• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

Driving headphones with speaker outputs?

P

p. rex

Active Member
Hey all,

I've been doing some reading about headphones and one consistent problem seems to be impedance matching: how can I get an acceptable damping factor when headphone outputs are all between 32 and 250 ohms? I was thinking, maybe I could get some very high impedance headphones (Beyerdynamic DT880 600-ohm comes to mind) and drive them directly off the speaker outputs on my Pioneer SX-1010. Some simple back-of-the-envelope calculations suggest that the amp should be capable of outputting about 1.5 watts into a 600 ohm load.

Now, Beyerdynamic specifies power handling for the headphones as 100 milliwatts, or about 12 decibels below the RMS maximum power (which is 1.5 watts at 600 ohms). Can anyone venture a guess as to what position on the volume knob this would be?

Also, from a less technical standpoint, has anyone actually tried this?
 
Register to hide this ad
You can drive the headphones directly from a power amp designed to drive speakers. I feel the only possible downside might be a high noise floor as the gain has been set for larger output swings.

Where would the volume position be? That is a problematic question. A) the volume control has a linear taper and B) the signal out is dependent on the power amp gain, volume position, and the level of the signal input. Just use the volume control as you would with your speakers.
 
It will "work" as in sound will come out (lots of it). The question is will it sound decent? I would think the issue will be that the volume control will be very near minimum for any reasonable sound level in the headphone. That would make me worry that cross-over distortion (assuming it is a class A-B amp) would be emphasized. If you measure the distortion of A-B amps with very low signals, they often will have pretty high distortion.
The other concern would be the potential to accidently over-drive the headphones and blow them to hell (along with your brains).
If I was going to try to drive headphones directly from speaker outputs, I would probably use a resistive attenuator so that the amp sees a lower impedance (near 8 Ohms) and the headset gets a severley attenuated signal so that at max volume it would be slightly below the rated value for the headset.
From your calculation that you can drive 600 Ohms to 1.5 Watts, I assume the amp can swing 30 V RMS. Maximum for 600 Ohm headsets at 100mW is about 8 V RMS. So you need to attenuate the voltage about 12 dB. Lets say you use an 8 Ohm series resistor in series with the headphones. That means you want about 2.875 Ohms in shunt (in parallel with the headphone output). That would be a damping factor of 215! That's a huge damping factor. The 8 Ohm series resistor would of course have to be a power resistor (like 100 watts if you were going to crank the hell out of it, although I would think 30 watts would be more than sufficient for anything realistic).
You could of course lower that power considerably by accepting an amplifier load of greater (more ohms) than 8 ohms and a reduction in damping factor. That is obviously what the manufacturers do when the include the dropping resistors in the headset circuit. You would have the freedom to decide what values you felt were reasonable.

I'm curious as to why you think damping factor is important in headsets? I'm not arguing that it isn't, I'm just trying to learn. I thought damping factor was important to speakers to damp the movement of the large mass of the woofer and better control it. I would have thought that this would be much less important in a headset with its much smaller driver.

Terry
 
What is your damping factor calculation? If the OP has a 600 ohm headphone and inserts a 8 ohm resistor in series, the 8 ohms is part of the output impedance or sorce resistance. Ignoring the amp's fractional ohm source impedance the damping factor is given as Rload/Rsource 600/8 = 75.

I am not saying that damping factor is important with headphones, I just disagree with the number. I also disagree about padding the output with resistors. If the the headphones impedance curve (maybe not flat) doesn't match the impedance curve of the divider circuit (more likely flat), there will frequency response errors.

I also disagree that a directly connected headphone is very dangerous to the headphones. Sure, there's people that think full clockwise is max power, but let's not dumb down the world for them. They can do what they want with their headphones and speakers.

I disagree about crossover distortion being prevalent at low volumes. A+A/B amplifiers are Class A until some output current. For a 600 ohm load, the amp may be in Class A mode until clipping.

I will add one more possible downside to headphones directly connected to a amplifier designed to drive speakers. Some volume controls have horrible tracking at very low settings.
 
My suggestion was to use a voltage divider with a resistor in parallel with the headphone. This means the amp sees 8 Ohms even if the headset is unplugged. The resistor is 2.875 Ohms, so the damping factor would be 600/2.875 = over 200. It would be even higher than that, since the 2.875 Ohm resistor would be shunted by the 8 Ohm resistor plus the output impedance of the amp (near zero), so it would actually be more like 280!
As far as cross-over, you may be right. I know from working on amps that the distortion will be low for very low signals since as you point out, they are running in class A. But as the signal level rises (still very low) to the point that it starts to switch from A to B, the distortion rises rapidly. You can even see this on an oscilloscope as crossover distortion on a sine wave output. It is a very high percent distortion at that point. But the amount of distortion stays constant, so that at full level, the percent of distortion becomes small numerically, and manufacturers can claim 0.1% or whatever. But measure it with just a few volts swing, and you will see much higher numbers. I don't know where it would lie with the headphone as a load, but it concerns me.
As far as the frequency response errors, again, having the lower output impedance with the shunt resistor mitigates most of that. The 600 Ohm impedance of the headphone in parallel with the 2.875 Ohm resistor means the 600 Ohm impedance variations have almost no effect on the voltage transfer ratio. Not zero effect, but pretty low. For instance if the headphone impedance changed from 600 Ohms to 300 Ohms (a pretty large swing), the voltage would change 0.03 dB. That should be good enough.

Terry
 
I am all over power testing and have observed when and where crossover distortion appears. Class A operation shifts to Class A/B when the output current is greater than the bias current. For a 600 ohm pair of headphones, that might not occur until the amp clips. It is not a function of output voltage.

Why does the amp need to be terminated into 8 ohms anyway?
 
Radio Shack used to have a box with a terminal strip for the speaker out connection and two headhone jacks with a level knob. I still have two of them somewhere. I don't think thay can handle high power
 
Jon_Logan said:
I am all over power testing and have observed when and where crossover distortion appears. Class A operation shifts to Class A/B when the output current is greater than the bias current. For a 600 ohm pair of headphones, that might not occur until the amp clips. It is not a function of output voltage.

Why does the amp need to be terminated into 8 ohms anyway?
Click to expand...

I see your point about the distortion. You are probably right about the outputs staying in class A with a headphone load. 20mA bias would stay pretty much class A up to about 1/4W into 600 Ohms.

As far as the loading, I'm mainly just paranoid. Ive seen plenty of amps that did not appreciate driving into an open. Usually I see huge high frequency peaking. Overshoots on rising edges of square waves and such. Even signs that it was getting close to oscillating (fuzziness on the rising edge of a square wave). I'm not familiar with the OP's amp. Might be fine. Usually, the amps that tolerated it worst were the more expensive, higher bandwidth ones.

Terry
 
TerryS said:
As far as the loading, I'm mainly just paranoid. Ive seen plenty of amps that did not appreciate driving into an open. Usually I see huge high frequency peaking. Overshoots on rising edges of square waves and such. Even signs that it was getting close to oscillating (fuzziness on the rising edge of a square wave). I'm not familiar with the OP's amp. Might be fine. Usually, the amps that tolerated it worst were the more expensive, higher bandwidth ones.

Terry
Click to expand...
welcome to the wonderful world of the Audiolab 8000A......can I smell burnt semi conductors.
 
egg_on_toast said:
welcome to the wonderful world of the Audiolab 8000A......can I smell burnt semi conductors.
Click to expand...

It is not a slam on an amp to say that it is unstable in to high impedance loads. Some of the best amps show this tendancy. The feedback loop is optimized for maximum bandwith and transient response for the expected load. This often means a compromise with other loads.

I appreciate the chance to discuss this with knowledgeable people. I've already learned a lot from Jon_Logan in just this short discussion. For some reason, I never realized that the distortion with a high impedance load would be low because it is always in class A mode. Duh! Thanks for pointing that out.

I would still like to understand the importance of good damping factor on headsets. How important is it? A lot of headset amplifiers that I see won't have very good damping factor. Some place fairly high resistors in series with the output to provide short circuit protection. With a fairly low impedance headset (say 32 Ohms), this would be a really crappy damping factor.

Terry
 
Man, thanks for all the useful responses. The question about the importance of damping factor is a good one -- what I've read is that most manufacturers now optimize for an amp output impedance of 0 ohms, but apparently Beyerdynamic still designs their cans with the old 120-ohm DIN spec in mind. Since the impedance of the SX-1010's headphone output is 150 ohms, that might be worth trying as well. Who knows, maybe a damping factor of 4 is adequate for headphones?

Terry, I'm worthless with a soldering iron, but I suppose such a setup would be a good starter project. As for the issue of strange behavior from the volume control, the SX-1010 has a nice -20db pad (which is unfortunately broken now, but can no doubt be sorted out once I get around to sending it to EchoWars). Glad to hear that crossover distortion should be no problem, since it'll be in class A operation at the relevant power level.

Also -- if accidentally blowing the headphones up is an issue, would it be possible to just wire fuses in line with the output (or do they not make fuses of values that low)? That combined with Terry's voltage divider/resistor trick seems like it should solve both the problem of damping and that of blown headphones.

Also, my grasp of electrical engineering is at about a high-school physics level -- rusty high school physics that I don't remember all that well. So I'd take my math with a grain of salt.

That said, I have the SX-1010 service manual here and I can confirm that the amp swings 30.5 watts. I derived the 1.5625 watt number just by halving output for each doubling of speaker impedance.
 
p. rex said:
Since the impedance of the SX-1010's headphone output is 150 ohms, that might be worth trying as well. Who knows, maybe a damping factor of 4 is adequate for headphones?
Click to expand...

Don't get caught up in the number, especially DF. The real question in all this is how does it sound?
 
Most receivers, amps, etc., drive the headphone jack directly from the speaker output stage. All they do is insert a 50 to 100 ohm resistor between the L+ and R+ to Ground lugs of the jack. That is enough to pad it down. This works for SS amps.
 
i don't know about you guys but i tried this little experiment with some headsets i had and it worked for me pretty well. soldered 16 gauge speaker wire to the drivers. hooked it back directly to my 1970's kenwood kr-720 receiver and to my surprise no fried drivers. i didn't experience any hissing issues either. i can't get the volume knob past dial 1 on my sansui au-d7 i tried too.very slight hiss but not too noticable at all and gone when source starts playing. it works fine.

here's some pics what i did today.

the driver itself:
http://i882.photobucket.com/albums/ac26/RexAeterna/DSC01055.jpg

the gutted enclosure:
http://i882.photobucket.com/albums/ac26/RexAeterna/DSC01059.jpg


completed project:
http://i882.photobucket.com/albums/ac26/RexAeterna/DSC01061.jpg


these are just links cause the images are pretty big resolution and i don't feel like resizing cause photobucket does it oddly i think. i know amps and receivers use a simple resistor between the power amp section and headphone out but no harm trying this idea i guess. it worked for me pretty well and the drivers i used are 55ohms (well it said on cyber snipa website) and looked very similar to akg drivers.
 
Last edited:
Thanks, Rex! It's good to see someone's actually tried this. However, I'm probably going to abandon the idea for the time being, since I've found out a little more about the Beyer headphones. Apparently, Beyerdynamic's own headphone amplifier, which they recommend for use with their headphones, has an output impedance of 100 ohms; since this is the case, the SX-1010's built-in 150 ohm headphone outputs seem close to a perfect match.
 
no prob. i always thought so and was told too that hooking headphones to speaker outputs was a no-no. i tried this on headset i never cared for and it worked. have to be very very careful with the volume knob i notice tho and make sure it's always set to 0 before turning the amp on. that will blow out both your ears and the drivers if you ever plan on to try it cause it's getting the full force of the power amp this time around.
 
Impedance matching of the source to the load is only important when transferring maximum power. This is not the case with headphones; they only need a few mw of power, and their high impedance protects them.

For almost all amps and headphones, virtually all of the listening will be in the class-A operation of the amp. For an AB class amp, the transition from A to AB occurs when the output current exceeds twice the bias current. You can take twice your bias current and calculate what power to which that corresponds into your load.

P=R*I^2 where R is the load and I is twice the bias

For example, if your phone impedance is 300 ohm and the bias in amp is set at 20ma, twice that current is 40ma, so the power in watts where the class of operation changes from A to AB is:

(300)*(0.04^2)=0.48W or about 1/2 Watt (with headphones)

With 8 ohm speakers on the same amp:

(8)*(0.04^2)=0.01W or 1/100W (with ordinary speakers)

Most speakers with this amp will be in class AB virtually all the time because it will be delivering more than 1/100th watt. Headphones can have very high sensitivity (mine are 102dB with 0.001W), so the class A area of operation under 1/2W for this amp would mean class A all the time.
 
I use a precision voltage divider that I built, off of the receiver's speaker terminals for my Sennheiser 580's and have used it for years. It's much better sounding than the standard headphone jack it came equipped with and it allows me to adjust the volume with the volume control at normal (speaker) settings.

BTW: I'm not too sold on the whole idea of how "valuable" damping factor is on any amplifier. It's more the result (a byproduct actually) of negative feedback than anything and often has a detrimental effect on bass extension. I believe that Nelson Pass has written extensively about this over on the diyaudio forum.

Best Regards,
TerryO
 
Headphone connected to amp outputs

Hey Guy's
Very easy solution...Divide the headphone impedance by the amplifer output resistance will give you the resistor to use in parallel with each driver connection....so 600 ohms / 8 ohm amp resistance gives you 75 ohm...put across headphone - and + and take reading with ohm meter and you get 8 ohms...as for the volume level problem get a stereo linear 8 ohm L Pad for tweeter levels and connect to amp and outputs of L Pad to headphone jack with your 75 ohm 1/2 watt resistors in parallel with headphones connections and you have your level control so you can get the amp volume off of its seat.
I suggest set you L Pad 1/4 way for a start but you want your amps volume 1/2 way for the best accuracy the use your L Pad for your absolute max level you can stand. But this is how you over come the impedance and level problem from using a power amp, and frequency response is not altered... Good Luck
John
 
You must log in or register to reply here.
Back
Top Bottom