100 mW into 32 Ohm - What constitutes a decent output?

[linnite]

I'm keen to find out what counts as an adequate headphone amp output. I've seen some amps quote an output of 1W (the manufacturer claims 1W will drive AKG K1000's directly), and then there is the above, 100 mW into 32 Ohm.

I realise output goes down as impedance rises, so in the real world the amp above will output even less into my Senns which I think are 62 Ohm.

But what sort of output is needed to drive cans to decent volumes?

Thanks

EDIT: Whoops, the Senns are 300 Ohm! It's the K701's that are 62 ohm. So the K701's are easier to drive loud, right?

 

[sacd lover]

Quote:

Originally Posted by linnite /img/forum/go_quote.gif I'm keen to find out what counts as an adequate headphone amp output. I've seen some amps quote an output of 1W (the manufacturer claims 1W will drive AKG K1000's directly), and then there is the above, 100 mW into 32 Ohm. I realise output goes down as impedance rises, so in the real world the amp above will output even less into my Senns which I think are 62 Ohm. But what sort of output is needed to drive cans to decent volumes? Thanks EDIT: Whoops, the Senns are 300 Ohm! It's the K701's that are 62 ohm. So the K701's are easier to drive loud, right?

You are missing some of the facts. The type of amp (tube or SS) and the design (OTL tube or transformer coupled tube) also determine power output in to differing impedances. An OTL tube amp will actually increase output power going from 32 to 300 ohms. A solid state amp will lose power going from 32 to 300 ohms .... and a transformer coupled tube amp will typically lose power as the impedance rises unless the transformer is specifically tapped to match a specific higher impedance.

There is also the question of the headphones sensitivity. The AKG has a low sensitivity and a lower impedance which makes the headphone harder to drive than looking just at the impedance. An OTL amp can run out of current delivery with the 701 unless the amp is very powerful. A transformer coupled amp designed for low impedances will have little trouble with the 701 assuming the amp has a decent amount of power. Some transformer coupled amps have seperate transformer taps for both high and low impedance headphones usually selected by a switch .... like the Woo 6. Almost any solid state amp should have plenty of current for the 701 and the 62 ohm impedance is not a difficult load for the solid state amps.

So, now you likely say a good powerful solid state amp is what I need for the 701. But, then there is the sound. The 701s are a little lightweight tonally and can be slightly edgy through the treble. I like the 701s best with tubes persoanally. The Singlepower Extreme is an example of an OTL amp with enough brute power to drive anything including the even the K1000. The Woo 6 has good power along with an impedance selector to tap the transformer for driving high or low impedances. The SS Headamp GS-1 will drive the low impedance of the AKGs very well and still has plenty of power to drive 300 ohm Senn 580/600/650s.

100mw is a good amount of power and generally adequate IF that is available to both headphones. The Mapletree Ear+ has a spec of 100mw @ 32 ohms. I owned the Ear+ amp and I didnt think the amp maintained enough power output into high impedances for the 300 ohm Senns. The GS-1 has a spec of 1 watt = 1000mw @ 32 ohms. The GS-1 is still putting out a minimum of 200-300 hundred mws into 300 ohms and drives the Senns easily from my experiences. The Singlepower Extreme puts out arouns 1.5-1.8 watts = 1500-1800mw @ 300 ohms and still will deliver 400mw or so @ 32 ohms, possibly more, and consequently as I said ..... drives anything.

So, in summmary you need to look at not just impedance but the headphones sensitivity. Once you know the headphones requirements than you need to look for an amp that will mate successfully with differing headphones by offering high power and/ or impedance matching capability. I like the brute force approach myself!

 

[linnite]

Thanks for the detailed reply, much appreciated.

 

[myinitialsaredac]

sacd lover got most of the importance of sensitivity and impedance, but just for a bit further delve into this, we need to look at ohms law ( amperes = voltage / resistance). So if we say that the resistance increases the voltage of the chain will also increase (amperes x resistance = voltage), but the wattage in the chain decreases (watts = voltage x amps so if the amperes decrease the wattage subsequently decreases). To break it down into some numbers we can say voltage is 120 and resistance is 62 ohms and 300 ohms respectivaly. This yields 1.93 amps and .4 amps respectively, or put that into our other equation and we end up with 231.6 miliwatts and 45 miliwatts respectively at the end of the circuit. Now if we plug in these wattages into our sensitivity equation noting that 3db is a doubling of power (so if 92db is 1mw 95 is 2, 98 is 4, 101 is 8, 103 is 16 106 is 32 on and on) we can get some interesting numbers and also see how low sensitivity phones eat away power. So if an amp is capable of putting out 100mw at 32 ohms at 115 volts, then its amperes would be .86. at 300 ohms with 115 volts the amperes look like .383. In wattages this is 44.045 miliwatts, or less than half the amount of wattage, and in terms of our sensitivity model above we can see a low sensitivity phone eating 44 miliwatts quickly if you take into account the source impedance also.

Now I dont want to pretend to be an electrical engineer, this is all stuff I learned from reading online and I may not be correct so I am open to learning something but I believe that is fairly accurate.

Dave

 

[linnite]

Thanks again.

The AKG's have 105dB/mW sensitivity - that's high (or average at worst) rather than low isn't it? I see the various Grado's average about 98dB/mW Grado Labs - Wikipedia, the free encyclopedia

Again very happy to be put right - n00b trying to get up to speed.

 

[majkel]

No, it's 105dB/V. The AKG website is unclear in the description while we all know the AKG's are silent/difficult to drive/power hungry.

 

[argentum]

It is more like 92 or 93 dB/mW with K701 . A little different example , but you will get the idea - K 271 MK II - Engineer's specifications

 

[linnite]

Thanks again everyone.

 

[QQQ]

AKG k701 is 93db\mw sensivity.

If 105 was real, they will be much easier to drive than KSC-75.

 

[MaloS]

Quote:

Originally Posted by myinitialsaredac /img/forum/go_quote.gif sacd lover got most of the importance of sensitivity and impedance, but just for a bit further delve into this, we need to look at ohms law ( amperes = voltage / resistance). So if we say that the resistance increases the voltage of the chain will also increase (amperes x resistance = voltage), but the wattage in the chain decreases (watts = voltage x amps so if the amperes decrease the wattage subsequently decreases). To break it down into some numbers we can say voltage is 120 and resistance is 62 ohms and 300 ohms respectivaly. This yields 1.93 amps and .4 amps respectively, or put that into our other equation and we end up with 231.6 miliwatts and 45 miliwatts respectively at the end of the circuit. Now if we plug in these wattages into our sensitivity equation noting that 3db is a doubling of power (so if 92db is 1mw 95 is 2, 98 is 4, 101 is 8, 103 is 16 106 is 32 on and on) we can get some interesting numbers and also see how low sensitivity phones eat away power. So if an amp is capable of putting out 100mw at 32 ohms at 115 volts, then its amperes would be .86. at 300 ohms with 115 volts the amperes look like .383. In wattages this is 44.045 miliwatts, or less than half the amount of wattage, and in terms of our sensitivity model above we can see a low sensitivity phone eating 44 miliwatts quickly if you take into account the source impedance also. Now I dont want to pretend to be an electrical engineer, this is all stuff I learned from reading online and I may not be correct so I am open to learning something but I believe that is fairly accurate. Dave

There is something very wrong with your math. 115 Volts into 300 ohms => .383 amperes, which means 115*.383 watts = 44.045 WATTS, which is enough to fry the headphone 10 times over, kill you with a small lightning since all of this is on your head, and burn your house down.

 

[bada bing]

Quote:

Originally Posted by myinitialsaredac /img/forum/go_quote.gif sacd lover got most of the importance of sensitivity and impedance, but just for a bit further delve into this, we need to look at ohms law ( amperes = voltage / resistance). So if we say that the resistance increases the voltage of the chain will also increase (amperes x resistance = voltage), but the wattage in the chain decreases (watts = voltage x amps so if the amperes decrease the wattage subsequently decreases). To break it down into some numbers we can say voltage is 120 and resistance is 62 ohms and 300 ohms respectivaly. This yields 1.93 amps and .4 amps respectively, or put that into our other equation and we end up with 231.6 miliwatts and 45 miliwatts respectively at the end of the circuit. Now if we plug in these wattages into our sensitivity equation noting that 3db is a doubling of power (so if 92db is 1mw 95 is 2, 98 is 4, 101 is 8, 103 is 16 106 is 32 on and on) we can get some interesting numbers and also see how low sensitivity phones eat away power. So if an amp is capable of putting out 100mw at 32 ohms at 115 volts, then its amperes would be .86. at 300 ohms with 115 volts the amperes look like .383. In wattages this is 44.045 miliwatts, or less than half the amount of wattage, and in terms of our sensitivity model above we can see a low sensitivity phone eating 44 miliwatts quickly if you take into account the source impedance also. Now I dont want to pretend to be an electrical engineer, this is all stuff I learned from reading online and I may not be correct so I am open to learning something but I believe that is fairly accurate. Dave

Except for the math error in confusing milliwatts with watts (as noted by MaloS above) it is a good attempt at calculating. Just as a very rough estimate, dynamic headphones typically are run about 2 volt output peaks, and most get uncomfortably loud around 100 mW. You need a large cushion in output power for quality sound - to prevent clipping and "sagging" . An amp that can output an honest 200-300mW into the connected load is actually plenty powerful for all but the most power hungry cans.

Your calculation methods yield some odd interpretations though. Generally, circuits are analyzed by looking at voltage and the resulting current is calculated. It is rather odd to see a fixed current and calculate the resulting voltage. That isn't the way it works in practice. Audio amps are voltage control devices, not current control devices. At a given volume pot setting, the output voltage of an amp is the same (ideally) no matter what impedance cans are connected. It's the volume knob, not the headphone impedance that varies the voltage - usually along an "audio" taper. The resulting current will be whatever Ohm's law predicts. The only bugaboo in the simplistic calculations is ignoring the output impedance of the amp. Ignoring it doesn't change much with high impedance phones or low impedance amps, but it becomes an increasing factor as lower impedance phones and higher impedance amps are used. The output impedance of tube amps is nearly always greater than SS designs and additionally tube amps have less flat impedance curves (the changing resistance over frequency) partly because of output coupling devices (capacitors or transformers) required by tube circuits. That makes calculating output in real world tube amp circuits a bit more complicated - especially with low impedance loads.

I sort of pretend to be an EE, at least I have 'em fooled to pay me for the sham. I work in power generation/distribution, not audio, but Ohm's law still works.

 

[Pricklely Peete]

Excellent thread guys....

Peete.