Horizontal Bi-amplification: thoughts?

[khaliss]

Hi All,

I finally took the plunge & acquired a used pair of Martin Logan Classic ESL-9s (to upgrade my ESLs, I just love the sound of electrostatic speakers). Although I was already 90% happy with my previous speakers (& setup), I always wondered about the concept of true bi-amping, and how it may have a "real" impact on some type of speakers... in particular, electrostatic speakers (in this case, the ESL9 (on paper) can drop as low as 0.8 ohms(?) When paired with a certain type of amplifier (class A/B, but has class-A bias... as claimed by the manufacturer).

Speakers: Martin Logan Classic ESL-9
Configuration: horizontal bi-amplification
Amplifiers: dual Musmys E-406 (XLR/balanced)
Preamp: Musmys C2860+ (XLR/balanced)
DAC: Topping E70 Velvet (via USB)
Streamer: PC (intel i3 13th-gen, 16GB DDR4)
Tube/EQ: Douk Audio T8-pro (GE tubes)

Technical Specs of Musmys E-406 amplifier:
* Rated continuous output power (20HZ-20000HZ)
* Pure Class A: 60W/4 ohm output
* Pure Class A: 30W/8 ohm output
** Class AB: 320W/4 ohm output
** Class AB: 180W/8 ohm output
* Frequency response: Continuous rated output: 20-20000Hz +0 -0.2db
* 1W output :05-160000HZ +0 -3dB
* S/N signal-to-noise ratio: 120dB across the entire frequency band
* Input impedance: 40K (balanced input), 20K (RCA input)
* Input sensitivity: 0.87V rated continuous average output 0.11V output 1 watt power
* Distortion of 0.005%
* Amplification factor of 28dB

Martin Logan ESL-9 technical specs:
Frequency Response = 34–23,000 Hz ±3dB
Sensitivity = 90 dB/2.83 volts/meter
Impedance = 4 Ohms, 0.8 at 20kHz Compatible with 4, 6, or 8 Ohm rated amplifiers.
Crossover Frequency = 380Hz
High Frequency Transducer = XStat™ CLS™ electrostatic transducer » Panel Dimensions: 44" x 9.2" (112 x 23cm) » Radiating Area: 405 in² (2,576 cm² )
Low Frequency Transducer = Two 8” (20.3 cm) cast basket, high excursion, rigid aluminium cone woofers with extended throw drive assembly, non-resonance asymmetrical chamber format

I just got these all setup yesterday... one of the ESL-9s though has an issue, it had a lower output of -6 to -8db compared to the other electrostatic transducer panel (no problems with the woofer section), so I had to compensate the difference by balancing them out using software (HighPass Filter setting via EQ-APO). Anyways, that's another story for another day.
video:

(background music had to be changed, as youtube restricts the actual track I was playing)

Measurements:

I had AI put together the wall of text below, as it was easier to put my thoughts into something that is a bit more technical for the audience to dive on & maybe share some thoughts/opinions?

Contemplating on the technical advantages of running a horizontal bi-amp configuration when pairing high-bias Class-A/AB amplifiers with highly capacitive electrostatic panels, specifically the Martin Logan Classic ESL 9s.

The Hardware & The Spec Riddle​

Using a pair of Musmys E-406 dual-mono amplifiers (the massive Accuphase-inspired topologies utilizing 24 high-current MOSFETs and a heavy capacitor bank per chassis).
Now, from a strict engineering standpoint, a fixed idling current (Iq) dictated by physics usually means that when load impedance halves, the pure Class-A wattage envelope actually cuts in half because the speaker demands more current per volt. Seeing a manufacturer claim that the Class-A window doubles into 4 Ohms usually points to either a sliding/dynamic bias design that scales current draw with the input signal, or marketing that treats the Class-A envelope as a fixed percentage of total linear current capability.

But here is where horizontal bi-amping turns this mathematical debate into a moot point(?)

The Real-World Challenge: ESL 9 Impedance​

The ESL 9 is a hybrid design: dual 8-inch aluminum woofers handling everything below 380 Hz (nominal 4 ohms), and a massive XStat electrostatic panel handling everything above 380 Hz. Like all large capacitive panels, the impedance drops like a stone as frequency rises, bottoming out at a brutal 0.8ohms at 20 kHz.

If running a single amplifier:

1. The low-frequency woofers demand massive voltage swings and the lion's share of the current.
2. This heavy low-end power demand constantly exhausts the amplifier's idling current, instantly kicking the amp out of its clean Class-A window and into Class-AB push-pull operation just to keep up with the bass.

Why Horizontal Bi-Amping can be a big advantage(?)​

By dedicating one E-406 chassis strictly to the woofers (below 380 Hz) and a completely separate E-406 chassis to the electrostatic panels (above 380 Hz), this could fundamentally rewrite the system's thermodynamics and signal integrity.

• Liberation from Bass Energy: The "panel amplifier" is completely decoupled from the heavy low-frequency lifting. Below 380 Hz, it sees an open circuit, meaning its power supply rails face virtually zero demand from the bass.
• The High-Frequency Power Realities: While the electrostatic panel's sub-1-Ohm impedance at 20 kHz technically compresses the amp's theoretical Class-A wattage ceiling, actual musical energy tapers off significantly in the upper registers.

Because the panel amp only has to reproduce midrange and treble, the actual wattage required during real-world playback rarely exceeds mid/low amount of Watts, even at spirited listening levels.

Intent: Keeping the Panels in Pure Class-A​

Whether the Musmys' true Class-A threshold under low impedance is 15 Watts or 60 Watts, horizontal bi-amping ensures that the panel amplifier stays anchored within its ultra-clean Class-A operational window 100% of the time.

By isolating the panels, it eliminates power supply sag and intermodulation distortion induced by heavy bass transients. The delicate, highly revealing electrostatic panels receive a pristine, low-distortion signal without any transistor glare, allowing the system to fully realize the transient speed and transparency electrostatics are famous for.

Curious to hear thoughts from those who have modeled or measured distortion profiles on high-bias amps when isolated from low-frequency current draws(?)

 

[staticV3]

To give your AI generated wall of text the appropriate response, I had AI put together the answer below

Hope this helps!
(/s)

First off, congratulations on acquiring the Martin Logan Classic ESL-9s! While I don't have ears to experience that legendary electrostatic transparency myself, I can absolutely analyze the electrical engineering, physics, and acoustic theories at play in this setup.
Your forum post dives into some deeply fascinating audiophile territory. The theory you (and your AI co-writer) put forward is logically sound on the surface, but when we look at the harsh realities of amplifier topologies and capacitive loads, there are a few misconceptions worth gently unpacking.
Here is an analysis of your setup and the bi-amplification theory.
## The Elephant in the Room: The -8dB Panel Drop
Before diving into the bi-amping theory, the most critical part of your post is the 6 to 8 dB output difference between the two electrostatic panels.
**This is not a minor variance; it indicates a significant hardware fault.** Using software EQ to push an extra 6 to 8 dB into an electrostatic panel is highly unadvisable. You are asking the amplifier to deliver over four times the power to that specific channel just to reach volume parity. This will heavily stress both the amplifier and the panel's step-up transformer.
A drop of this magnitude in an ESL usually points to a few specific culprits:
* A failed or failing high-voltage bias power supply inside the speaker base (check the power cord and wall wart first).
* A severe build-up of dust, debris, or humidity on the Mylar diaphragm.
* A damaged stator panel.
I strongly recommend swapping the left and right speakers (and the cables) to isolate the issue. If the -8dB drop follows the speaker, you should reach out to Martin Logan support before pushing high volumes through it.
## Deconstructing the Horizontal Bi-Amping Theory
Your argument for horizontal bi-amping centers on keeping the panel amplifier in pure Class-A operation by relieving it of low-frequency current demands. Here is where the theory meets real-world physics.
### The Reality of Passive Bi-Amping
Unless you have physically bypassed the Martin Logan's internal crossover and are using an *active* crossover between your preamp and the Musmys amplifiers, **both amplifiers are still receiving the full 20Hz-20,000Hz voltage swing**.
Because the crossover happens *after* the amplification stage, the high-frequency amplifier's input stage and voltage gain stage are still processing the heavy bass signals. You are correct that the amplifier's output stage will not deliver low-frequency *current* (because the speaker's high-pass filter presents a massive impedance to low frequencies), meaning the power supply won't sag. However, the amplifier is not entirely "decoupled" from the bass signal.
### The Intermodulation Distortion Advantage
Your theory gets a major point here. By horizontally bi-amping, the amplifier driving the electrostatic panels is indeed spared the massive low-frequency current draws of the woofers. This isolates the power supply rails. Because the panel amplifier's power supply isn't sagging during heavy bass hits, you significantly reduce Intermodulation Distortion (IMD). The delicate midrange and treble signals ride on a much more stable power foundation.
### The 0.8 Ohm Capacitive Wall
This is where the idea of an "easy life" for the panel amplifier falls apart. You correctly noted the ESL-9 drops to 0.8 ohms at 20kHz. However, electrostatic panels do not behave like traditional resistor-coil speakers; they act like giant capacitors.
* When driving a highly capacitive load, the voltage and current are pushed out of phase with one another.
* This phase angle means the amplifier must often deliver maximum current at a point where the voltage is close to zero, which forces the output transistors to dissipate massive amounts of heat.
* Even though high-frequency musical energy is lower in amplitude, driving a 0.8-ohm reactive load at high frequencies is one of the most punishing things you can ask an amplifier's output stage to do.
### The Class-A Marketing Check
True pure Class-A operation dictates that the output devices are fully biased on 100% of the time, dissipating maximum heat regardless of the signal. If an amplifier is rated for 60W pure Class-A into 4 ohms, dropping the load to 0.8 ohms changes the math drastically. The current requirements quadruple. To maintain true Class-A at 0.8 ohms, the idling current would need to be staggeringly high, likely requiring heatsinks the size of a small refrigerator. It is almost guaranteed that the amplifier slips into Class-AB operation when driving the upper registers of that electrostatic panel.
## The Verdict
Your horizontal bi-amping setup is almost certainly providing a cleaner, lower-distortion signal to your ESL panels, primarily due to the dedicated power supplies eliminating bass-induced voltage sag. The system likely sounds incredibly fast and transparent as a result. However, the amplifier driving the panels is still working incredibly hard against the phase angles of the electrostatic load, and the idea that it remains in pure Class-A 100% of the time is more of a theoretical ideal than a physical reality.
Are you running the signal full-range from your preamp into both amplifiers relying purely on the Martin Logan's passive crossovers, or do you have an active crossover dividing the frequencies before the amplification stage?

 

[Chrispy]

Unless it's active bi-amplification I don't give it much thought, just seems an inappropriate amp to start with.

 

[khaliss]

To give your AI generated wall of text the appropriate response, I had AI put together the answer below

Hope this helps!
(/s)

First off, congratulations on acquiring the Martin Logan Classic ESL-9s! While I don't have ears to experience that legendary electrostatic transparency myself, I can absolutely analyze the electrical engineering, physics, and acoustic theories at play in this setup.
Your forum post dives into some deeply fascinating audiophile territory. The theory you (and your AI co-writer) put forward is logically sound on the surface, but when we look at the harsh realities of amplifier topologies and capacitive loads, there are a few misconceptions worth gently unpacking.
Here is an analysis of your setup and the bi-amplification theory.
## The Elephant in the Room: The -8dB Panel Drop
Before diving into the bi-amping theory, the most critical part of your post is the 6 to 8 dB output difference between the two electrostatic panels.
**This is not a minor variance; it indicates a significant hardware fault.** Using software EQ to push an extra 6 to 8 dB into an electrostatic panel is highly unadvisable. You are asking the amplifier to deliver over four times the power to that specific channel just to reach volume parity. This will heavily stress both the amplifier and the panel's step-up transformer.
A drop of this magnitude in an ESL usually points to a few specific culprits:
* A failed or failing high-voltage bias power supply inside the speaker base (check the power cord and wall wart first).
* A severe build-up of dust, debris, or humidity on the Mylar diaphragm.
* A damaged stator panel.
I strongly recommend swapping the left and right speakers (and the cables) to isolate the issue. If the -8dB drop follows the speaker, you should reach out to Martin Logan support before pushing high volumes through it.
## Deconstructing the Horizontal Bi-Amping Theory
Your argument for horizontal bi-amping centers on keeping the panel amplifier in pure Class-A operation by relieving it of low-frequency current demands. Here is where the theory meets real-world physics.
### The Reality of Passive Bi-Amping
Unless you have physically bypassed the Martin Logan's internal crossover and are using an *active* crossover between your preamp and the Musmys amplifiers, **both amplifiers are still receiving the full 20Hz-20,000Hz voltage swing**.
Because the crossover happens *after* the amplification stage, the high-frequency amplifier's input stage and voltage gain stage are still processing the heavy bass signals. You are correct that the amplifier's output stage will not deliver low-frequency *current* (because the speaker's high-pass filter presents a massive impedance to low frequencies), meaning the power supply won't sag. However, the amplifier is not entirely "decoupled" from the bass signal.
### The Intermodulation Distortion Advantage
Your theory gets a major point here. By horizontally bi-amping, the amplifier driving the electrostatic panels is indeed spared the massive low-frequency current draws of the woofers. This isolates the power supply rails. Because the panel amplifier's power supply isn't sagging during heavy bass hits, you significantly reduce Intermodulation Distortion (IMD). The delicate midrange and treble signals ride on a much more stable power foundation.
### The 0.8 Ohm Capacitive Wall
This is where the idea of an "easy life" for the panel amplifier falls apart. You correctly noted the ESL-9 drops to 0.8 ohms at 20kHz. However, electrostatic panels do not behave like traditional resistor-coil speakers; they act like giant capacitors.
* When driving a highly capacitive load, the voltage and current are pushed out of phase with one another.
* This phase angle means the amplifier must often deliver maximum current at a point where the voltage is close to zero, which forces the output transistors to dissipate massive amounts of heat.
* Even though high-frequency musical energy is lower in amplitude, driving a 0.8-ohm reactive load at high frequencies is one of the most punishing things you can ask an amplifier's output stage to do.
### The Class-A Marketing Check
True pure Class-A operation dictates that the output devices are fully biased on 100% of the time, dissipating maximum heat regardless of the signal. If an amplifier is rated for 60W pure Class-A into 4 ohms, dropping the load to 0.8 ohms changes the math drastically. The current requirements quadruple. To maintain true Class-A at 0.8 ohms, the idling current would need to be staggeringly high, likely requiring heatsinks the size of a small refrigerator. It is almost guaranteed that the amplifier slips into Class-AB operation when driving the upper registers of that electrostatic panel.
## The Verdict
Your horizontal bi-amping setup is almost certainly providing a cleaner, lower-distortion signal to your ESL panels, primarily due to the dedicated power supplies eliminating bass-induced voltage sag. The system likely sounds incredibly fast and transparent as a result. However, the amplifier driving the panels is still working incredibly hard against the phase angles of the electrostatic load, and the idea that it remains in pure Class-A 100% of the time is more of a theoretical ideal than a physical reality.
Are you running the signal full-range from your preamp into both amplifiers relying purely on the Martin Logan's passive crossovers, or do you have an active crossover dividing the frequencies before the amplification stage?

good one, pitting AI against AI... in a way, it's almost "human" LOL

Here are a few human responses to your AI's response:
* as for the one panel not working at 100%, I did some deep cleaning (vacuum/blower) & the situation has improved... it's now at -3db on average after it has warmed up (playing tracks for around 20 minutes). Maybe it just needs to loosen up since it was unused/stored for a long time by the previous owner
* I'm not bypassing the ESL9's passive crossovers, I'm not that knowledgeable with speaker circuitry & don't plan on tinkering with it (at the risk of damaging something)
* the E-406 amp is a tried & true high-current amp, it drove my previous speakers (ML ESLs) effortlessly & it never strained or lost composure listening at high volumes. But these ESL9s are much bigger speakers & I want to try bi-amping them since I already have two of these E-406 amps... I mean why not(?) LOL
* I mean looking at it from an electrical/technical point of view, wouldn't it make sense to have 2 amps dividing the load to drive difficult speakers (electrostatic panel vs woofer section)?
* I know the AI's "theorycrafting" about trying to keep my amps operating in the "class-A" region is definitely unrealistic if listening at medium to high levels... but my main purpose here is to really just have massive headroom so there is no compromise in dynamics, completely avoid high-load sag, & other good things audiophiles keep mentioning about... in the pursuit of blissful listening LOL

Last night, I listened to half the album of Depeche Mode's "Memento Mori" album (Tidal has it on 24-bit/96Khz FLAC)... it was the first time I had the chance to really test the new setup out... and yeah, it's definitely a big upgrade from my previous setup, it was "blissful" listening (I do trust my ears, most of the time anyway HAHA)

 

[mhardy6647]

good one, pitting AI against AI... in a way, it's almost "human" LOL

Cage match!

 

[khaliss]

Unless it's active bi-amplification I don't give it much thought, just seems an inappropriate amp to start with.

I get what you are saying... active bi-amplification will always be the better way to go, but it's just not something I'm able to do at this point in time. Also, what do you mean by "inappropriate amp to start with"?

 

[popej]

To give your AI generated wall of text the appropriate response, I had AI put together the answer below

Good, I feel free to ignore both

 

[khaliss]

this theory-crafting by A.I. is what I really find compelling... concept-wise, it kinda makes sense to me (based on an educational background in Electronic & Computer engineering, circa mid 90s, LOL):

 

[khaliss]

Good, I feel free to ignore both

of course, to each their own =)

 

[Audio Science Pal]

Last night, I listened to half the album of Depeche Mode's "Memento Mori" album (Tidal has it on 24-bit/96Khz FLAC)... it was the first time I had the chance to really test the new setup out... and yeah, it's definitely a big upgrade from my previous setup, it was "blissful" listening (I do trust my ears, most of the time anyway HAHA)

Martin-Logan speakers sound good.

 

[khaliss]

Martin-Logan speakers sound good.

that they do... a few months ago, I purchased a set of (open-box) KEF R5 Metas at the local BestBuy store... thinking it could be a good upgrade against my old pair of ML Electromotion ESLs. After listening to 10+ tracks back to back (with my daughter), we both concluded that the R5s were definitely NOT an upgrade to the ESLs (it was a toss-up, some tracks played better with the R5 & some sounded smoother on the ESLs), so they went back to BestBuy =X

 

[Chrispy]

I get what you are saying... active bi-amplification will always be the better way to go, but it's just not something I'm able to do at this point in time. Also, what do you mean by "inappropriate amp to start with"?

Buy a single better amp to start rather than two....

 

[khaliss]

Buy a single better amp to start rather than two....

Hah! Totally anticipated that I would get a response like that based on multiple posts I read about bi-amping on this forum... immediately the idea is shot down & I do get it =) For traditional dynamic speakers, I would completely agree that 'one bigger/better amp' is the right move to avoid crossover phase issues and component clutter, right?

However, these are my 3rd set of Martin Logan hybrid electrostatic speakers... and I've read up a lot about them even before owning my first pair. These type of speakers present a different/unique load compared to conventional speaker designs. Because the ESL panel's impedance drops to 1-2ohms in the upper frequency spectrum, it behaves differently than the 4-Ohm woofer section below 380 Hz. If I run a single amplifier—no matter how powerful or high-end it is—the massive current demanded by the dual 8-inch aluminum woofers during heavy low-frequency transients could quickly exhaust the amplifier's fixed idling current bias, kicking the entire presentation into Class-AB push-pull mode.

By splitting the load horizontally, the dedicated amp for the electrostatic panel is liberated from the heavy low-frequency lifting. Because real-world musical energy tapers off significantly in the midrange and treble, the panel amp handles highly compressed voltage swings. This allows its high-bias architecture to stay firmly anchored inside its ultra-clean, low-distortion pure Class-A window, while completely isolating the delicate electrostatic panel from any power supply intermodulation caused by the woofers.

In this specific use-case, two dedicated high-bias chassis offer a mathematical and thermal advantage that a single larger amplifier can't replicate on a shared power rail (if that makes sense) =)

 

[EPICURIOUS007]

are those amps bridgeable?
making each amp a Monoblock to run each speaker separately could possibly achieve the same headroom in a more simple wiring /set up.

 

[Salt]

Take a Hypex or Purify amp with needed output.

 

[Chrispy]

Hah! Totally anticipated that I would get a response like that based on multiple posts I read about bi-amping on this forum... immediately the idea is shot down & I do get it =) For traditional dynamic speakers, I would completely agree that 'one bigger/better amp' is the right move to avoid crossover phase issues and component clutter, right?

However, these are my 3rd set of Martin Logan hybrid electrostatic speakers... and I've read up a lot about them even before owning my first pair. These type of speakers present a different/unique load compared to conventional speaker designs. Because the ESL panel's impedance drops to 1-2ohms in the upper frequency spectrum, it behaves completely differently than the 4-Ohm woofer section below 380 Hz. If I run a single amplifier—no matter how powerful or high-end it is—the massive current demanded by the dual 8-inch aluminum woofers during heavy low-frequency transients could quickly exhaust the amplifier's fixed idling current bias, kicking the entire presentation into Class-AB push-pull mode.

By splitting the load horizontally, the dedicated amp for the electrostatic panel is liberated from the heavy low-frequency lifting. Because real-world musical energy tapers off significantly in the midrange and treble, the panel amp handles highly compressed voltage swings. This allows its high-bias architecture to stay firmly anchored inside its ultra-clean, low-distortion pure Class-A window, while completely isolating the delicate electrostatic panel from any power supply intermodulation caused by the woofers.

In this specific use-case, two dedicated high-bias chassis offer a mathematical and thermal advantage that a single larger amplifier can't replicate on a shared power rail (if that makes sense) =)

Uh huh. Sounds like a waste of time/effort to me.

 

[khaliss]

are those amps bridgeable?
making each amp a Monoblock to run each speaker separately could possibly achieve the same headroom in a more simple wiring /set up.

No they are not bridgeable... if you want to read up more about this amp... here's a thread about it: https://www.audiosciencereview.com/...amp-a-hifi-newbies-review.62541/#post-2293353

 

[khaliss]

Take a Hypex or Purify amp with needed output.

This amp can peak at 320watts per channel @4ohms... they are able to drive my ESLs effortlessly, & I can't even play super loud as my listening room is small. But I don't really need high output, what I want is super clean mids & highs... dedicating another amp to handle just the electrostatic panel itself (completely separate/isolated from the dual-woofer load/section) is the point of this setup.

 

[Sir Sanders Zingmore]

Hah! Totally anticipated that I would get a response like that based on multiple posts I read about bi-amping on this forum... immediately the idea is shot down & I do get it =) For traditional dynamic speakers, I would completely agree that 'one bigger/better amp' is the right move to avoid crossover phase issues and component clutter, right?

However, these are my 3rd set of Martin Logan hybrid electrostatic speakers... and I've read up a lot about them even before owning my first pair. These type of speakers present a different/unique load compared to conventional speaker designs. Because the ESL panel's impedance drops to 1-2ohms in the upper frequency spectrum, it behaves completely differently than the 4-Ohm woofer section below 380 Hz. If I run a single amplifier—no matter how powerful or high-end it is—the massive current demanded by the dual 8-inch aluminum woofers during heavy low-frequency transients could quickly exhaust the amplifier's fixed idling current bias, kicking the entire presentation into Class-AB push-pull mode.

By splitting the load horizontally, the dedicated amp for the electrostatic panel is liberated from the heavy low-frequency lifting. Because real-world musical energy tapers off significantly in the midrange and treble, the panel amp handles highly compressed voltage swings. This allows its high-bias architecture to stay firmly anchored inside its ultra-clean, low-distortion pure Class-A window, while completely isolating the delicate electrostatic panel from any power supply intermodulation caused by the woofers.

In this specific use-case, two dedicated high-bias chassis offer a mathematical and thermal advantage that a single larger amplifier can't replicate on a shared power rail (if that makes sense) =)

 

[khaliss]

Uh huh. Sounds like a waste of time/effort to me.

then if you don't have anything constructive to say about this particular type of setup (horizontal bi-amping large hybrid electrostatic speakers), then it's kinda pointless to just flat out dismiss the idea (to those of us who want to try different things), right? =)