passive, horizontal bi-amping with VTLs

[Ron Resnick]

I have the opportunity to purchase a pair of VTL Siegfried Series Is. I presently have VTL MB-750s, and I anticipate getting MartinLogan Neoliths in the future.

I am considering the Siegfrieds to passively, horizontally bi-amp the future Neoliths.

My view is that there pretty much is no such thing as too much power on big MartinLogan hybrid speakers. The MB-750s and the Siegfrieds should output approximately the same power.

If I wanted more low-frequency oomph or I wanted some flexibility to tinker with the frequency response of the system in my room from 23 to 400 Hz the Siegfrieds would afford me the option to experiment with driving the panels with my current MB-750s in triode mode and driving the woofers and subwoofers with the Siegfrieds in tetrode mode (twice the power of triode mode).

If I needed to tinker with the frequency response of the Neoliths from 23 to 400 Hz, with the Siegfrieds in tetrode mode I could boost both the 15" driver (covering 23 hz to 60 hz) and the 12" driver (covering 60hz to 250-400 hz (based on crossover setting)) by 3 dB, or I could boost only the 12" driver by 3dB and keep the 15" driver the same (by switching in the built-in 4 dB attenuator for the 15" driver).

For those who object to passive bi-amping — I totally understand. If MartinLogan follows its past practice it will release an active-crossover version of the Neolith. I would select that product if it is released next year.

1) Do you agree or disagree that whatever is the difference in output power between the MB-750s and the Siegfrieds it is likely to be small and inconsequential (and who knows what the sample-to-sample output variance is from even the same model)?

2) Do you agree or disagree that being able to raise by 3dB the output of a) the 12” driver or (b) the 15” driver or (c) both the 12” driver and the 15” driver might afford some useful flexibility in getting flat frequency response?

Thank you!

 

[AJ Soundfield]

Hi Ron,

I would ask ML to send you an actual measured impedance/phase curve and post it here. The website indicates the Neoliths are 4 ohm nominal, 0.43 ohm @ 20k. That could give your tube amp fits, despite their (high) rated power at 8 ohms.

cheers,

AJ

 

[esldude]

While passive biamping is not optimum. I agree with both 1) and 2) in your original post. I believe you will need to use balanced cabling to the power amps for closest match on overall gain.

I wouldn't worry too terribly much about the low impedance at high frequencies. It will perhaps soften the treble a touch, but the output impedance of the VTL's being not far from an ohm won't let much bad happen otherwise. If it became an issue with distortion on the treble amp, inserting a half ohm or ohm power resistor between amp and speaker would likely clear it up.

Of course using two such amps for 3 db more output and minor happenstance adjustment of the low end is inefficient electrically and economically. If none of that bothers you go ahead with it.

 

[Ron Resnick]

Dear AJ, I see the .43 ohm spec for the Neolith at 20 khz. I wonder why that design results in half the impedance at that frequency as did the Prodigy or the Monolith III. The high frequency roll-off from tube amplifiers on Martin-Logans has never bothered me.

esldude, Thank you for your advice!

 

[AJ Soundfield]

Dear AJ, I see the .43 ohm spec for the Neolith at 20 khz. I wonder why that design results in half the impedance at that frequency as did the Prodigy or the Monolith III. The high frequency roll-off from tube amplifiers on Martin-Logans has never bothered me.

esldude, Thank you for your advice!

Hi Ron,

Here is my concern. Let's take your current Prodigys. As Dr Toole would say, there is better/more relevant information on the side of tires, than current audio specs. Lucky you we have independent measurements, not "specs":

As you can see, the impedance varies all over, worst being a HF where it drops to around 1ohm at 20k. That's not the issue per se, it's what is the impedance (and phase angles) at frequencies that do have significant more music content? If you read JAs comments, he alludes to just that. Luckily, your VTLs seem to have plenty power, even at 4 ohms.
However, as you can see in the power table, output has dropped significantly at 2 ohms, to around 50-60 watts at 3% THD+N.
I think for such an expensive purchase, it might be best for you to know all the relevant information. I'm not saying the VTLs won't work with the Neoliths, just that more info is better than less.
By all means consult with ML to see if they recommend adding a series resistance to the main panel.

Glad the HF roll off doesn't bother you. I listened to a pair of tube driven monoliths over the weekend and the missing triangles with Rimsky-Korsakov certainly stood out for me! Sound otherwise was very nice. YMMV.

cheers,

AJ

 

[esldude]

Hi Ron,

Here is my concern. Let's take your current Prodigys. As Dr Toole would say, there is better/more relevant information on the side of tires, than current audio specs. Lucky you we have independent measurements, not "specs":
View attachment 21919
As you can see, the impedance varies all over, worst being a HF where it drops to around 1ohm at 20k. That's not the issue per se, it's what is the impedance (and phase angles) at frequencies that do have significant more music content? If you read JAs comments, he alludes to just that. Luckily, your VTLs seem to have plenty power, even at 4 ohms.
However, as you can see in the power table, output has dropped significantly at 2 ohms, to around 50-60 watts at 3% THD+N.
I think for such an expensive purchase, it might be best for you to know all the relevant information. I'm not saying the VTLs won't work with the Neoliths, just that more info is better than less.
By all means consult with ML to see if they recommend adding a series resistance to the main panel.

Glad the HF roll off doesn't bother you. I listened to a pair of tube driven monoliths over the weekend and the missing triangles with Rimsky-Korsakov certainly stood out for me! Sound otherwise was very nice. YMMV.

cheers,

AJ

Yes, all of what you have posted is true. I have run into two instances where the sub-1 ohm treble of an ESL and the phase angle on some musical peaks could cause a bit of distortion of moderately powerful amps (both SS amps btw). The addition of a fractional ohm resistor in series took care of it. Combine that with the fact rarely does music have significantly high levels at or near 20 khz and usually nothing bad happens. Tube amps, even of rather modest power haven't been audibly bothered by the condition. What also happens not infrequently is the transformer loaded with a large ESL will have a resonant bump in the upper range. Often right near 20-25 khz. That isn't very sharp and sometimes will lift the upper half octave a bit partly compensating. Of course all of that varies from pair to pair of amp and speaker matching.

 

[AJ Soundfield]

Yes, all of what you have posted is true. I have run into two instances where the sub-1 ohm treble of an ESL and the phase angle on some musical peaks could cause a bit of distortion of moderately powerful amps (both SS amps btw). The addition of a fractional ohm resistor in series took care of it.

Hi esldude, I understand what you are saying, but we don't know what the impedance curve of the neolith is. So how will one know whether adding a fractional (<1ohm) resistor will help? If the impedance at 8k is 2 ohms with a steep phase angle, how is a 0.5 ohm or so resistor going to help when the amp is clipping badly at 50 or so watts? That's a lot of continuous power, but not peak.
I still think Ron should at least ask. That also begs the question of why ML doesn't just include a switchable 0.5-1ohm or so resistor if this fixes things?

cheers,

AJ

 

[esldude]

Hi esldude, I understand what you are saying, but we don't know what the impedance curve of the neolith is. So how will one know whether adding a fractional (<1ohm) resistor will help? If the impedance at 8k is 2 ohms with a steep phase angle, how is a 0.5 ohm or so resistor going to help when the amp is clipping badly at 50 or so watts? That's a lot of continuous power, but not peak.
I still think Ron should at least ask. That also begs the question of why ML doesn't just include a switchable 0.5-1ohm or so resistor if this fixes things?

cheers,

AJ

Well with an esl 2 ohms plus a bit of purely resistive is less capacitive overall, and a touch higher impedance. A step in the right direction. Enough of a step? depends on the particulars. Maybe, maybe not, but a step in the right direction. I only suggest it as it has worked a couple instances for me.

Why does ML or these other ESL makers not include such a thing? Well some sort of do. Nevertheless it all comes down to what sells, what fits your marketing strategy etc. The real story here is why anyone uses passive crossovers whatsoever in this day. It is simply a shoddy way to do things. Inefficient, relatively ineffective, and gives clearly less good results. The important result it gives is in the marketplace. Audiophiles don't like active speaker crossovers, multiple amps, or speakers permanently wedded to a built in amp no matter how well it works. It is illogical, and ineffective and yet people in the business of high end speakers continue to find it a bridge too far.

 

[AJ Soundfield]

Enough of a step? depends on the particulars.

Then hopefully we can agree that Ron is better off knowing the actual impedance curve/phase than not. With an $80k purchase. Hopefully ML can provide it.

The real story here is why anyone uses passive crossovers whatsoever in this day. It is simply a shoddy way to do things. Inefficient, relatively ineffective, and gives clearly less good results.

Well, some of us luddites still know how to make them for other luddites.
But yes, personally, I've been active DSP for 13 yrs and now make products that are available as such, for the non-luddites.

cheers,

AJ

 

[microstrip]

Well with an esl 2 ohms plus a bit of purely resistive is less capacitive overall, and a touch higher impedance. A step in the right direction. Enough of a step? depends on the particulars. Maybe, maybe not, but a step in the right direction. I only suggest it as it has worked a couple instances for me.

Why does ML or these other ESL makers not include such a thing? Well some sort of do. Nevertheless it all comes down to what sells, what fits your marketing strategy etc. The real story here is why anyone uses passive crossovers whatsoever in this day. It is simply a shoddy way to do things. Inefficient, relatively ineffective, and gives clearly less good results. The important result it gives is in the marketplace. Audiophiles don't like active speaker crossovers, multiple amps, or speakers permanently wedded to a built in amp no matter how well it works. It is illogical, and ineffective and yet people in the business of high end speakers continue to find it a bridge too far.

If the impedance of the panel assembly (including transformer) is .5 ohm, adding a serial resistor will attenuate the treble. It is why manufacturers do not add it.

And fortunately some audiophiles know perfectly why they choose passives - it has been explained many times by high-end designers. Until now it gives better final results. Perhaps thinks will change now digital people are changing their minds

 

[Fitzcaraldo215]

If the impedance of the panel assembly (including transformer) is .5 ohm, adding a serial resistor will attenuate the treble. It is why manufacturers do not add it.

And fortunately some audiophiles know perfectly why they choose passives - it has been explained many times by high-end designers. Until now it gives better final results. Perhaps thinks will change now digital people are changing their minds

I am a little confused by the specs on the Neolith x-over at "60, 250-400 Hz". If I understand it correctly, then inserting a resistor on the panel will attenuate all frequencies above 250 or 400 Hz. In effect, the resistor would reduce system efficiency and boost the mid/upper bass on down relative to the mids on up in the panels. The result would not sound like what you auditioned and bought. So, I think it is a very bad idea. It would have to be a high wattage resistor, by the way.

The issue of very low impedance in the extreme highs is something all MLs have had since day one. I think in most cases with most amps it is benign, and, in any case, it is once again part of the sound you auditioned and bought. I would either not worry about it or move on to some other speaker you think sounds better, of which there are quite a few choices even in this price range.

Also, the efficiency here is typical of MLs at about 90 dB, which is about average for most all speakers. I do not think they can benefit much from huge amounts of amp power in the panels, at least. Panel excursions are limited by the spacing between the stators. More power only potentially causes the panel excursions to hit the stators and distort - unpleasant clipping distortion. There should be no damage from this within reason because the stators are insulated. Only the woofers might benefit from extreme added power.

In truth, I do not like passive biamping. No more power is delivered to the speakers using it, since all channels are amplifying a full range signal. If the panels clip, you do not want to continue to increase bass level and power. Even active biamping has its problems and limits, less so to a true subwoofer which has all the necessary sonic integration tools built in. But, active biamping still "sees" the internal passive crossover, unless you modify and rewire the speaker internally to bypass it. And, then, once again, you have changed the sound you bought, and you are bypassing the internal passive crossovers you paid dearly for.

I also do not like analog active crossovers vs. digital ones. Analog xovers, their added cables and connectors, can affect system transparency on the hi pass side by introducing subtle noise and distortion up into the mids and highs, whereas digital xovers tend not to. Anyway, either biamping method adds complexity and cost better served by a bigger single amp, if needed. So, I would try to live with it with a singe amp/channel, upgrading to a bigger one only if necessary. Auditioning and setup are also dramatically simpler.

I like MLs very much, in spite of the Floyd Toole tests. I have had a number of them ever since 1986, starting with CLS Is.

 

[microstrip]

I am a little confused by the specs on the Neolith x-over at "60, 250-400 Hz". If I understand it correctly, then inserting a resistor on the panel will attenuate all frequencies above 250 or 400 Hz. (...)

No it will not attenuate all the frequencies in the same way. The impedance dependents on frequency and the panel capacitance will start dominating the impedance progressively somewhere around 5 KHz - we need proper data to know the values and shape. Anyway there is a reason for ML not inserting it - probably they want it to sound that particular way!

And IMHO it is not a question of liking or not liking biamping. The central question in this thread is what gives us the best subjective results in a particular situation.

 

[Fitzcaraldo215]

No it will not attenuate all the frequencies in the same way. The impedance dependents on frequency and the panel capacitance will start dominating the impedance progressively somewhere around 5 KHz - we need proper data to know the values and shape. Anyway there is a reason for ML not inserting it - probably they want it to sound that particular way!

And IMHO it is not a question of liking or not liking biamping. The central question in this thread is what gives us the best subjective results in a particular situation.

Ok, but then the choice of resistance value is critical and might require experimentation in order to get the desired result in the desired frequency range, assuming there is a subjective sonic problem caused by the panel's HF impedance in the first place. My point is that if there is a perceptible subjective problem in audition, it might be smarter to avoid the speaker entirely in place of something that sounds better from the get go. No sense buying something that expensive in the hope that you can cure problems you subjectively hear later on via modification and experimentation.

 

[microstrip]

Ok, but then the choice of resistance value is critical and might require experimentation in order to get the desired result in the desired frequency range, assuming there is a subjective sonic problem caused by the panel's HF impedance in the first place. My point is that if there is a perceptible subjective problem in audition, it might be smarter to avoid the speaker entirely in place of something that sounds better from the get go. No sense buying something that expensive in the hope that you can cure problems you subjectively hear later on via modification and experimentation.

I referred before why probably the manufacturer feels the resistor is not needed - please see the SL3 review link. However I am sure that Martin Logan will advise prospective buyers of what is possible to do! Wilson Audio do not have level adjustments, but the factory will supply officially approved resistors to adjust bass Q, medium and treble levels on consultation.

If someone wants to have the Neolith sound in his system he must supply the panel with the adequate current at high frequencies - capacitive loads ask for it! IMHO it is not a problem, it is a feature. SET owners should not dream about owning Neoliths ...

 

[Ron Resnick]

After research and discussions I have come to the tentative conclusion that passive bi-amping makes little to no sense.

Perhaps I will wait for MartinLogan to release an external, active crossover version of the Neoliths. I suspect ML will do with the Neoliths what it did with the Monolith IIIs: offer an external passive crossover and an external active crossover. The IIIx for the Monolith IIIs had solid state circuitry in the crossover, but ML will never offer a tube-based crossover.

I have a new idea for bi-amping the Neoliths: VTL MB-750s in triode mode (350 watts) on the panels and VTL MB-450 Series IIIs in tetrode mode (425 watts) on the dynamic drivers. I also would consider Aesthetix Atlas monoblocks on the dynamic drivers (with EVS Ultimate Attenuators to adjust the volume on the Atlases).

The solid state output section of the Atlas would control the cones nicely, but it might be too risky to blend different amplifiers in the lower midrange of the Neoliths (450 Hz). The MB-450s might not control the cones as well as the Atlas but setting the adjustable damping of the MB-450s on high could help there, and I would avoid two different circuit topologies mixing in the lower midrange.