Vsonic GR07 MK2 - stunning new cable! | bio-cel tech IEM | now with review - see 2nd post - Page 6

Quote:

Originally Posted by goodvibes 

Everything there is correct but not typical or necessarily significant for the argument of dynamic vs BA. The TF10 is extremely unique to drop below 1 ohm in the highs and using a 22 ohm amp designed for high impedance cans for demonstration purposes isn't real world.

I wish it was true - many portable amplifiers have on the order of 10 Ohms, certain (e.g. Hifiman's except 90x) have 32 Ohms.

Many (I'd be careful with "most") DAPs have low output impedance - otherwise they'd have even more current limits at low frequencies. However quite a few have severe limitations of maximum output current which will cause bass rolloff.

It is typical for a single BA to have sharply increasing impedance at high frequencies. TF10 is not that unique in the multi-armature world, but it is the more rare shape of the curve.

It is typical for dynamic driver and orthodynamic headphones and IEMs to have very flat impedance curve. (Within 5%)

If you think otherwise, please provide valid sources.

This assertion is wrong. Output impedance (of the amplifier) should be linear within audio spectrum - otherwise it will act as a filter - which can happen when someone uses a misdesigned lowpass.

What is visible on the graphs is lack of current capability in J3 and YP-R0.

Sony A845 has too high output impedance - the slight variations in the impedance curve of the dynamic driver are visible in frequency response. Note they're within 0.5 dB, so still hardly audible. If you used Sony A845 with a BA IEM of that low impedance, it could have massively altered frequency response.

Again, if you think otherwise, please provide valid sources. Also, please try harder to write proper English.

Quote:

Originally Posted by kiteki 

 

 

Anyway, when I said I'm excited about the new cable, I meant the looks and pliability, that's all. ^^

Ah, that's another story..... I thought we lost you to the dark side.  I think the "old" GR07 cable is pretty slick (figuratively and literally) , the only silver cable I have. Pictures of MKII look similar except you can see the braids.

Old cable is excellent if slightly stiff and springy which made it paramount to remove all "axial coil" from it. Otherwise it tended to escape from behind ears if not using the ear guides. The splitter also could've used some slight improvement - it moved too easily.

I hope these very minor issues have been resolved. The simplest fix would be to add a bit of memory wire near the ears.

We disagree so now it's the grammar police? Really? I'm happy to rephrase if asked without an insult.

HiFiman?  How many do they sell? We know every current Apple, Sony, Sansa and Cowon has below a 5 ohm output impedance and I'm not sure any of them exceed 1/2 that. Pretty much qualifies for most. That would include most phones though here you can find a couple models that will have a higher impedance. Check your own link to qualify most.

It's not like I'm unaware of what you're describing. I still own an Ipod touch 3 and it's highish output impedance doesn't audibly affect the highs of my Grados or Phonak that have rising impedance's but I do notice a bit of roll in the low bass where the impedance is quite linear to the mids. Something you don't see in the next pic.

Not that it affects my statement but I suspect that the output impedance of the HiFiman products are lower than the derived measurements would indicate due to interpretation of the built in high frequency roll. RMAA results are often a bit off because much is calculated. I'm not saying they're worthless but that they sometimes need a bit of scrutiny.

The low frequency response into a 16 ohm load would indicate a lower output impedance. The highs are a function of the low phase shift (shallow) passive analog DAC filtering built into the circuit ahead of the amp. You'll find the same frequency response characteristics on a Hm-602 from the line out.  http://www.stereophile.com/content/head-direct-hifiman-hm-602-digital-audio-player-measurements

No one clipped an amp during frequency response tests. It's not current limiting. It's the transfer function.

I wasn't asserting anything about varying output impedance. The 2 ohm output impedance of the 845 was assumed constant to frequency and that it would have a negligible effect on a single driver BA's high frequency response. The 2 ohm output impedance is not high enough to affect the high frequencies. Why would you assume I was saying anything else?

We can take this to PM if you'd like to discuss it in a friendly manner. Probably don't want to take this thread any further off track.

Quote:

Originally Posted by goodvibes 

Stuff. Sometimes right, sometimes wrong.

Sony A845 is 35 Ohms, not 2 Ohms. That's why even with a dynamic driver there are ripples - you'd need around 100 Ohm impedance to smooth this out and this DAP doesn't have enough power to do so.

--

Low frequency response into the load indicates nothing in particular other than the source having enough current to provide enough power at low frequencies.

Lack of current in the output stage will typically not show as clipping - just the amp providing less gain as the load impendance falls, looking like a bass rolloff and slight THD+N increase.

It is *correlated* with higher output impedance, as more power is lost on the impedance. It's very different issue from straight voltage clipping. Even an amp with huge output impedance can have linear bottom end if it has lots of current capability.

In short, a voltage mode amplifier with high output impedance will have the low end rolloff. Current mode amplifier with high output impedance will likely not be rolled off.

HM-601 is (AFAIR) 38 Ohm impedance. You can see the peak at 3.5 kHz due to it. (It's being tested with an easy load of almost constant impedance, you'd see more dramatic peaking with many low impedance BA IEMs. E.g. ER-4P, you'd get a high boost relative to flat impendance - an U shape in the highest end.)

--

The difference in power handling equired is about 4x for e.g. ER-4P and many amps (current mode) actually can deliver less power with increasing impedance.

For decent loudness and dynamics, ER-4P needs ~2mW into 22 Ohm at 1kHz (it's 109 dB SPL/mW), this means it requires ~24 mW into 100 Ohm at 20 kHz. (if the square root for sensitivity holds, this isn't always true in a mechanical system like a balanced armature - it might need more than that.)

Some cheap amps used in cheap portable devices might not provide enough power at this impedance. Most portable headphone amplifiers will have no problems.

The better DAPs won't have an issue either, e.g. Clip+ and iPod.

--

I agree we should take this to another thread but not in private, e.g. to the Sound Science forum.

Your 1st link shows a sony A844 with the same output section at under 4 ohms. It reflects the difference between RMAA and real equipment testing. I don't trust RMAA for certain measuements. Under 5 ohms is not significant for the issue discussed. You should also be aware the even the 4p has 20 ohms of series resistance so it's also not a normal representation. You're saying it has these characteristics as a BA when 20 ohms os the load is already a resistor. The TF10 and 4P are good examples for the link but poor ones to use with a 20 ohm output impedance as examples for normal situations, at least for me. We may have found out why there is such a varied opinion on TF10s. With a good amp, they're more U shaped. We can agree to disagree but on single driver rising impedance iems, I don't believe that there's any issue under normal conditions.  Amps are not clipping in these tests and while listening to music, the energy over 5k is an extremely small fraction of what's being used. It would also show up as massive distortion in test. To go further, I want nothing to do with any portable device that has over a 10 ohm output impedance unless I have a special use. Yuk! I also wanted to mention that I read a graph wrong and the TF10 doesn't drop to under an ohm. It drops to about 7 ohms which will still cause issues but it's not in the ridiculous range. For some reason I saw .8 instead of 8. 

Back on topic, there's no way that you could find a cable characteristic that would influence the measured frequency response unless additional parts were added. That's not to say you can't hear a different cable but it wouldn't be something discoverable under normal type testing. Some of what we discussed is also a good reason to not add bits as load since the results may vary.

Wanted to post that earlier but couldnt on my mobile.

They are pliable and fit comfortably on your ears. They come with a little bend near the housings and seem to maintain that shape. Other than the area around the housings the cable just flops down straight as if there was no memory. The chin slider helps to keep the wires in place. I needed to pull them up a little higher than chest level.

@h20

agreed. I found the ear guides a bit too big and clunky tbh. You won't need them dapple.

Quote:

Originally Posted by kiteki 

Long story short there is a lot more to an amplifier than output impedance, yes?

Output impedance tends to be most noticeable.

There's obviously noise performance, voltage gain and current capability. THD+N, IMD and jitter as of nonlinearities, but it's a really lousy amplifier that fails there horribly.

And in case of portables, battery life.

About FiiO E6, it is a decent amp if you don't require lots of drive power - ok for IEMs and many headphones. Not strong enough to drive orthodynamics or high impedance versions.

(But definitely can drive something 80 Ohm and reasonable sensitivity like Beyerdynamic DT1350, unlike many portable sources.)

It has pretty short battery life. Not sure if anything can beat it at its price level - it can definitely beat many amplifiers built into cheaper DAPs and cellphones and due to impedance issues, even some soundcards.