Lucky I'm trying to continue my Spiritual Journey of Love.......There was another day, where I could add to this Comment

LOL Yeah how's that rev coming along 

 

I'm not sure you're not trolling, as you're not making any sense. There's a difference between statements and facts. You're not giving any of the latter and lots of the former. I'm giving this one final reply.

 

Let's start with the basics.

THD+N is a direct measure of linearity of a single sine wave delivery. And any other bandlimited wave by the Fourier theorem. IMD is a measure of delivery of two sine waves close in frequency and the resulting modulation products. (These products cause a real life "violation" of Fourier theorem.)

 

Square waves and impulses don't apply, they're not bandlimited, but still, anything stable will not ring badly and almost any amplifier is fast enough. It still happens that someone releases a borderline unstable amplifier, e.g. the AMB Mini3, which shows broken behavior even on my cheap 25 MHz scope; or a DAP with a broken lowpass filter/not fast enough, like aforementioned Hifiman HM-601.

Jitter is a bit more elusive, but it's typically not a problem, with an exception of misdesigned class D amps.

 

---

About soundcards, it highly depends on what we're talking about.

HDA Intel implementation are either hit or miss. (For instance, mine is miss with all BA IEMs I have due to relatively high output impedance. Otherwise excellent quality.)

Many internal soundcards have a real hard time dealing with RF interference and this is what kills the performance. Like hearing that buzzing noise every time you move the mouse or use the hard disk? I don't either. The better HDA implementations tend to avoid the noise by careful placement of the chip and analog section, but you can't do that with a card - it will almost always be near the mainboard chipset.

So there's typically a noticeable quality gap between internal soundcards and sound chips in bad implementations versus sound chips in good implementations and external soundcards/DACs.

 

---

"Higher grade" (whatever that means) parts are relatively unimportant. You can make a good amp out of 80s parts by applying every trick in the book and being careful about component placement. Heck, certain 80s parts still can beat or equal modern ones. (Example: NE5532 which is darn good and cheap. NJM4580 which is really low noise and also cheap.)

 

The only "high end" parts I'd pick is a good electrolytic capacitor for power supply decoupling - so that it doesn't break after a year (that's about $0.3 difference) and a very good potentiometer (otherwise channel imbalance can ensue), unless the amp is to have fixed volume (really cheap) or highly switchable gain. (quite expensive)

Capacitors, inductors and resistors in the audio route don't matter too much unless you use something entirely wrong (e.g. carbon capacitor or a high ESR inductor) or mismatch values badly. (Can happen with certain cheap parts - not really an issue when building a single piece, huge QC issue for mass production.)

[EDIT: I forgot one place where I'd pay for "high end": sockets and jacks. There's nothing more frustrating than those failing.]

 

For example, a simple act of placing volume control at the output instead of input has major benefits for IEM performance by cutting current noise. Similarly trying to keep low impedance values everywhere. (Relatively easy with SMD parts.) Also not mixing power supply paths with signal paths and properly grounding everything. (E.g. keeping digital ground far from analog ground, important for DACs with low PSRR opamps like AD8620.)

 

On the other hand, you can make a lousy amplifier full of buzzwords (e.g. class D monoblock with heavily filtered power supply) as well as RF, jitter and other kinds of distortion. Done that once on a breadboard, got smarter since.

 

---

Feel free to have a blind test of volume matched amplifiers some day. It's relatively easy to set up even, need a switchbox, identical looking cables, something to hide the amplifiers and another person to wire it up. (For single blinding - double blind requires extra careful planning and impersonal means of communicating the end of trial.) You need to decide on a number of trials before the test (minimum 8 for a pair according to blind Bayesian conditional probability) and generate a random wiring list.

 

Then you have to do multiple attempts to label each of the amplifiers (e.g. "FiiO E6", "Leckerton UHA-4", "Teclast X16HD", "FiiO E17") to prove that you can discern them and assign a numeric "quality" value to it.

 

For 8 trials, you need all of them to succeed perfectly. For 12 trials, you need to succeed at minimum 10 for 95% confidence.

 

---

The main issue with FiiO devices is the relatively cheap build quality in places and the fact that their wide assortment has some outright bad devices (E3, E5, E9), mediocre (E10, E11) and some good ones. (E6, E7, E17)

 

Why do you assert that E6 isn't accurate? Care to back up your assertion with some proof?

It measures and sounds quite decently, but is not completely transparent. Has some definitely audible minor RF issue that can be picked by highly sensitive IEMs. (buzzy noise floor, but very quiet)The slight THD+N lack of performance is almost inaudible, but there.

All this is only noticeable when directly compared to an amplifier that's ~5x more expensive. (For example, E17, Prodigy Cube's output.)

 

The main issue is that E6 is relatively sensitive to RF noise, e.g. "cellphone" blips and that FiiO has some QC issues with the volume control, causing tracking and channel balance issues in some models. It's better than some of the more exotic CMoys out there in fact.

 

I didn't know about Teclast [X19HD specifically], but from what I could find, it's does not seem clearly superior to E6 and has even lower output power. Might be too quiet with even some relatively easy to drive headphones as E6 is already marginal. (50 mW at 32 Ohm in X19HD; 60 mW at 32 Ohm in E6)

 

E6 is quite a bit stronger than what you have in many smartphones, but not a class better quality. You can also use it to reduce output impedance if it's too high. You shouldn't buy it to improve output quality in general, unless the source has lots of common mode noise.

 

(HTC and LG are the worst offenders on impedance due to capacitive coupled outputs. Motorola and certain Sony models have relatively high noise which is cleaned up nicely by an amp. Samsung, iPhone and other Sony are good regardless of amplification.)

Edited by AstralStorm - 7/10/12 at 7:55pm

I'm un-subbing from this thread. It's turned into anything but information about GRO7 MKII. Honestly, people are very interested in the GRO7 MKII, information about it or the cable for a large majority of the posts is anything but.. If you two want to debate other topics go and get a room somewhere else.

Nice one guys.

In my experience, reterminating a stock cable of cheap IEM's is pretty much a nightmare.

 

Hey Kiteki, any hints or tips to try with the GR07 MkII? I'm getting one fairly soon! 

Yeah, I dont think that any cable is pure silver, its almost costly forbiden and doesnt improve that much from the good ol' cooper. Plus I think that silver as gold cant handle too much mechanical stress.

Well K isn't really a roman numeral... 

 

Hmm, that's something I'll try. Did you find that it played well with any type of amp (e.g.  analytical ones, or warmer ones)? I'm definitely going to pick a new one up in the near future.

Pure silver isn't that expensive at quantities used in a cable. That said, there are better silver alloys for cables if you *really* want to use one.

(It's pointless at audio frequencies.)

 

Kiteki, you're raising and misrepresenting points randomly, this is called trolling.

 

DT880 32 Ohm vs 250 Ohm vs 600 Ohm are same headphone. The difference is extremely minor, in damping.

http://www.innerfidelity.com/content/comparison-beyerdynamic-dt-880-32-ohm-dt-880-250-ohm-and-dt-880-600-ohm-headphones

 

If these were dynamic driver IEMs, the ringing would be even less affected, as the excursion length is tiny.

GR07 (mk2 or not) is a dynamic driver and a small one at that. (but not an undampened microdriver like in RE272)

 

Moving armature and especially BA IEMs might be more dependent on electrical damping (there are versions with ferrofluids for mechanical damping, as normal BA is quite resonant), so I'm not so sure about those.

Edited by AstralStorm - 7/10/12 at 9:47pm