What's best in highly sensitive/efficient speakers.

[audioquattr]

Hi Dave,

Agree with your points. Have you tried a line level passive crossover to alleviate the “burden” on the SET powered mid/high? I am thinking of building line level passive for my current project such that my 46 SET amp only “sees” >250hz. My rationale is lower distortion due to not being required to reproduce the current heavy stuff. Also I am using active Mosfet driven bass.

I would think it doesn't see < 250hz when you cross at speaker, there is no load than so no current.
You don't dissapate power with a 250hz highpass crossover, so there's no power generated there (<250hz) so no distortion created (there).
In my experience you loose more at tiny level in signal than at speaker level, imo.
It is only much more expensive with good components at Sp level.

 

[morricab]

If it were me, I would go active for your open baffle subs. Also not sure how many drivers you were planning per side, but I would advise to not be shy when designing lol. I feel people underestimate how many you actually need to get the bass response you desire.

Your bass is active, right?

 

[morricab]

Hi Ron,
Mostly coherence I’ll use exactly the same neodymium 15 inch woofers already used for the pap horns OB but just in a narrow band application sub with isobaric loading.

Two things can potentially happen at this point I’m thinking. I’ll either find I just want to add a mild extension and authority to the paps but not forego ultimate coherence and so just integrate it all in a passive 2.5 way crossover being driven from just the one pair of 805 output tubes. The minimalist in me is very attracted by that notion of SET purity. But only by building and listening will I ever be sure.

I will also try a plan B and go active for the subs with a mosfet SS. Bill’s logic in this approach is also very much worth investigating.

I am just guessing that 4 x 15 inchers for a sub will give it the authority I’m looking for and that using the same drivers all the way from 300hz down (2 x 15 OB woofers and 4 x 15 OB subwoofers will be a good approach).

Either way I’d very much like to start simple and build up to it and enjoy the discovery of what works and what doesn’t along the way.

I’d likely try slot loading the four subwoofers before then going for more woofers. I’m trying to avoid overscaling with the subs and it’s just simply easier (and more economical) to just work upwards.

I’m a complete newb at this so just moving towards learning and having some fun along the way.

I do think that the further you reach down the higher you need to reach up so if all this lovely bass foundation requires more air to bring it to a good weighted sense of centred-ness I could then also try floating a supertweeter in on top.

So Plan C is to just then give myself a completely modular and detachable OB platform which I can then experiment with a different approach altogether for then a four way horn setup as alternative with a pair of iwata horns for a 300hz and above... so an iwata wide bander with an iwata suoertweeter. I do love the form of the iwata horns and so may just end up with two alternative horn setups. Surely two pairs of horns can’t hurt

Kind of doing like me now playing with elliptical horn mids and/or tweeters. Get an active xover with DSP for design work at least.

 

[audioquattr]

Totally agree - the infamous B&K curve / research says it all (seen below).

Actually I recommend all audiophiles to have a play with target curves using dsp packages if you are into computer audio. It taught me a lot about what my ears liked and disliked. A 20hz to 20khz flat profile is ear bleeding for me.

I agree. I have played a lot with target curves and learned a lot. Also learned that i don't want to touch a signal digitally, you allways can here it makes the system less transparent even with the best Dacs and system behind that.

But playing with different speakers through the years and measuring them (in room) i find it hard to refer to a target curve.
Than the way of measuring the spl curve is very arguable. Pinknoise playing for a few seconds give different results than a swipe and short full pikes even more. A very fast speaker will measure and for sure perceive different.

Even a very fast and high efficient speaker can be slower at both ends than from 200-2000hz.
And can show different Frequency response also dependent how is measured.
Some speakers reach 90% of a spl in e.g. a microsecond and 100% after a milisecond, others 40% in e.g. a microsecond and 100% after a second..

 

[the sound of Tao]

Kind of doing like me now playing with elliptical horn mids and/or tweeters. Get an active xover with DSP for design work at least.

Thanks Brad yes, but I’m strangely completely dsp agnostic... I always figure what dsp does kind of inherently works against what I’m trying to listen for in the big picture anyway. I’ll just go old school and keep it analogue in the post source stages where possible. I’m patient and happy for the right way to open up as we go.

 

[morricab]

Thanks Brad yes, but I’m strangely completely dsp agnostic... I always figure what dsp does kind of inherently works against what I’m trying to listen for in the big picture anyway. I’ll just go old school and keep it analogue in the post source stages where possible. I’m patient and happy for the right way to open up as we go.

I understand, which is why I was suggesting for "design purposes only". You would not believe how many iterations you can go through in a very short time this way....something that would take months or even years can be done in a few weeks max. Then you can go try to build an analog equivalent or analog active xover etc...assuming you haven't relied heavily on equalization to make it all work...then you are kind of stuck with needing DSP.

I could have done all my rapid prototying on the Behringer digital xover as it has all the bells and whistles one would need to optimize a design but I was always thinking that maybe staying digital (I won't use analog most likely with this system) would work great if I could match it with good DACs, as I have done now...much better than the built in DACs in most commercial digital xovers. So, now with this system I can perhaps envision staying with a digital xover from minidsp and two dacs and two amps. If my ideas expand to a three way (adding deeper bass) then I might mix active and passive or add another dac and amp...or maybe just go all passive...let's see.

 

[Audiophile Bill]

I would think it doesn't see < 250hz when you cross at speaker, there is no load than so no current.
You don't dissapate power with a 250hz highpass crossover, so there's no power generated there (<250hz) so no distortion created (there).
In my experience you loose more at tiny level in signal than at speaker level, imo.
It is only much more expensive with good components at Sp level.

My stereo amp (46 amp) is getting the full signal from the preamp - this same amp then sends this to the speaker crossover normally where is deals with bass and mids/highs. Using the passive line level crossover my stereo amp will only see >250 because the remaining signal will be dealt by active Mosfet bass amp. Hope that is clearer.

 

[Audiophile Bill]

Your bass is active, right?

Yes my current bass is in the Swings and my horn project is also.

 

[audioquattr]

My stereo amp (46 amp) is getting the full signal from the preamp - this same amp then sends this to the speaker crossover normally where is deals with bass and mids/highs. Using the passive line level crossover my stereo amp will only see >250 because the remaining signal will be dealt by active Mosfet bass amp. Hope that is clearer.

Yes i understand.
But i mean when you do it with (highpass) speaker crossover than your 46 amp will also only see >250, because of what i wrote before.

 

[Audiophile Bill]

Yes i understand.
But i mean when you do it with (highpass) speaker crossover than your 46 amp will also only see >250, because of what i wrote before.

I don’t know what you mean, Jeroen. If you mean biamping with a passive crossover?

This is just the tip of the iceberg though. There are countless further advantages of PLLXOs or ALLXOs.

Best.

 

[audioquattr]

No i mean when you use a passive crossover (just high pass 250hz) than your main (tube)amp will only see >250hz load.

There is no impedance (load) under 250hz, so no current there, no power, so no distortion created there.
So the same as you want to achieve in a crossover before the amps.

Offcourse there is still some load when you have a first order or 2nd or 3rd, but that's also the same as a crossover before the amps.
I hope you understand me Bill, otherwise i try in dutch ;-)

 

[DaveC]

I don’t know what you mean, Jeroen. If you mean biamping with a passive crossover?

This is just the tip of the iceberg though. There are countless further advantages of PLLXOs or ALLXOs.

Best.

I think I now understand your previous post fully...

Jeroen is saying that the end result of line level vs speaker level xo will be the same, with the 250 Hz speaker level xo your amp will not have a load at <250 Hz, so it won't produce power. It's kind of counter intuitive in a way since the amp is getting a full range signal in.

I also agree that a stand alone PLLXO isn't a great idea, but it can be done with coupling caps in preamp outputs or between stages in tube amps pretty easily. Otherwise I'd go with speaker level, it has less overall effect.

And, maybe try a Crown amp w/dsp for woofer, it's excellent imo. NCore is almost too good (I have NC500 monos), my Crown sounds more cohesive with my EL34 SET driving the mids and highs, but NCore has an iron grip and very low distortion.

 

[audioquattr]

I think I now understand your previous post fully...

Jeroen is saying that the end result of line level vs speaker level xo will be the same, with the 250 Hz speaker level xo your amp will not have a load at <250 Hz, so it won't produce power. It's kind of counter intuitive in a way since the amp is getting a full range signal in.

I also agree that a stand alone PLLXO isn't a great idea, but it can be done with coupling caps in preamp outputs or between stages in tube amps pretty easily. Otherwise I'd go with speaker level, it has less overall effect.

....

Yes Dave, that is exactly what i mean.

 

[Audiophile Bill]

I think I now understand your previous post fully...

Jeroen is saying that the end result of line level vs speaker level xo will be the same, with the 250 Hz speaker level xo your amp will not have a load at <250 Hz, so it won't produce power. It's kind of counter intuitive in a way since the amp is getting a full range signal in.

>> I don’t get it. The net result will not be the same unless in the speaker level scenario, I would biamp and use another amp on the woofer.

I also agree that a stand alone PLLXO isn't a great idea, but it can be done with coupling caps in preamp outputs or between stages in tube amps pretty easily. Otherwise I'd go with speaker level, it has less overall effect.

And, maybe try a Crown amp w/dsp for woofer, it's excellent imo. NCore is almost too good (I have NC500 monos), my Crown sounds more cohesive with my EL34 SET driving the mids and highs, but NCore has an iron grip and very low distortion.

 

[Audiophile Bill]

Yes Dave, that is exactly what i mean.

It doesn’t make any sense to me. It will *only* be the same net effect if one were using 2 amps biamping with the passive speaker level crossover.

 

[Audiophile Bill]

Here is an interesting discussion on PLLXO - outlines a lot of the advantages clearly.

https://www.audiocircle.com/index.php?topic=57219.0

 

[Audiophile Bill]

It is very obvious that an amplifier requires less “work” in this configuration if you read the above.

 

[DaveC]

IMO the "correct" answer depends a lot on the speaker in question. If you want/need Zobels, baffle step, etc. that's not as practical to do with PLLXO. If you're talking a simple slope or rolling off the bass to a single driver then yes, this can be done just by exchanging values of coupling caps in many systems.

In the context of a multi-way traditional horn system, many drivers will benefit from impedance correction networks to make load optimal for a high output impedance amplifier and then imo it's just better to go all passive.

I'd be suspicious of statements like Roger's in the AC thread that make declarative blanket statements, they are always wrong.

 

[morricab]

It is very obvious that an amplifier requires less “work” in this configuration if you read the above.

I think you are right Bill. In the PLLXO the amp is not asked to produce the bass in the first place and the second it produces the full signal which is then partially absorbed by the speaker crossover...the distortion generated in the amp will be different and overload potential different as well.

 

[Audiophile Bill]

I think you are right Bill. In the PLLXO the amp is not asked to produce the bass in the first place and the second it produces the full signal which is then partially absorbed by the speaker crossover...the distortion generated in the amp will be different and overload potential different as well.

Hi Brad,

Yes precisely - the speaker crossover parts are massively higher values absorbing more current in heat. So using PLLXO you are dealing with no inductors at all, caps in the nano farad range and tiny resistors. You then also have the advantage of higher damping factor for the exact same amp since it is directly coupled to the driver itself without needing to passing through the passive crossover.