[twelvebears]

Is it just me, or are amp makers that officially state the output impedance for their products in the minority?
 
I know it's not the be-all, end-all, but making sure the output/load impedance ratio is at least 1/8 is quite important to make sure that the frequency response of the 'phones isn't unduly affected.
 
This can still be a factor with full-size cans which have low impedance, even when they aren't power-monsters (e.g. AKG K701s), but is particularly important and relevant with portable amps which may well be used with low impedance IEMs, especially those which use balanced armature drivers.
 
Amp makers often quote compatible ranges but that is that enough?
 
To give a specific example, the Audinist HUD-MX1 quotes 16-300ohms, but the output impedance is actually 5ohms.  So for 16ohm phones, ideally the output impedance should be a max of 2ohms, preferably less.
 
This only came to light for me because I chose my ACS T1s (17ohms) for their even-handiness across the range and didn't want the amp to start changing this too much.  They are my only phones so I needed something I could use either at home or mobile which would be a good match.
 
Interestingly a Fiio E7 has a commendably low output impedance of about 0.2ohms, so is a great match for IEMs, but their own matching E9 desk-top unit has a much higher figure.  In fact the figures for the 1/4" and 1/8" jacks on the E9 are different and the smaller jack is actually higher, not lower, which seems counter intuitive given that if you had IEMs, you'd naturally use the 1/8" socket.
 
Of course this doesn't mean these amps can't sound great with low impedance phones, but it may mean that your favorite 'phones will sound different than you might expect just because of this mis-match.
 
 

 

[skip129]

I agree.
 
Just a word of caution though - it appears that Beyerdynamic design most of their headphones to be used with a high output impedance amp (120 ohms), according to standard [size=small]DIN-EN 60268-7, which specifies that the output impedance should be 120 ohms.  I contacted them, and they confirmed this. Their own headphone amp (the A1) has an output impedance of 100 ohms, which is further evidence of this. However, I have heard second hand from someone who said he spoke to a senior person at Beyer, and that Beyer person told him that they design their low impedance phones assuming a very low output impedance. Very confusing, especially given that the DT 48 E has an impedance of 25 ohms, and it actually states in the specs that the "nominal source impedance" is 120 ohms!  Anyway, if I were to buy some Beyer headphones, I think I'd at least try adding impedance to bring the total output impedance (internal + external) to about 120 ohms, to see whether it improved the sound.  My main headphones at the moment are AKG K601, and AKG said that the ideal output impedance should be very low - as close to zero ohms as possible.  [/size]
 
[size=small]Also, the Etymotics Research ER4S transducer is optimised for a 75 ohm source impedance, but they INCLUDE that 75 ohms in the cable, which means that the ER4S product per se is optimised for a very low impedance amp. The portable version, which uses the same transducer, is the ER4P, and they simply omit the 75 ohms resistors, to make it easier for portable devices to drive it. However, the downside of this is that the frequency response of the ER4P is not quite as accurate, and so if you want your ER4P to be more accurate, you can actually insert a pair of 75 ohm resistors between your amp and the transducer, if your amp has enough voltage output. Etymotics sell this adaptor themselves. Doing this essentially converts the ER4P to an ER4S.[/size]
 
Greg.

 

[maverickronin]

I don't know they most manufacturers don't even bother tell us such a simple figure and why they don't optimize it.
 
I'd guess its some combination of several factors.  Some probably don't know or don't care.  Others may just want to make their gear sound "different".  Many times its just a bandaid for another problem like short circuit protection or because it can't take a low Z load.

 

[twelvebears]

 
Quote:

I agree.
 
Just a word of caution though - it appears that Beyerdynamic design most of their headphones to be used with a high output impedance amp (120 ohms), according to standard [size=small]DIN-EN 60268-7, which specifies that the output impedance should be 120 ohms.  I contacted them, and they confirmed this. Their own headphone amp (the A1) has an output impedance of 100 ohms, which is further evidence of this. However, I have heard second hand from someone who said he spoke to a senior person at Beyer, and that Beyer person told him that they design their low impedance phones assuming a very low output impedance. Very confusing, especially given that the DT 48 E has an impedance of 25 ohms, and it actually states in the specs that the "nominal source impedance" is 120 ohms!  Anyway, if I were to buy some Beyer headphones, I think I'd at least try adding impedance to bring the total output impedance (internal + external) to about 120 ohms, to see whether it improved the sound.  My main headphones at the moment are AKG K601, and AKG said that the ideal output impedance should be very low - as close to zero ohms as possible.  [/size]
 
[size=small]Also, the Etymotics Research ER4S transducer is optimised for a 75 ohm source impedance, but they INCLUDE that 75 ohms in the cable, which means that the ER4S product per se is optimised for a very low impedance amp. The portable version, which uses the same transducer, is the ER4P, and they simply omit the 75 ohms resistors, to make it easier for portable devices to drive it. However, the downside of this is that the frequency response of the ER4P is not quite as accurate, and so if you want your ER4P to be more accurate, you can actually insert a pair of 75 ohm resistors between your amp and the transducer, if your amp has enough voltage output. Etymotics sell this adaptor themselves. Doing this essentially converts the ER4P to an ER4S.[/size]
 
Greg.

Thanks Greg.
 
This just demonstrates what nonsense this all is.  Beyer's approach to using the 120 Ohm 'standard' that was never really a standard in any way that means anything (i.e. actually being adopted consistently), is very odd indeed.  And the 100 output impedance of their amp means you'd probably best avoid unless you're using Beyer 'phones designed for it!
 
The most reliable and simple approach would be for amp and 'phone designers to work around the basis of a zero, or near zero output impedance, that way any 'phones can be used (if the amp can swing enough voltage for high impedance 'phones), and also accounts for the need for low impedance 'phones for mobile devices due to the low voltages they use and the efficiency they need.
 
Worth noting that most high-end headphone amps (apart from Beyer it seems) usually have near-zero output impedance, which means the   interaction between amp and 'phones won't influence the frequency response, which should surely be the point?
 
Commendable that Fiio seem to have cracked this.  I believe that pretty much everything they make (apart from the E9) has an output impedance of less than 1 Ohm, so basically good with anything subject to sensitivity and volume requirements.
 
All this only came to light for me after I noticed a vast difference between the sound of my ACS IEMs on my lad's 2yr old iPod Touch vs a friends brand new model.
 
It seems that the one significant difference between them, is that the new iPod Touch and iPhones have a much lower output impedance, but you'd never be able to find that out from anything Apple publishes.... 
 

 

[skip129]

I have read somewhere (in that same standard, I think) that one reason the 120 ohm output impedance is recommended is simply to reduce the discrepancy in volume between headphones of different impedances. (which it would). It went on to say that headphones generally have a rather constant impedance vs frequency, which justifies this recommendation. We know this is absolute nonsense.  In any case, I think most headphone manufacturers do now optimise for near-zero output impedance.
 
I can't notice any difference between a 5 ohm amp and a 220 ohm amp, when driving my K601s, and the impedance plot is rather flat for these phones.  However, I CAN notice a difference with some Sony PX-100-II 32 ohm headphones, when I move from my circa 50 ohm laptop output to the 5 ohm amp (being driven from the laptop), and the impedance plot is NOT flat for these phones. (there could be other factors at play here - this doesn't prove that the only cause of the difference is the impedance, but it's certainly possible)
 
Greg.

 

[ClieOS]

Quote:

Interestingly a Fiio E7 has a commendably low output impedance of about 0.2ohms, so is a great match for IEMs, but their own matching E9 desk-top unit has a much higher figure.  In fact the figures for the 1/4" and 1/8" jacks on the E9 are different and the smaller jack is actually higher, not lower, which seems counter intuitive given that if you had IEMs, you'd naturally use the 1/8" socket.
 

E9 is not designed to have high output impedance, but it is required because of the specific opamp used. Unlike many other general purpose audio opamp, the TPA6120 is designed specifically to drive headphone, especially for pro-audio purposes with typical impedance of 300~600 ohm, and it is well known to be fairly good sounding. But the high output of the opamp does have a weakness - that is it becomes unstable on very low output impedance. Thus is why the manufacturer, TI, recommends putting a 10ohm resistor on the output to prevent the opamp ever becoming unstable. That's the reason why E9 has an 10ohm output impedance on the 1/4 jack. As to why the 1/8 jack has an 44ohm output impedance instead of the minimum 10ohm - it is because how much power TDA6120 is pumping out. The 1/8 jack is more likely to connect to a low impedance earphone (i.e. IEM). Without the extra 34ohm, the output of the amp will be too high and render the volume pot being useless, that is, ever a tiny turn on the volume pot it will become too loud for the listener. So the user might end up with an overly loud amp with really bad channel imbalance (which is typical when the pot is in low position). The 44ohm output impedance is there to address that problem.
 
E9, for its intended purpose, is really for high impedance headphone, much like E7 is for low impedance headphone. That is why E9 will never be perfect for low impedance headphone and E7 will never drive high impedance headphone with good result. These are all compromises, no doubt.

 

[twelvebears]

 
Quote:

E9 is not designed to have high output impedance, but it is required because of the specific opamp used. Unlike many other general purpose audio opamp, the TPA6120 is designed specifically to drive headphone, especially for pro-audio purposes with typical impedance of 300~600 ohm, and it is well known to be fairly good sounding. But the high output of the opamp does have a weakness - that is it becomes unstable on very low output impedance. Thus is why the manufacturer, TI, recommends putting a 10ohm resistor on the output to prevent the opamp ever becoming unstable. That's the reason why E9 has an 10ohm output impedance on the 1/4 jack. As to why the 1/8 jack has an 44ohm output impedance instead of the minimum 10ohm - it is because how much power TDA6120 is pumping out. The 1/8 jack is more likely to connect to a low impedance earphone (i.e. IEM). Without the extra 34ohm, the output of the amp will be too high and render the volume pot being useless, that is, ever a tiny turn on the volume pot it will become too loud for the listener. So the user might end up with an overly loud amp with really bad channel imbalance (which is typical when the pot is in low position). The 44ohm output impedance is there to address that problem.
 
E9, for its intended purpose, is really for high impedance headphone, much like E7 is for low impedance headphone. That is why E9 will never be perfect for low impedance headphone and E7 will never drive high impedance headphone with good result. These are all compromises, no doubt.

Completely agree the E9 does a great job with full-size, high-impedance cans, but 'power management' solution or not, having a 1/8" socket would naturally lead owners of IEMs or 'mobile' cans to use this by default (why use an adapter eh?) if they have an E9, and they will then definitely be using an output that will not suit those type of 'phones well.
 
I know the E9 has a primary purpose, but I can also see some seeing it as a convenient 'docking station' for their loved E7, only to find that it then doesn't sound as good.
 
I'm not knocking Fiio - far from it, I have an E10 and love it - but I just think the whole impedance matching thing could be made a lot less hit and miss.
 

 

[estreeter]

No question that the E9 is a fundamentally flawed design, but mine still sounds damned good. Even Voldemort admits that all lower impedance phones arent automatically impacted by the high output impedance - his warning was mainly directed at the owners of BA IEMs.

 

[wuwhere]

I use at least a 10 to 1 ratio, but that's just me.