Audeze LCD 4 - Which amp?

[Deke609]

Most scopes will handle a considerable amount of AC voltage, generally far more than you're going to see on the speaker jacks of a single ended tube amp.

For measuring distortion, a scope isn't going to tell you all that much unless it has FFT functionality built in.  A $50 soundcard and some free software will though!

Thanks again Paul.  I have a cheapish usb scope (Picoscope 2204A), which I figured wouldn't be as robust as as a benchtop model.  But it apparently has FFT (although I have no idea how to access that functionality for reasons made clear by what follows).

I may need to go the soundcard and software route, at least initially -- I spent about 6 hours last night trying to figure out how to get the o-scope "working" and cursing it for not being immediately intuitive (I'm sure it works just fine - it's my brain that's the problem).

To a more technically adept onlooker my last night's misadventure in oscilloscoping would have looked like a scene from Zoolander: "The files are in the computer?" ... Smashes computer on floor and looks for physical files.  I couldn't even figure out how to connect its output to the amp - solved that this morning by soldering some banana posts to a rca plug and connected to the right channel input of the amp.  So far I've managed to get a frequency sweep from the scope to play through the right headphone, with the scope input connected to same speaker taps that connect the headphones, but I have no idea (yet) how to get the scope to measure and plot dB's by frequency.  Sheesh ... plug-and-play has ruined my ability to figure things out.

I'll look into the soundcard/freeware option this evening.

[Deke609]

Derek, keep up the good work.
Keep posting about the upgrades and how they affect the Stereomour II with your headphones.

If your still planning on a 45 conversion; be sure to let us know how it worked out!

And could you perhaps test/compare with other headphones?

I'm very happy to learn that there's more after Mainline for headphones (and speakers ).

Thanks for the encouragement Tom!

Re 45 conversion: I hadn't actually planned to do it, b/c I'm concerned the drop in wattage (from 3.5 Wpc to I think 1.75 Wpc) would leave me with an amp too quiet for the LCD4s.  But that may change depending on: (a) the results of my future experiments with the other impedance configurations for the SII (4, 8, 16 ohms): and (b) whether I can resist getting a Kaiju next year.   

As currently configured (2 ohms) and in its present state (not even burned in), the SII is just stunning.  The DCF upgrade completely eliminated the hum, and the Shunt Regulator upgrade was mindblowing - among other things, the bass is now punchy and precise.  A good test track for slam (to my ears) is ZZ Top's "Master of Sparks" from their Tres Hombres album - the opening drums out of the SII through the LCD4s are incredible.

Re other headphones: I only have 3 pairs: the LCD4s, Focal Elex and HD6xx.  I am listening to the the Elex as I write.   The Elex is a Massdrop/Focal collaboration that is essentially the Elear with Clear pads, and reported lowering of the Elear's original bass response (I've never heard the Elears, so can't comment on this).  They are easier to drive and much more sensitive than the LCD4s, so I think they make a nice comparison. 

Some specs for the Elex (from the Massdrop website):

Impedance: 80 ohms
Sensitivity: 104 dB SPL / 1 mW at 1 kHz
THD: < 0.3% at 1 kHz / 100 dB SPL
Frequency response: 5 Hz–23 kHz

Specs for the LCD4s (from the Audeze website):

Impedance:   200 ohms
Sensitivity:   97 dB/1mW (at Drum Reference Point)
Voltage sensitivity:   95 dB/1V
THD:   <0.1% @ 100dB
Frequency response:   5Hz – 50kHz

Both pairs of headphones sound great out of the SII configured for 2 ohms.  I prefer the LCD4s, but only because they are for my tastes a superior headphone. There is no hum that I can notice through either pair of headphones - but that may change when I configure the amp for 8 ohms - particularly, I imagine, with the more sensitive Elex.

My (loud-ish to loud) volume listening ranges on the Mourquiet Attenuator upgrade for the 2 headphones:

Elex:  -27 dB to -22.5 dB
LCD4: -9 dB to -3 dB

So, my tentative conclusion is that a fully upgraded SII is a stellar option for a wide range of headphones.  I'll try the HD6xx later this week.

cheers,

Derek

[Deke609]

Switched the OT configuration from 2 Ohms to 8 Ohms this afternoon and, my oh my, it is GLORIOUS. 

The SQ change is immediately obvious.  Most noticeable: bass is A LOT MORE detailed, present and authoritative. At 2 Ohms I felt I could still use a bit more bass (even after the Shunt Regulation upgrade, which gave the bass a lot more punch and precision). But at 8 Ohms, the bass is perfect for my tastes (and I like a fair amount of bass - I like my music visceral). Overall, at 8 Ohms music is both more precise and more fluid - which I would have thought contradictory, but that's the best way I can think of describing it.  Everything is more coherent.  And the added gain has given me a good deal of headroom. I went from regularly listening at 2 Ohms with the course attenuator at 0 dBs to listening at 8 Ohms withe course attenuator at -9 dBs. Sweet!

There are no downsides that I can hear.  No hum with the DC Filament even at 8 Ohms.

Conclusion: A fully upgraded SII configured for 8 Ohms is a SUPERB match for the LCD4s. 

Many thanks to everyone at BH who played a role in creating this fantastic amp.

cheers,

Derek

[Deke609]

@Doc, PJ and PB - I'm considering putting a BeePre between my DAC and the SII, for use with my LCD 4s. The preamp gets rave reviews here for soundstage, detail, fullness, and other wonderful sonic qualities. My only holdback is a concern about noise. I suspect I'll be ok, but want to check with you guys first in case this is a "no way, it'll be way too noisy" pairing for my headphones. 

The SII has been dead silent since I installed the SR upgrade [oops, I meant the DCF upgrade], even configured for 8 ohms.  Doing the hum pot adjustment/testing, I am down to 0.1 mV on each channel. But even if I adjust the pot to allow 1+ mV, I still hear nothing.  The LCD 4s are not that sensitive (97 dB) and have 200 Ohms impedance.  And they can take a fair bit of current and voltage. At 2 Ohms, I could max out the volume on some tracks, and the headphones seem to love the added gain at 8 ohms (the bass really opened up and everything sounds both more fluid and controlled). And at 8 Ohms I haven;t had to turn the volume attenuators down that much - just one turn of the coarse attenuator (-9dBs for loudish and -18 dBs for normal/quieter listening). So I think I found a winning combination with the LCD4s and the SII.  But if I can get even more of the BH sound with the BeePre ...

Thoughts?

[edited to correct error]

[Doc B.]

You don't need a preamp with the Stereomour II. The BeePre is super quiet in terms of power supply, but there is the potential issue of microphonics in the directly heated 300Bs if you can't put the preamp somewhere where it will be isolated from external vibration.

OK, the disclaimers are out of the way. Sonically it could be a very nice combination. Let us know when you are ready to order. 

[Deke609]

Thanks Doc.

Ready!

Microphonics, if it becomes an issue, are something I can probably fix.  My rack rests on spring-loaded self-leveling feet.  I've read the threads you started about sorbothane feet and lead tube rings. And Jamie R's recent posts about his vibration control box using an inflatable cushion and heavy plate.  So there are lots of potential, stackable solutions.

cheers and thanks,

Derek

[Deke609]

Order for Beepre and BeeQuiet placed    I had originally earmarked this money for a SS amp, also American made, but canceled my order of same after hearing just how good the SII is.  The SS amp was for greater bass extension - but now I'd rather focus my efforts on getting more of the BH sound from my system. (That said, if BH ever makes the Kaiju iron upgrade for the SII, I will be the first to order - I bet there are enough SII owners on this forum that would jump at the upgrade to make it do-able).

[Deke609]

Guys - Now that the amp has 100+ hours on it, and I've experimented with all 4 output impedance configurations, I'd like to return to the power/distortion and damping issues.  And just to be clear, I'm not trying to "correct" anything. I am perfectly happy with the amp - it sounds stellar at 4, 8 and 16 ohms configurations (in comparison with which, 2 Ohms sounds a bit dry and lifeless - but I was perfectly happy with 2 Ohms before hearing 4, 8 and 16  ). So this is just for fun and education - if I somehow manage to squeeze out a bit more SQ improvement from all this, that's just a bonus.

You previously suggested building a stereo L-pad that would present a fixed 8 ohm load to the amp.  I have the required parts, but I don't know whether I should proceed because I'm unsure whether my headphones need the amp output to be "padded down". I suspect not, given that I can comfortably listen to them with the amp configured for 16 Ohms with the volume attenuators at -18 dB, at -9 dB when configured for 8 ohms, and pretty much wide open at 4 Ohms. 

But I may well be missing something - so maybe it's helpful to return to something PB wrote a month ago in reference to using an L-pad:

If the resistor values get too large, you'll lose damping of the headphones and the bass will get lumpy on you. If the resistor values are too small, you'll load the amp down and it will distort.

Based on some reading I've done, my understanding is that the damping/lumpy-bass issue is a function of a headphone's impedance curve across the audible frequency bandwith. My guess is that if the headphone impedance dips appreciably at certain bass frequencies, a high output impedance could result in loss of damping at those frequencies and the bass will be loose and boomy. Do I have it right?

If my understanding is correct, then I don't think I need to worry about damping. The LCD 4s have an input impedance of 200 Ohms and the impedance versus frequency plot is dead flat - literally a straight line parallel to the x-axis for all frequencies.  This has been tested and verified by various reviewers (Inner Fidelity, Stereophile, etc.). My understanding is that this is a common characteristic of planar magnetic headphones.

So my current thinking is that (a) no padding down is necessary, and (b) headphone damping is not a problem.  BUT I have come across numerous references in online forums about "keeping an amp happy" by putting a resistor in parallel with each headphone channel -- e.g., 2 8 ohm resistors for an amp configured for 8 ohm speakers.

1st issue: Does the SII need to "see" the resistance for which it is configured to be "happy"?  If so, is this what PJ was referring to when he mentioned "parafeed circuit resonance affected by amp load impedance"?  And what PB meant when he mentioned too little resistance "loading the amp down" and causing it to distort?

One other thing I don't understand is how power hungry headphones can be be driven fully if resistors having a value much smaller than the headphone's impedance are placed in parallel with the headphones. While the headphones will still get the same voltage, wouldn't most of the current flow across the parallel resistor, bypassing the headphones?  My thinking is that if voltage is fixed by the OTs and the amp cannot exceed 3.5 to 4 Wpc, then current is necessarily limited such that, with the parallel resistors in place, the headphones may not be able to draw as much current as they need.  But maybe I misunderstand this too.

2nd issue - Distortion. I've tried to read up on this, but there are so many different types of distortion and I don't know which one or ones I should be educating myself about - what kind(s) of distortion are we talking about? And is it produced at the amp or the headphones?

Many thanks,

Derek

[Paul Birkeland]

I suspect not, given that I can comfortably listen to them with the amp configured for 16 Ohms with the volume attenuators at -18 dB, at -9 dB when configured for 8 ohms, and pretty much wide open at 4 Ohms. 

Yeah, no reason to pad things down.  If you were all the way down on coarse and fine and things were loud, then you'd want to pad down the output a bit.  If the general noise floor of the amp was too high, the padding would also remedy that.

Based on some reading I've done, my understanding is that the damping/lumpy-bass issue is a function of a headphone's impedance curve across the audible frequency bandwith.

When load impedance gets lower than output impedance, fluctuations in the load impedance become very audible.  For example, an 8 ohm rated speaker that dips to 4 ohms and up to 50 ohms will have the potential for lumpy bass issues unless run from the 4 ohm setting (or lower).

If my understanding is correct, then I don't think I need to worry about damping. The LCD 4s have an input impedance of 200 Ohms

Yeah, plenty of damping on any impedance setting.

I have come across numerous references in online forums about "keeping an amp happy" by putting a resistor in parallel with each headphone channel -- e.g., 2 8 ohm resistors for an amp configured for 8 ohm speakers.

Our amps remain quite happy with just about anything you want to load them with.  They are unconditionally stable.

1st issue: Does the SII need to "see" the resistance for which it is configured to be "happy"?  If so, is this what PJ was referring to when he mentioned "parafeed circuit resonance affected by amp load impedance"?  And what PB meant when he mentioned too little resistance "loading the amp down" and causing it to distort?

What PJ is describing and what I brought up are two different issues.  There's no harm in putting loading resistors in parallel with the headphones, I wouldn't pretend to be able to predict whether you'd like the outcome or not.  A load that is too low in impedance relative to what the amp is expecting will generate extra distortion and a little more power.  You don't need either.

One other thing I don't understand is how power hungry headphones can be be driven fully if resistors having a value much smaller than the headphone's impedance are placed in parallel with the headphones. While the headphones will still get the same voltage, wouldn't most of the current flow across the parallel resistor, bypassing the headphones?  My thinking is that if voltage is fixed by the OTs and the amp cannot exceed 3.5 to 4 Wpc, then current is necessarily limited such that, with the parallel resistors in place, the headphones may not be able to draw as much current as they need.  But maybe I misunderstand this too.

The resistor L-pad reduces the power available to the headphones.  Putting an 8 ohm resistor in parallel with each headphone driver will slightly reduce the output of the amp, so you will see slightly reduced power to your headphones as well. 

2nd issue - Distortion. I've tried to read up on this, but there are so many different types of distortion and I don't know which one or ones I should be educating myself about - what kind(s) of distortion are we talking about? And is it produced at the amp or the headphones?

We don't really go into a whole lot of discussion here on distortion generated by transducers.  The distortion we are mentioning is harmonic distortion. 

[Deke609]

Many thanks again, PB.

It's good to know that my SII is happy no matter the load. 

If I were to play around with parallel resistors, what value(s) would you suggest?  I think you previously stated that twice the intended load was a good starting point.

And thanks for clarifying that it is harmonic distortion that we are talking about.  That might be something fun to experiment with. In a 1972 paper for the AES (see link in "Why do I like tube amps" thread under "General Discussion), Russell Hamm suggested that it may actually be harmonic distortion that is responsible for why many people find tube sound to be fuller bodied and punchier than SS. As Hamm explained:

"Transistor amplifiers exhibit a strong component of third harmonic distortion when driven into overload. This harmonic produces a "covered" sound, giving a recording a restricted quality. Alternatively a tube amplifier when overloaded generates a whole spectrum of harmonics. Particularly strong are the second, third, fourth, and fifth overtones, which give a full-bodies "brassy" quality to the sound. The further any amplifier is driven into saturation, the greater the amplitude of the higher harmonics like the seventh, eighth, ninth, etc. These add edge to the sound which the ear translates to loudness information. Overloading an operational amplifier produces such steeply rising edge harmonics that they become objectionable within a 5-dB range. Transistors extend this overload range to about 10 dB and tubes widen it to 20 dB or more. Using this basic analysis, the psychoacoustic characteristics stated in the beginning of this paper can be related to the electrical harmonic properties of each type of amplifier."

In the 1972 article, Hamm noted that significant/audible distortion only occurs when the amp is overloaded - which might lead one to think that at normal listening levels, harmonic distortion isn't an issue and that tube and SS should sound pretty much the same.  However, Hamm points out that transients -- the attack of a note, drum hit, etc. -- often overload the amp, if only momentarily - which suggests (to me) that tube amps should sound more dynamic even at normal (non-overloading) listening levels -- which is something that I find.

I can't claim that this is a definitive explanation of the difference between tube and SS, but it is very interesting. 

I wonder whether through tweaking the amp, it is possible to reduce 2nd harmonics that produce a "covered" sound, while leaving the other harmonics undisturbed?  And would this be a good thing? Or is the 2nd harmonic only detrimental to SQ when out of balance with the other harmonics, such that removing it or lessening it from the full range of tube harmonics would remove some of the body of the sound.  Dunno. But it would a very interesting to experiment with, were it possible to do so.

cheers,

Derek

[Paul Birkeland]

The methods for reducing harmonic distortion in a zero feedback directly heated SET amplifier are generally more detrimental to the sound than they are helpful. 

[Deke609]

Thanks PB. I think I'll stop with the tweaking for a while.  With the impedance switching board, balanced speaker outs, and some signal wire shielding, I've pretty much scratched my tweaking itch. Maybe down the road a bit I'll play with capacitors.

But for now I have the BeePre on order.  I plan to do the flux band mod you discussed on another thread, and go a little crazy with shielding the signal wire with grounded copper and where possible some MCF5 film - just for S & G's. And I'll cover with non-conductive tubing/wrap any shielded section that could possibly come into contact with a component.

[Paul Birkeland]

I'd be a little cautious withe modifications on our higher end kits.  The signal wiring and a fair amount of the AC wiring is shielded already and arranged reliably and out of the way. 

The copper tubing I put in the Neothoriator did double duty as a shield and as a support for a bunch of less than rigid twisted pairs of our solid core Teflon jacketed wire.  In the BeePre, these wires are secured to the chassis and won't move.  These wires also take a gentle curve down the center of the amp that won't be so easily done with copper tubing. 

[Deke609]

Thanks PB. Wiring paths will not be changed - for the main section through the middle, I'll be using refridgerant copper tubing, which is very soft, unlike regular copper plumbing, and I will form it to match the specified curved path.  For trickier spots, I'll use 1/8 or 3/16 OD copper braid, depending on the width of wire to be shielded.  I will be very careful in the build process and take my time. 

cheers,

Derek

[Deke609]

You don't need a preamp with the Stereomour II. The BeePre is super quiet in terms of power supply, but there is the potential issue of microphonics in the directly heated 300Bs if you can't put the preamp somewhere where it will be isolated from external vibration.

OK, the disclaimers are out of the way. Sonically it could be a very nice combination. Let us know when you are ready to order. 

So, the BeePre w/ BeeQuiet has been up and running for a week. Caps and JJ 300Bs are nowhere near to broken in, but from the first switch on it has sounded superb paired with my SII.  Build is stock except for (1) teflon 4 pin sockets, (2) copper shielding around as much of the signal wire as I could manage and some sections of the power wiring (refridgerant copper tubing for long sections, and copper braid everywhere else -- sometimes both), and (3) copper flux bands on the transformers. Flux band and shielding is wired to chassis ground. 

It is still early days, but the change in SQ is dramatic.  Doc is right, the SII doesn't need the BeePre, but boy do they ever get along well. For me, adding the BeePre in front of my SII has already had as much of an impact as the Shunt Reg upgrade (my favorite of the 3 Stereomour upgrades). With the Shunt Reg upgrade added, the SII really started to punch.  With the BeePre, it is punching harder and now has SERIOUS BITE! The dynamics are incredible. This is in addition to greater clarity/detail and "openness".

So far, I mostly prefer to listen with the BeePre wide open. This may be pushing it into distortion - but if so, it is a very euphonic distortion, particularly for rock. For classical and some jazz I turn down the Coarse attenuator by a click.

I am impressed by everything about the BeePre, including the very clever layout and implementation of the schematic.  There's a lot of stuff going on in this amp -- it not only all fits, but does so in a way that makes amateur assembly possible and fun.  Bravo!

cheers,

Derek