[runeight]

I'm sure that many folks on this forum know the answer to this, but I'm discovering pretty quickly that I don't.

I want to know what the maximum current is that a particular pair of headphones can handle. I think I can figure this out for headphones where the spec gives the maximum power. Assuming that this means maximum continous power.

But the other thing I want to know is what is the maximum instantaneous current that can be applied. And over what period of time before things are damaged. I know that this is an esoteric question, but I'd like to have some sense of this if I can.

If there is a website please point me that way. I am happy to read.

Thanks much for the help.

 

[digger945]

I too would like to understand the answer to this dynamic question

.

 

[pabbi1]

I smell sulfur... is this thread worthless without pics?

Sounds like the Bat light needs to be truned on for Pete Millett and Dr Gilmore.

 

[runeight]

Now that's a good idea. I hadn't thought of a thread just for pics of destroyed headphones. But, before we do that, if I could just get a tiny bit of help on this . . .

Don't know that either of those two gentlemen would know more than some other folks who visit this thread. I'm happy to hear from whoever has insight into this.

 

[nikongod]

Quote:

Originally Posted by runeight /img/forum/go_quote.gif I'm sure that many folks on this forum know the answer to this, but I'm discovering pretty quickly that I don't.

you are probably less alone than you think.... Im not sure I really have any solid answers, but I have some good things to think about.
Quote:

I want to know what the maximum current is that a particular pair of headphones can handle. I think I can figure this out for headphones where the spec gives the maximum power. Assuming that this means maximum continous power.

as an exercise of estimates, this is probably the easiest way. There are very few standards for headphones. My own assumption is that most power ratings on headphones are little more than penis-waving show-points. They can probably back these claims up, but does any of that matter? Hey look at him, his "driver" can handle 1W (and its 75mm diameter too)! Too bad he has no SQ to back it up... i'l stop before they ban me.

You will probably be more interested to just see how much power YOU use. hook up whatever headphones you like, and go to an average level. measure a few songs with a AC volt-meter (its very helpful if it has max level memory and even better if it shows RMS as well as peek-peek values). calculator away.

If you are concerned with estimating when a given amp will drop out of class-A (or hit the power rails...), use peek values. If you are just curious with average current use RMS values.
Quote:

But the other thing I want to know is what is the maximum instantaneous current that can be applied. And over what period of time before things are damaged. I know that this is an esoteric question, but I'd like to have some sense of this if I can.

this will more likely come down to the size of the wire IN the driver.

The smallest I could find a rating for was 40ga: 0.014A some headphones may use larger, but thats probably a good starting point.
source
If you want to see what that does to a headphone put 14ma(RMS) into a set of koss KSC-75.

 

[runeight]

Thanks nikongod. I should have crafted my question better, but the second half of your answer is beginning to get at what I'm looking for. I am interested in what's needed to protect headphones from high currents resulting from failure conditions as opposed to listening levels.

Here's more of what's on my mind.

Speaker protection can often use fuses because speakers can handle fairly high currents for fairly long periods of time before the voice coils fry. In this case there is plenty of time for a fuse to blow before the coils are damaged (unless the current is extraordinarily high).

Headphones can't be protected with fuses because if the current is high enough the fuse may not blow fast enough to prevent damage.

We can use electro-mechanical devices (such as the e12) to help with this. But even here, if the transient is fast enough the offset detector may not have enough speed. IIRC the e12 operates on the 10's of ms timescale electrically and that further delayed by how long it takes the relay to flip the contacts once the voltage is applied. Other offset detectors have to work on similar principles because if they are too fast they will trip from ordinary audio signals. So the integration time constant has to be long enough for the DC offset to really be of significance but short enough to actually protect the headphones. This works very well for "slow" transients.

However, let's make up a hypothetical case for discussion. Suppose a 32V transient is applied to 32R headphones. At full voltage the headphones will be sustaining 1A. There is a rise time and fall time. How long can this transient last before the coils are damaged? 1us? 10us? 1ms? Etc. This number is different from the amount of continous current that a particular wire gauge can pass (although those number are extremely helpful).

I realize that there are many conditions here that are not specified that determine the answers, but I am hoping to find some handle to start with.


Edit: Another aspect of this is damage to the diaphragm. If the current is really high it will pull the diaphagm very hard in one direction. When is the diaphragm permanently warped by this? If ever?

 

[holland]

I don't think anyone will ever know the answer to this, and I'm not even sure if headphone manufacturers know or will divulge that information.

People rely on ears to say whether or not something is damaged, where the damage may begin much sooner than what is audibly detectable.

I honestly would start with a vendor, if you can get in touch with an actual engineer, that would be best. Someone from Beyerdynamic or Sennheiser may be more open to discussion.

AFAIK, Beyerdynamic has some of the lower power ratings for headphones at a max capability of 100mW @ 80ohm.

From some of the threads, particularly on amps with cap coupled outputs, it's been stated that it's fine to leave the headphones in on power on/off. If you figure an output cap of 470uF and a load of 32 ohm (some of that is dissipated by the load resistor as well). At worst case, you will deal with the peak charge currents exponentially decreasing as the capacitor charges. With a 470uF and 32-ohm and a SSMH peak voltage of about 19.6, you'll probably look at a peak current of about 600mA @ t=0 and exponentially decreasing till about 12 seconds out where it reaches quiescence. Again, need to take into account the load resistor to ground in addition to the headphone load, but usually it is such that the bulk of the current goes to the headphones.

This in no way indicates the max, but can definitely be a starting point.

 

[amb]

There isn't going to be a simple answer. There are many variables, such as the wire gauge of the voice coil, the shape of the voice coil wire (both cross-sectionally and in the way it's arranged), the diameter of the voice coil bobbin and its heat dissipation capability, etc. As for diaphragm or suspension damage, it also depends on their design and material.

This is not much different than, say, a speaker tweeter, except at a different scale and impedance. But the same variabilities exist. For example, two seemingly similar 1" dome tweeters could have very difference power handling capabilities (and therefore maximum current capacity) due to the design. A higher powered unit might have an aluminum voice coil bobbin, with hexagonal voice coil wires wound over a fairly large area, suspended in ferrofluid which not only damps the resonance but transfers heat to the magnetc structure...

 

[nikongod]

Quote:

Originally Posted by runeight /img/forum/go_quote.gif We can use electro-mechanical devices (such as the e12) to help with this. But even here, if the transient is fast enough the offset detector may not have enough speed. IIRC the e12 operates on the 10's of mV scale electrically and that further delayed by how long it takes the relay to flip the contacts once the voltage is applied. Other offset detectors have to work on similar principles because if they are too fast they will trip from ordinary audio signals. So the integration time constant has to be long enough for the DC offset to really be of significance but short enough to actually protect the headphones. This works very well for "slow" transients.

NOTE! this is all my own "gut" instinct and should be checked up on if you want really concrete info... If anyone does more research on this, I would happily be proven wrong (or better right!) to better the community.
I dont think that the transient is the problem, even the spike up to high voltages and instantaneous currents higher than what the size of the voice coil wire would suggest. At some point, I agree, the current from the offset will burn through the actual voice-coil wire basically instantly, but I think in real-world failures it will take longer than most people expect. There have been several threads throughout the ages where people have "observed" heat coming off of a driver that was exposed to DC. The drivers later produce "some" sound, and impedance measures OK, but the sound is clearly off.

I am (personally) more concerned with long-term exposure to even lower DC offset and the fact that it will eventually melt the plastic or glue that holds everything together, OR soften the plastic of the "cone" and then it can warp (see more below). At lower offset levels, this will take more time AND voltage offset than it takes the e22 to trip so you are well protected. It is possible that the offset is so bad, and the current so high that the voice coil basically burns up instantly, but even in the 32V/1A into a grado example, you may still be OK (I dont really know) while you wait for the e22 to trip. Your amp is TOAST even if the headphones are ok...
Quote:

However, let's make up a hypothetical case for discussion. Suppose a 32V transient is applied to 32R headphones. At full voltage the headphones will be sustaining 1A. There is a rise time and fall time. How long can this transient last before the coils are damaged? 1us? 10us? 1ms? Etc. This number is different from the amount of continous current that a particular wire gauge can pass (although those number are extremely helpful).

As another point, the 32V offset would only happen if a transistor *suddenly* failed in a DC coupled amp which is possible but rare. Historically, the damage is done by a low-voltage long term problem. part drift in an amp with no DC servo, servo lag-time (EG: an amp whose DC servo dosnt "kick in" for about 20-30 seconds) or an amp that drifts beyond what the DC servo can accommodate due to over-heating or some other part-drift. thermal-run-away anyone? In these cases, the offset may be a few-hundred mV which can destroy a headphone if its allowed to remain for more than a few moments, but if removed quickly should pose no long term harm to the headphone.
Quote:

Edit: Another aspect of this is damage to the diaphragm. If the current is really high it will pull the diaphagm very hard in one direction. When is the diaphragm permanently warped by this? If ever?

You can wrinkle some drivers with excessive SPL, but most simply bounce back to their original form. It is possible to "heat" them just enough to warp them with long-term mild DC offset though.