Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Questions about voltage/current headphone requirments (Read 7739 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Questions about voltage/current headphone requirments

As a sort of mental exercise (to satisfy personal curiosity) I've been trying to either verify or find evidence against many of the claims made for dedicated headphone amplifiers.  I've been able to verify some claims, such as needing something with a larger voltage swing when using a high impedance headphone for a portable music player (trying using an AKG 240M 600 ohm with a Sansa Clip).

I'm not an EE so that's why I'm asking for help.

Rather than just throw theoretical numbers out there for how much current/voltage a headphone amp needs to deliver, I've been trying to figure these out empirically.  I would appreciate if somebody could read over what I'm trying and point out if I'm making any mistakes.  Assume that when I say something like "required" I mean how much is needed to drive the headphones to volumes slightly above what I prefer listen to.

I wanted to test the required current when driving a pair of low impedance headphones (25 ohm) headphones.  I took my portable music player, set the volume to maximum, then used a Y splitter at the headphone out.  A headphone was connected to one end of the splitter and a multimeter in AC volts mode was connected to the other.  The multimeter is a cheapo radio shack auto ranging meter.

I then played a range of music ranging from highly compressed pop (very loud all the time) to very dynamic classical.  I never ever saw a voltage reading higher than .16 volts.  As a multimeter can't possibly change the display fast enough to accurately represent music, I just rounded this up to .20 volts.  Then I doubled it for an even greater margin of error.

So we have .4 volts.  Using V(.4)=IR(25), I = .016 amps.  That's 16 milliamps with what I assume (am I assuming wrong?) is enough of a window for dynamic swings.

Many dedicated headphone amps have current buffers able to provide current in excess of 100mw.  From my calculations this seems like a total overkill. 

16mA does not seem much at all.  Can a typical sound card or a basic amp like a OPA2227 based Cmoy (no current buffers) output 16ma?

I don't own a pair of high impedance headphones so I am not able to do a empirically based test without building a dummy load.  So I'm just going to extrapolate from the above.

16ma at 25 ohms calculates into 6mw of power.  Let's throw in all sorts of variance and assume that a 300 ohm headphone with less sensitivity needs 15mw of power with doubling to 30mw for volume/voltage swings.  That turns into 3 volts of power.

So, based on these two calculations, a 3V headphone amp with no current buffers could drive most headphones without issue.  If we throw in some more headroom for things like impedance varying with volume etc, we could double the voltage to 6v or 4 AA batteries.

So my conclusion that a basic opamp (such as one based around the OPA2227) circuit based amp with a 6V supply should be enough to drive any headphones, and that includes all the random doubling of "Just in case" numbers I've thrown in.

Does all of this make sense?  Have I made a mistake?  Am I right to use a multimeter to measure voltage swings in music?

Following from the above, even a computer sound card should be able to provide the required current/voltage to drive most headphones.  Based on some volume matched listening tests comparing my sound card to dedicated amplifiers, I have come to the same conclusion based on what I heard.

Thanks for your help.

Questions about voltage/current headphone requirments

Reply #1 –
Quote from: odigg on
I wanted to test the required current when driving a pair of low impedance headphones (25 ohm) headphones.  I took my portable music player, set the volume to maximum, then used a Y splitter at the headphone out.  A headphone was connected to one end of the splitter and a multimeter in AC volts mode was connected to the other.  The multimeter is a cheapo radio shack auto ranging meter.


You would obtain more definitive results if you used steady-state test tones instead of music.

Also, using an actual resistor (they still sell them at RS, right?) would also give your results more meaning.

Quote
I then played a range of music ranging from highly compressed pop (very loud all the time) to very dynamic classical.  I never ever saw a voltage reading higher than .16 volts.  As a multimeter can't possibly change the display fast enough to accurately represent music, I just rounded this up to .20 volts.  Then I doubled it for an even greater margin of error.

So we have .4 volts.  Using V(.4)=IR(25), I = .016 amps.  That's 16 milliamps with what I assume (am I assuming wrong?) is enough of a window for dynamic swings.

Many dedicated headphone amps have current buffers able to provide current in excess of 100mw.  From my calculations this seems like a total overkill.


Yes.

Quote
16mA does not seem much at all.  Can a typical sound card or a basic amp like a OPA2227 based Cmoy (no current buffers) output 16ma?


Well you actually are talking 16 ma average, which corresponds to more like 22 ma peak. Peak current is how op amps are rated. Most op amps are current protected at some level above 30 ma.  For the op amps you mentioned, the current limit is more like 40 ma.

Quote
I don't own a pair of high impedance headphones so I am not able to do a empirically based test without building a dummy load.  So I'm just going to extrapolate from the above.


High impedance headphones generally take far less current than low impedance headphones.

Quote
16ma at 25 ohms calculates into 6mw of power.  Let's throw in all sorts of variance and assume that a 300 ohm headphone with less sensitivity needs 15mw of power with doubling to 30mw for volume/voltage swings.  That turns into 3 volts of power.

So, based on these two calculations, a 3V headphone amp with no current buffers could drive most headphones without issue.  If we throw in some more headroom for things like impedance varying with volume etc, we could double the voltage to 6v or 4 AA batteries.


I use 5 volts for my headphone amp supply voltage. Emprically, 4.2 volts seems to be enough, but 3 volts does not.

Quote
So my conclusion that a basic opamp (such as one based around the OPA2227) circuit based amp with a 6V supply should be enough to drive any headphones, and that includes all the random doubling of "Just in case" numbers I've thrown in.


I don't know about some pro audio, high impedance, low sensitivity headphones.

Quote
Following from the above, even a computer sound card should be able to provide the required current/voltage to drive most headphones.  Based on some volume matched listening tests comparing my sound card to dedicated amplifiers, I have come to the same conclusion based on what I heard.


I use a gain=2, 5 volt supply headphone amp with some headphones and my computer. It's a Boostaroo - about $25.

Questions about voltage/current headphone requirments

Reply #2 –
Thanks for reading over my post.  Just a few more questions/comments.

Quote
You would obtain more definitive results if you used steady-state test tones instead of music.

Also, using an actual resistor (they still sell them at RS, right?) would also give your results more meaning.


I tried that and the output voltage was slightly higher to my prior measurement.  Considering how much extra room I put in my old calculations,I think most of my prior conclusions/calculationss till hold.  The headphones I used have a measured flat 25 ohm impedance in the audible range.  I'll have to try this with a resistor.

Quote
Well you actually are talking 16 ma average, which corresponds to more like 22 ma peak. Peak current is how op amps are rated. Most op amps are current protected at some level above 30 ma.  For the op amps you mentioned, the current limit is more like 40 ma.


If I was looking for the current limit on the datasheet, would this be the short circuit current?  For the OPA2227 this is rated at +-45mv.  I don't know why the - is in there.

Quote
I use 5 volts for my headphone amp supply voltage. Emprically, 4.2 volts seems to be enough, but 3 volts does not.


Thanks for this.  I wanted some empirical evidence to either support or go against my calculations.  So 3 volts is not enough.  6 should be.

Quote
I use a gain=2, 5 volt supply headphone amp with some headphones and my computer. It's a Boostaroo - about $25.


The results of all this slightly depressing.  I can't seem to find any reason, through measurement, calculation, or listening tests, that helps me understand all the high priced headphone amps out there.  It seems like a Cmoy with a 9V supply can provide all the current or voltage swings required for most headphones excluding those 2000 ohm pro audio monsters.  The OPA2227 also has a relatively low slew rate 2.3 V/us, but that's fast enough for the voltages talked about with headphone amps.

It's not depressing in a bad way, just kind of sad to see how much fluff there is out there.

Questions about voltage/current headphone requirments

Reply #3 –
I own a PA2V2. I'm quite satisfied with it and the guy who builds them is a nice person. It easily powers a 300 ohm pair of Sennheiser HD 600 from 2.4V input voltage.

I'm not using it anymore, though, since the HD 595's 50 ohm are a perfect match for the iPod without a separate amp.

Questions about voltage/current headphone requirments

Reply #4 –
Quote
Rather than just throw theoretical numbers out there for how much current/voltage a headphone amp needs to deliver, I've been trying to figure these out empirically.
  You can  good (useful) calculations if you know two things... 

1.  You need to know how loud the peaks need to be for a given listener and a given type of music.  That's difficult!

2.  You need to know the sensitivity of the headphones.  For good headphones, that information is published...  The problem is, that you (or the listener) may change headphones sometime in the future. 

Quote
...a multimeter in AC volts mode was connected to the other. The multimeter is a cheapo radio shack auto ranging meter...  As a multimeter can't possibly change the display fast enough to accurately represent music...    Am I right to use a multimeter to measure voltage swings in music?
  A digital multimeter is pretty useless with program material (as opposed to constant test tones).  The meter is designed/calibrated to give an RMS reading from a constant AC sine wave.  (Most meters will also give the wrong RMS reading for a constant square wave.)  If the signal is not constant, different meters will give different readings.  Unless you've got a special meter that has "peak hold" won't get a good reading.    An analog meter is slightly better, but the movement is usually too slow to catch infrequent peaks, and with the needle jumping around, its difficult to make sense of the reading. 

If you were an EE, you'd probably be checking the peak voltage with an oscilloscope.  And with an oscilloscope, you could see if you are clipping under real-world listening conditions. 

This might not be any easier for you than getting your hands on an oscilloscope, but with an op-amp, you can build a peak detector (see page 10 & 11 of this data sheet).  This circuit will hold the peak voltage, so you can measure it with your multimeter set for DC.  A peak detector won't directly indicate clipping, but you could intentionally drive the amp into clipping, and take a maximum measurement.  Then, if you never reach that peak under normal listening, you could be reasonably sure you're not clipping.

Test tones are fine for making measurements, but not that helpful for listening or evaluating loudness.  A constant test tone doesn't need any headroom.

Quote
I can't seem to find any reason, through measurement, calculation, or listening tests, that helps me understand all the high priced headphone amps out there.... It's not depressing in a bad way, just kind of sad to see how much fluff there is out there.
When it comes to high-end or audiophile equipment, the cost can be very high for very little benefit.  I can get a 300 - 400W power amp for $200 USD, so I wouldn't expect to pay that much for a headphone amp, which is much easier to build.  If was building a (non-portable) headphone amp, I'd probably use plus-and-minus 15 volts (or more) for the power supplies.  There is little or no cost in providing the extra headroom, except if you go too far, I suppose someone could "fry" their headphones (perhaps when they weren't listening).


...Looking at the specs for the Sennheiser HD280*, I see the following:
Sensitivity = 113dB @ 1V RMS.
Impedance = 64 Ohms.

1V RMS is (2.828 V peak-to-peak), and you should be able to easily get that with a 5V supply.  At 64 ohms, that's about 16mA RMS. 

* I don't know if 113dB peaks are loud enough for you, and I don't know if these specifications are "typical", I just chose the HD280s because they are popular, and I own a pair.

Questions about voltage/current headphone requirments

Reply #5 –
Quote from: odigg on
Quote
Well you actually are talking 16 ma average, which corresponds to more like 22 ma peak. Peak current is how op amps are rated. Most op amps are current protected at some level above 30 ma.  For the op amps you mentioned, the current limit is more like 40 ma.


If I was looking for the current limit on the datasheet, would this be the short circuit current?  For the OPA2227 this is rated at +-45mv.  I don't know why the - is in there.



Short circuit current is about the same as maximum current output. I notice that they derate it  a bit for higher package temperatures.  The minus is in there because op amps can both source and drain current to/from the load. They are often operated with both plus and minus power supplies, and even when DC coupled, their output can rotate about ground.

Quote
I use 5 volts for my headphone amp supply voltage. Emprically, 4.2 volts seems to be enough, but 3 volts does not.


Thanks for this.  I wanted some empirical evidence to either support or go against my calculations.  So 3 volts is not enough.  6 should be.

Quote
Quote
I use a gain=2, 5 volt supply headphone amp with some headphones and my computer. It's a Boostaroo - about $25.


The results of all this slightly depressing.  I can't seem to find any reason, through measurement, calculation, or listening tests, that helps me understand all the high priced headphone amps out there.



How about that!  ;-)

Quote
It seems like a Cmoy with a 9V supply can provide all the current or voltage swings required for most headphones excluding those 2000 ohm pro audio monsters.  The OPA2227 also has a relatively low slew rate 2.3 V/us, but that's fast enough for the voltages talked about with headphone amps.


If you want more slew rate, pick the 2228 version that is compensated for higher gains.

Quote
It's not depressing in a bad way, just kind of sad to see how much fluff there is out there.


Yes, there is a lot over overkill in audio.

Questions about voltage/current headphone requirments

Reply #6 –
Quote from: rpp3po on
I own a PA2V2. I'm quite satisfied with it and the guy who builds them is a nice person. It easily powers a 300 ohm pair of Sennheiser HD 600 from 2.4V input voltage.


Thanks for the info.  So it seems even a 2.4V can do it, which supports my earlier rough calculation of 3 volts.

DVDDoug - If only I had an oscilloscope.  Unfortunately I can't justify the cost or even the space it would take up on my desk considering considering how little I would use it.  I've been thinking about picking up an inexpensive USB ADC and using a software oscilloscope.  Do you have any experience with this?

I realize my numbers are very rough and my measurements are far from ideal.  Part of reason I've been doing this is to get a rough estimate of what voltage/current source I would need to drive most any headphone I would consider owning.  I've personally tried a DT880 (250 ohm) with a Cowon D2 and a Sennhesier HD600 (300 to 600 ohms depending on frequency) with an very expensive, well rated in the audiophile world, portable amp powered by a 3.7V lithium battery.  I was able to achieve painfully loud volumes without distortion, clipping, or anything else that would lead me to conclude those amps couldn't swing the voltage required to drive those headphones.

Based on my research 30 v (+-15v) is clearly overkill.  I suppose with a desktop amp you might as well build in as much headroom as you want, but I don't see the logic in doing so.  If I was to build an amp I'd build something that was portable and ran off 6V (4 AA batteries) or 9V.  I'd then put in a jack for a wall-wart so I could use it as a desktop amp.

Thanks for the help and input everybody.  You have been a great help.

Questions about voltage/current headphone requirments

Reply #7 –
Quote
DVDDoug - If only I had an oscilloscope. Unfortunately I can't justify the cost or even the space it would take up on my desk considering considering how little I would use it. I've been thinking about picking up an inexpensive USB ADC and using a software oscilloscope. Do you have any experience with this?
Yeah...  I'm not suggesting that you go out and buy (or rent) a 'scope.  But, since I'm saying that a multimeter is the "wrong" tool, I needed to mention the "right" tool.

No, I don't have any experience with PC-based oscilloscopes.  I've seen them (hardware & software) in catalogs, but from what I recall, these are a few hundred dollars.  I don't know what software is available (or comes with) a USB ADC.  Depending on what software tools it comes with, it might be rather involved to make a 'scope or some sort of peak measuring tool.

I've also seen software that can turn you soundcard into an oscilloscope.  A search turned-up Zelscope which only costs $10 USD!    You might even be able to find something like that free.    You'd have to calibrate it, and you'd need to make a voltage divider (2 resistors), since the soundcard's input itself will clip at about 1V.

Questions about voltage/current headphone requirments

Reply #8 –
Free software oscilloscopes for the PC...

Winscope

BIP Oscilloscope

Cheers, Slipstreem. 

Questions about voltage/current headphone requirments

Reply #9 –
Quote from: odigg on
DVDDoug - If only I had an oscilloscope.  Unfortunately I can't justify the cost or even the space it would take up on my desk considering considering how little I would use it.  I've been thinking about picking up an inexpensive USB ADC and using a software oscilloscope.  Do you have any experience with this?


For what you want to do, the real-time presentation of an oscilliscope, implemented via hardware or software, is probably overkill.  It might even be less than what you want, as it can be handy to record a signal and then examine it at your leisure.

Just record the signal you want to look at, and then examine it using a sound editor such as Audacity (freeware).

You've got the problem of calibrating it, which is facilitated by the fact that you already have some kind of meter. Just generate a 60 Hz tone with Audacity, and then play it back through a Y cable, one side of the Y going to your voltmeter, and the other side of the Y  looped back into the input of your PC's audio interface.

Note the equivalence between what you measure with the voltmeter and what you see on the screen with Audacity.

Questions about voltage/current headphone requirments

Reply #10 –
Quote from: rpp3po on
I own a PA2V2. I'm quite satisfied with it and the guy who builds them is a nice person. It easily powers a 300 ohm pair of Sennheiser HD 600 from 2.4V input voltage.


I must revise that statement and post my results for reference. After suspecting audible differences several times over the last year, I have tried a level matched comparison with a Y splitter cable today. Already at 0.7 V I can identify clipping for a 440 Hz tone, that I don't hear when it is plugged into my DAC's output. The test wasn't really as blind as it should have been. I had a friend's help and could hear him fiddling with cables. But the distortions were really pretty obvious.

It took a while to get to this point, since I mostly fed the HD 600 directly from my DAC and used the HD 595 for portable use.

Questions about voltage/current headphone requirments

Reply #11 –
Quote from: rpp3po on
I must revise that statement and post my results for reference. After suspecting audible differences several times over the last year, I have tried a level matched comparison with a Y splitter cable today. Already at 0.7 V I can identify clipping for a 440 Hz tone, that I don't hear when it is plugged into my DAC's output.


Well I suppose that makes sense as a 0.7V tone is basically 2Vpp.  Assuming some losses on the part of the circuit, 2Vpp is probably at the upper limit of what one can expect from a 2.4V supply.

But I do have a further question.  What type of music were you listening to?  Was is quiet (in average RMS terms) music?  0.7V is is a little above 1mw, which should be quite loud with certain types of music (e.g. modern pop music) as the HD600 has a sensitivity of 102db/mw.

And since you've brought it up I'd like to point out a serious flaw in my mathematics.  All the values for volts are in Vrms, yet when I did the math I quite stupidly thought about them as Vpp.  In those terms, the HD600 probably needs (as Arnold Krueger pointed out) a supply closer to 5Vpp at 10mw (a 10db gain!) of power.
 

Questions about voltage/current headphone requirments

Reply #12 –
I didn't listen that much via the PA2v2 at all, since I use my IEMs and 595s for mobile use without amp. And for stationary use I don't need a mobile amp, so the PA2v2 was basically a waste of money. Back then I had tried some Jazz and classical music and thought it was fine - probably lower to medium territory RMS wise. Honestly, I mainly checked if I could get enough volume and if it sounded basically fine. I did not expect that you can, by design, turn the volume knob into so much clipping.