Audeze LCD 4 - Which amp?

[Deke609]

Many thanks again Paul.  The damping vs distortion issue is very helpful. I'm doing a bit of self study about these and other issues and hope to get back to you in a few weeks with what I hope at least will be the right questions. 

By the sound of things, finding the right damping/distortion balance is matter of "try and listen" experimentation.  But it also sounds like that, with enough experimentation, I should be able to get the amp nicely dialed in for the LCD 4s. My current working idea is to create a box with stepped attenuators that would plug into the binding posts and then allow connection of the headphones - so that the load at the output could be adjusted to see what sounds best.  I think this might be a bit of construction challenge as the resistors for the attenuators would probably need to be rated for 4W or so to cover off the 3.5Wpc? and would therefore be quite large and possibly expensive.  But the ability to switch loads would allow for quick A/B comparisons.  I'd try to get less expensive (and consequently noisy) resistors for the attenuators. Once I find R values that worked best for me, I assume I should get some low noise resistors (e.g., Vishay naked bulk foil?) for the final implementation?  Does that make sense? 

many thanks in advance,

Derek

[Paul Birkeland]

I would try wiring the amp for 2 ohm output and driving the headphones straight from the speaker jacks first.  There's no need to spend money on a problem that you aren't certain you will have.

[Paul Joppa]

I see that PB posted while I was writing. He's right; comparing the 2-ohm and 8-ohm outputs will give you two data points on the power/distortion tradeoff, at essentially zero cost, and you are likely to be happy with either. The rest of this post is in case you (or someone else reading this) wants to go into more depth to explore these variables.

You can get adjustable L-pads for loudspeaker crossovers. They are not expensive, handle plenty of power, and have wirewound resistance elements (the good kind!). If you load the output of an 8-ohm L-pad with an 8-ohm power resistor (also widely available and inexpensive), it will present an 8-ohm load to the power amp no matter the setting, and still provide an adjustable signal to the headphones (which are in parallel with the 8-ohm load resistor). As long as the phone impedance is much larger than 8 ohms, it will be dominated by the paralleled 8-ohm resistor.

This way, the damping will be excellent no matter the setting, and you can explore the distortion vs. power tradeoff in a fairly clean experiment.

Once you know your sweet spot for power/distortion, you can separately explore the damping issue. We can get into that another time - there are actually two in the case where an output transformer is involved: 1) headphone damping affected by source output impedance, and 2) parafeed circuit resonance affected by amp load impedance. The latter is usually a smaller effect, and often can be ignored, but if you want to explore all the relevant variables it might be interesting.

I recommend this approach because the results are more easily perceived, understood, and communicated.

[Deke609]

Many thanks to you both.  I agree that it makes no sense to problem-solve a problem that may not even materialize.  But curiosity is getting the better of me and I "just want to see what happens if ..."  This is all quite new and fascinating to me.  I've tried numerous times over the years to learn the basic principles of electronics from texts that taught the principles in the abstract or only provided boring experimental applications -- in both cases, that part of my brain that asks "Who cares? So what?" would kick in and I'd close the books after a short time. I've probably relearned "how to calculate voltage between two resistors" at least 10 times by now and never cared, and it wasn't until last night that I had an "Aha" moment when I saw that a volume pot acts as a variable voltage divider - the simple beauty of it! ... So maybe I'm getting the bug.

Regarding the premade L-pads for power amps - that is very helpful.  I will look into that.  My idea of putting together large step attenuators was my attempt to solve the (perceived) problem of how to determine the required resistance values (they could be read from the position of the attenuator). I came across a lot of discussion about the use of L-pads for guitar amps, and assumed that such devices would use a potentiometer and that this would making it difficult to determine the "good sounding" resistor values b/c I'd have to estimate the position of the knob as a point on the resistance divider range of the pot - but maybe that's not an issue.  I'll look into it.  But perhaps you (Paul J.) are suggesting that the L-pad could be part of the "permanent solution" -- i.e., that it would be good enough sonically and not need to be replaced by a fixed value resistor setup?

I had also planned to ask you (Paul J.) about capacitors, as I came across an earlier post of yours that mentioned tweaking "parafeed capacitors" and I did some initial online reading about that last night.  But I will wait until I understand more about it before asking questions.

Again, many thanks to you both for having taken the time to respond so helpfully. I really appreciate it.  I'm not sure what I'm looking forward to more: good sounding music or the learning/experimentation   But why choose between them?  Hopefully I can have both (barring a major screw-up that melts my amp or blows up my cans).

[Deke609]

Oops. I misunderstood the L-pad point.  I think you are saying that it presents a fixed and not variable resistance load to amp - e.g., so that on the 8 ohm tap the amp "sees" an 8 ohm load.  I had in mind a device (perhaps non-existent) that could vary the load that the amp sees.

[Paul Joppa]

In my post, I mentioned exploring the damping issues - one of them is the effect of changing the load on the amp. It's rather involved, and as I said often a small enough effect to be ignored - or postponed. In any case, a different thing from the power/distortion balance question.

[Deke609]

Thanks Paul J. I think I understand now: the L-pad is to take the load issue out of the equation so that power/distortion can be cleanly experimented with.   Once that is done, I can return to the load issue if I want. 

[Doc B.]

I would hope that if you are going to explore these small issues you plan to have some equipment besides your ears to measure the results. I would suggest getting some measurement software that will work with a sound card so you can quantify frequency response and distortion. That will help a lot towards understanding what is really going on when you change the loading.

[Deke609]

Ha! Thanks Doc.  I hadn't thought past my ears, but that makes obvious sense (obvious, now that you've pointed it out to me).  Any recommendations?

[Deke609]

OK. I've made some small progress in preparing for the load and power/distortion experimentation. Following Doc's suggestion that I get some test equipment, I've ordered a USB oscilloscope (Pico 2204A), together with banana plug terminated connectors, and a basic audio signal generator (B&K Precision 3001), for a total cost of about $250 excl. shipping. I hope this will be adequate for my limited testing purposes. (Reason for the B&K: a reviewer of the scope indicated that the built-in audio generator isn't so good for hifi amplifier testing and suggested getting an external one).

I've also looked into L-pad attenuators and have some questions.

(1) Which one(s) to get?

Parts Express sells 8 ohm speaker L-pads in either mono or stereo, and with 15W, 50W or 100W ratings.  Regarding mono vs. stereo: I assume that a stereo L-pad will allow me to connect both right and left speaker outputs to it - so that the single pot will attenuate both right and left earphones simultaneously and more-or-less evenly. Is this correct?  In which case I need only 1 L-pad?

Regarding wattage rating: Which is likely to be sonically better, if any? The price difference between them is small, so I'll get the 100W if that is likely to sound/work best.

Link to the Parts Express L-pad page: https://www.parts-express.com/cat/speaker-l-pads/306

(2) Implementation: 8 ohm L-pad(s) PLUS 8 ohm power resistor(s), or just 8 ohm L-pad(s)?

I am confused by Paul J's suggestion, quoted below, which I took to mean that I need BOTH an 8 ohm L-pad and an 8 ohm power resistor (or possibly two power resistors - one for each pair of speaker outputs?):

You can get adjustable L-pads for loudspeaker crossovers. They are not expensive, handle plenty of power, and have wirewound resistance elements (the good kind!). If you load the output of an 8-ohm L-pad with an 8-ohm power resistor (also widely available and inexpensive), it will present an 8-ohm load to the power amp no matter the setting, and still provide an adjustable signal to the headphones (which are in parallel with the 8-ohm load resistor).

I had assumed that the 8 ohm L-pad would take care of presenting the amp with a constant 8 ohm load and that nothing else would be required to do so.  I'm not questioning the suggestion - I just want to make sure that I've interpreting it correctly and order the right stuff. So I need BOTH the L-pad and an 8 ohm power resistor?

So my questions w/r/t L-pads: How many L-pads (1 or 2), mono or stereo, and at what wattage rating?  How many power resistors, if any, (1 or 2)?

Testing -- Just to be sure I'm understanding this right, once I find the sweet spot on the L-pad pot for the power/distortion issue, I can determine the corresponding impedance value of the L-pad by (a) cutting the power, and (b) leaving the pot turned to the sweet spot, measuring the resistance between two terminals of the L-pad pot?  I.e., the L-pad would have two resistors values - one fixed (8 ohms) and the other adjustable by turning the knob (X) - and I find X by following the above procedure.


Many thanks in advance,

Derek

[edited to add missing word]

[Paul Birkeland]

It's pretty easy and inexpensive to just use ARTA and a soundcard (the soundcard will act as both the generator and the signal measuring device) for your distortion measurements, but it looks like the Picoscope has an FFT.  For years I got away with using a $40 M-audio Fast Track soundcard, but I recently splurged on an Mtrack Plus MKII to get some extra bandwidth.

Yes to the stereo L-pad, that is the way to go.

I would not pay the extra money for an L-pad rated for higher power.

Yes, you need an 8 ohm load.  The L-pad will present an 8 ohm load to your amplifier provided the L-pad itself is loaded with 8 ohms.

As far as finding the sweet spot, if you do your measurements at 100Hz, you can easily measure the AC voltage coming out of the amp and coming out of the L-pad with your DVM, then you would have a good idea of the fixed resistor values you might want to buy.

[Deke609]

Awesome.  Many thanks Paul B.

Just a quick follow-up question: do I need (a) 1 or 2 L-pads? and (b) 1 or 2 8 ohm power resistors?  E.g., do i need one L-pad/power-resistor combo for the right channel and a second for the left (2x2)?

[Paul Birkeland]

Awesome.  Many thanks Paul B.

Just a quick follow-up question: do I need (a) 1 or 2 L-pads?

One stereo or two mono L-pads.

and (b) 1 or 2 8 ohm power resistors?

Two 8 ohm resistors.

[Deke609]

Perfect.  Many thanks Paul B.

[Deke609]

I just got the my SII up and running in stock configuration. That was a bit more involved than putting together the Crack! Only screwed up one thing (installed a 5W 3K instead of a 5W 2K -- but caught that mistake and fixed it). Passed all resistance and voltage checks. Fired her up and no smoke. Woohoo!

Following previous advice, I plan to compare the amp configured at 2 and then 8 ohms with my Audeze LCD 4s plugged directly into the speaker outs. It is currently configured for 2 ohms.

Question: should I install the 3 upgrades before comparing 2 and 8 ohm configurations, or compare 2 and 8 ohms first?

Observations and first impressions with zero burn-in and cold amp configured for 2 ohms (Roon -> Yggdrasil DAC -> SII -> LCD4):

(1) Slight hum when no music is playing - but mourquiet upgrade should minimize that and perhaps even eliminate it. 

(2) With the volume pot ranging from 7 (no sound) to 5 (max) o'clock, my normal listening level is between 2 and 3. I expect this will change when I try the amp configured for 8 ohms -- I assume I'll get a big jump in gain from this.

(3) Overall sound presentation is remarkable for an amp with no burn in or warm-up - Marvin Gaye's album, What's Going On, sounds wonderful: "human" and spacious.

(4) Bass is muffled -- like someone strapped a pillow to the bass drum: "whumphh".  Rush's "Tom Sawyer" lacks punch.

(5) Details are there if I listen for them, but a bit veiled -- opening guitar on "Roundabout" from Yes' Fragile Album missing the micro details and clarity that make you believe you're in the recording studio with the band.

I expect installing the upgrades and switching to 8 ohms (or possibly 4 if 8 goes too far in the other direction) will take care of (4) and (5) above.  That, and letting the amp warm up and burn-in 

If the 3 upgrades are even half as good as others have described them on this forum, the SII and LCD 4s are going to be a helluva match!

Liking what I'm hearing so far!

Derek