, various op amps etc. Third, I cannot believe that running a

signal through some passive devices (inductors & capacitors) would

produce relatively more distortion (these are "quiet" components)."

I tri-amp my LaScalas. I used an autoformer to lower the hiss level from the amp that feeds the tweeters.

Before I put the autoformer inline I measured it. Where I am using it

(above 8k) it looked fine. But down at 400hz (where they are run down

to and below) it added odd order distortion to the signal. The active

crossover (with the ADCs, DSP, DAC an opamps) don't add any measurable

harmonic distortion there.

It is also possible to bi/tri-amp using passive line level crossovers

for those that think they are allergic to opamps and digital.

Shawn

DrWho · October 25, 2005

When running an active digital processor you are sending the

coax/optical digital out from your source (say a cd player). All the

processing is performed in the digital domain so there is no loss in

quality. At the output you will have a set of 3 digital to analog

converters. However each individual speaker channel only sees one DA conversion

so the net result is still one DA conversion in the signal path....and since you would

still have a single DA conversion from the output of the cd player,

you've really added nothing while getting rid of the entire passive

crossover network.

If you are running an analog source (say an LP or reel to reel), then

you will have one AD conversion and then one DA conversion that could

possibly affect the sound quality. I believe very strongly in the

listening tests performed on the "analog freak audiophiles" that show

they couldn't detect a difference between a straight wire and a wire

with

an AD to DA conversion in the middle. Once in the digital domain there

is no loss of sound quality regardless of the number of operations

(provided you don't screw it up with the operations you're

implementing) [] So the net result here is the reduction of the

passive crossover in exchange for an extra AD and DA conversion.

But nevertheless, if you're still uncomfortable with the digital domain

then by all means implement an analog line-level network. I can think

of a few units off the top of my head that sound great and don't add

noise, but you sacrifice the ability for time alignment. In this

situation you end up with the same net signal chain (each of the three

speaker channels still see a single passive crossover network and an

amplifier). Theoretically it should be easier to implement a lower

distortion line-level network because the power handling capabilities

of the components are much less of a concern (so you can avoid

saturation and large signal nonlinearities while also implementing

steeper slopes more easily not to mention the impedance response in the

line-level domain tend to be much more constant).

artto · October 25, 2005

So, are you trying to convince me that 3 amplifiers of that same make & model per speaker, especially of that particular type, will produce less distortion and noise than one per speaker? And that introducing an additional electronic device is going to decrease distortion and noise? LOL. Hell, just all the additional connections/signal breaks in that situation is enough to introduce additional distortions and noise. There is absolutely no benefit to bi or tri amping in domestic situations other than attempting to tailor the sound of each amplifier to each driver.

I haven't read Roy Delgado's collective thoughts on the subject of time-delay correction but I will say that its been pretty common knowledge for decades that time coherence is certainly important (under certain conditions), at least the middle frequencies where our hearing is most sensitive, and where the Klipschorn basically utilizes one driver across that range. While digital (anything) has its virtues, I also think it has a long way to go. To me, this is a situation where the cure is worse than the disease.

<?xml:namespace prefix = o ns = "urn:schemas-microsoft-com:office:office" />

And while I agree in principal with what they are teaching you in school about the reactive aspects of diaphragm motion, I ask you to consider, what are the reactive properties of diaphragm motion in a Klipschorn? Isnt it designed to keep diaphragm motion to a minimum? Maybe this has some significance for the typical professional near-field studio monitor with a five inch woofer, or some insensitive power hungry audiophile speaker, but its of far less significance with a Khorn in a domestic situation.

The time-delay correction exercise with a Klipschorn is an interesting one (and has been done before in some of the aforementioned Klipsch publications), but not with three 225 watt P.A. stoves hanging on each driver. Those amps are great for everyday always on use. I do the same with my Crown stuff (but again, not 225 watts per speaker). IMHO leedal would be far better off dumping the BGW on Ebay and reinvesting in some high quality low power tube or class A solid state amps. As DM mentioned, there is a point of diminishing returns, and there are other more important areas to improve performance first at less cost.

sfogg · October 25, 2005

"