Low impedance phones, Please suggest amp design?

noobman92

Member
2010-11-26 7:33 pm
I own a pair of Low Impedance IEMs, the RE272s, rated at 20Ohms at 1KHz.
I recently read somewhere that Headphone impedance is different at different frequencies, which i could infer from manufacturers rating Headphone impedance at one particular frequency. Well, how much the impedance varied with frequency, i did not have an idea of that.

Quoting NwAvguy at NwAvGuy: Headphone Impedance Explained

"YOUR IMPEDANCE MAY VARY: Virtually all passive headphones (without their own powered electronics) have a varying impedance that changes with frequency. As described above, they don't behave like a simple resistance when driven with typical audio signals. The gold colored graph below shows the impedance, in ohms, of the Ultimate Ears SuperFi 5 Pro in ear monitors. You can see they’re only at their rated 21 ohm impedance below about 200 hz. The impedance rises to nearly 90 ohms at about 1200 hz and drops to below 10 ohms at 11,000 hz"

This scares me a bit. I have the 272s at 20Ohms, and I imagine their impedance varying from 10ohms to 30-40ohms maybe.

Can i have an amplifier designed for the current needs of my IEMs? Would it actually help me in getting better sound? i was looking at
http://www.ti.com/lit/an/sboa031/sboa031.pdf

But they list the nominal impedance as 40-300Ohms in this design. What else should i look at?

The aim of the amplifier is going to be able to drive the IEMs effortlessly, as in, being able to have all the current the IEMs can ever require.

Regards
 
Loudspeakers and earphones have a mechanical diaphram that is made to swing back and forth by an electromagnet coil. The diaphram and coil is a mass that is allowed to move against a restoring force of a springlike material.

If you put a mass on a spring it will vibrate best at its natural frequency. Also, at its natural frequency, the coil produces the most back emf. This back emf will hold back any current the amplifier attempts to push into the headphone.

There is nothing you can do about this. The headphone designer has worked hard to keep this effect at a minimum by controlling a sort of friction on the springlike motion (also called damping).

The only thing you need to be concerned about is that any amplifier will give enough current at the lowest impedance. Then at the higher impedance, the current the amp will provide is more than enough. The higher impedance (or changing) impedance need not scare you.

Feeding a lower impedance into a higher one is not a problem. Feeding a higher impedance source into a lower impedance will result in current limiting, which may be a problem.

Total frequency response and distortion are generally much more important to concern yourself about.

Also, a peak in the response curve is more annoying than a dip of similar size.

Most opamps, being very small, do not give enough current for a 30 ohm earphone.

You are looking for a suggestion.

Look around in this forum for headphone amps with larger discrete output devices. Many of them are excellent. They will give plenty of current for any headphones.

Better yet. Build two different ones, and report on your results. Learning is life.
 

noobman92

Member
2010-11-26 7:33 pm
The only thing you need to be concerned about is that any amplifier will give enough current at the lowest impedance. Then at the higher impedance, the current the amp will provide is more than enough. The higher impedance (or changing) impedance need not scare you.
That is exactly what I was talking about in the post, its the Current that the IEMs get at Low impedance


Total frequency response and distortion are generally much more important to concern yourself about.
Does it mean that i should not really be worried about the IEM getting enough current or not? Also, To keep Distortions at the lowest, an amplifier should be able to provide all the current required by the IEM easily right? as in, the max current requirements of the IEM should fall under moderate operating range of the amplifier? Did i understand you correctly?


Most opamps, being very small, do not give enough current for a 30 ohm earphone.

You are looking for a suggestion.

Look around in this forum for headphone amps with larger discrete output devices. Many of them are excellent. They will give plenty of current for any headphones.

Better yet. Build two different ones, and report on your results. Learning is life.

I should be looking into Discrete amplifiers? or specifically amplifiers with discrete output stages? i have never handled a Discrete amps before, can you please provide a link maybe if its not too much to ask for?

Regards
 

noobman92

Member
2010-11-26 7:33 pm
@Avro Arrow,

I cant seem to find enough info on that particular amplifier.

Also, can you explain why it can do what i am looking for?

Also, i need to ask and don't know if i should be starting a new thread for this,
Why exactly cannot OPAMPs drive low impedance phones properly? In all the datasheets of OPAMPs, i have never read the nominal impedance of the speaker to be less than 30 ohms. Why is it that way? I know for sure that low impedance phones mess up an amplifier's graphs, as in Distortion levels, TDH+N and others. Why is it that way? Is it because they demand so much current? i calculated a little bit here.

My RE272s are rated at 20Ohms. I assume that the lowest impedance they get at is 10Ohms. Max rated input is 30mW, Rated input is 10mW. So, i assume i want the volume at 15mW.

SO
15mW= R x I^2 (assuming purely resistive load, tell me if i am wrong here)
15mW=10x I^2
I^2 = 15mW/10
I^2 = 0.0015
I = 38.72mA

38mA is the typical output current that a NE5532 can provide.

Why then, do OPAMPs have problems driving low impedance loads??

Regards
 
@Do you know of any OPAMPs which have relatively low distortion at Low Impedance? What is the lowest impedance OPAMPs are ok in driving with typical distortion in them?

Regards

You can chip amp that design for headphone amplifier. Many chip amp is capable to use split power supply and work exactly like an OP-AMP.
 

jcx

Member
2003-02-17 7:38 pm
..
the iem are pretty sensitive - 640 mV, 32 mA peak will drive them to 115 dB SPL

because of the low V that is only 10 mW_rms

I like to point out that "the best" interface to such highly sensitive, low impedance devices is an audio bandwidth output step-down transformer - but "audiophile quality" transformers could cost US$50 each
which isn't so bad compared to $250 headphones - and its much harder to toast your pricey iem with xmfr output naturally blocking DC faults

using direct drive, common diy project amps, toasting the iem will be an issue
 
The first reply to the original post directed me towards Discrete amplifiers. Do you think i might have some luck there? From all the Datasheets i have read, of OPAMPs and Power Amps, i have found that OPAMPs just do not support low impedance and have high distortion at these loads. For Chip amps,the best i found was TI TPA6111A2 which had 0.005% THD+N at 1Khz and less than 0.5% through the Audio region. 0.5% is too high right?? The good thing about the chip amp was that they can easily drive low impedance loads. Problem is high distortion.

That makes me ask the question again, For OPAMPs, low impedance loads are a problem. Is it that OPAMPs cannot provide enough current and therefore there is distortion.

-OR-

OPAMPs cannot provide current and have high distortion -BY NATURE- at low impedance loads??

@jcx,
Can you provide a link to one of the Transformers? and also, is distortion fairly low on the se transformers??

also, do you think discrete design might help lowering the Distortion and keep the current high?

Regards
 
Yes.

But you can use a high quality opamp with a current buffer in the feedback loop to get higher current without any noticeable increase in distortion.

Something like this?
 

Attachments

  • Dragonfly v2.png
    Dragonfly v2.png
    735.6 KB · Views: 162
Looks an awful lot like a The Wire...

Is that bad, is that good, is that wrong or what is then?

Looks fine to me.

It looks more like:
a high quality opamp with a current buffer in the feedback loop to get higher current without any noticeable increase in distortion.

Or is "a high quality opamp with a current buffer in the feedback loop to get higher current without any noticeable increase in distortion." a "The Wire" when you use a lme 49990 and lme 49600?

:)


Gr GJ
 
Last edited:
Is that bad, is that good, is that wrong or what is then?

The Wire by OPC is a no compromise design that he spent a large amount of effort on and that has measurements that nothing else compares to (and isn't crazy expensive either). It is capable of driving down to 8 ohm loads.

Boards don't appear to be available at the moment because he's busy designing the next revision, which promises to be even better. You could always design a copy of it, try your hand a board design. It might not measure as well, but it shouldn't be too hard to make something that sounds great. I made some for myself with a couple of compromises to make things simpler (such as one board with power supply and two channel amp), but I would never sell them because that's just disrespectful.

"The Wire" Headphone Amp Build Wiki - diyAudio

http://www.diyaudio.com/forums/head...igh-performance-headphone-amplifier-pcbs.html
 
RE272s were measured as follows: 21 ohms @ 1 kHz (quite constant actually), 29 mVrms or 0.04 mW for 90 dB SPL. That's ~131 dB / 1 V, or ~104 dB / mW.

Aside from their lowish impedance, they are not particularly hard to drive. Actually noise performance should be of at least equal concern. Keep noise across the audible range under 10 µV or so, preferably a fair bit. (That already rules out, say, the FiiO E6 and E11.)

Note that "The Wire" merely is a headphone driver module, intended to provide a headphone output inside an existing amplifier or whatever. It does not constitute a complete self-contained headphone amp. Input impedance also is a bit on the low side for direct connection of a volume pot, you'd rather be driving this one with a preamp of sorts.

I'd look at an O2, arguably the most prominent DIY amp when it comes to low noise (almost 20 dB quieter than the E6, and among the very quietest headphone amps full stop). A configuration with low gain = unity and high gain = 2.5 ought to be just right. This one is a self-contained amp.
 
Last edited: