[nwavguy]

Quote:

 
I posted exact cut and paste copies of what you posted and you have failed to respond.  You seem to have an axe to grind.  Why did you make those charges and then fail to back them up?
 

I'm trying to keep this focused on what matters. In post #46 you said "let's stick to the known facts", and in post #48 you said "propagation of hysteria never helps anything" and I'm only trying to oblige. In post #55 you quoted several statements of mine which are clearly my opinion and then asked me for objective proof to back them up. One could argue that's encouraging hysteria rather than discouraging it. I can defend the more objective statements, and I'll be happy to, but you and others here would just accuse me of "trashing" Schiit. So I think, for now, we really should focus on the OP and main topic of this thread rather than trying to discredit me. The questions at hand are: What's the peak voltage of the power off transient into a typical load? Is the Schiit Asgard hard on headphones?
 

 

[kwkarth]

Quote:

I'm trying to keep this focused on what matters. In post #46 you said "let's stick to the known facts", and in post #48 you said "propagation of hysteria never helps anything" and I'm only trying to oblige. In post #55 you quoted several statements of mine which are clearly my opinion and then asked me for objective proof to back them up. One could argue that's encouraging hysteria rather than discouraging it. I can defend the more objective statements, and I'll be happy to, but you and others here would just accuse me of "trashing" Schiit. So I think, for now, we really should focus on the OP and main topic of this thread rather than trying to discredit me. The questions at hand are: What's the peak voltage of the power off transient into a typical load? Is the Schiit Asgard hard on headphones?

Well, I guess if you could un-say all the unsubstantiated negative things you've said in this thread, we could get back to a modicum of objectivity, couldn't we. 

 
 

 

[Prog Rock Man]

Kwkarth please test your headphones as requested and hopefully we will then find out about the state of Shike's amp from Shiit Audio and then the comments can be made.

 

[kwkarth]

Ok, here are my findings:
 
 

The diaphragm of my K-701 deformed similarly to the video in the thread.  I measured the right channel only.

 

I used one of my Fluke meters set to trap min & max potentials.  After visually testing the K701 diaphragm movement, I measured electrically using a pair of LCD-2s, so the load was 50 ohms resistive, again, right channel only.

 

turn off

-2.2 VDC peak

0.584 VAC peak

 

turn on

1.112vdc peak

0.332vac peak

 

I repeated measurements several times to verify results.

 

The voltage measurements represent just under 100mW of power for less than 0.25 seconds.  That's not enough energy to harm any full size headphone I know, but I wouldn't recommend this with sensitive IEM's.

 

[deadlylover]

Wow, thanks for doing the test, the results are very interesting.
 
Fluke's are pretty good, I don't recall if they were immune to overshoot readings though, not many multimeters are, we really need to get one of these under a scope.

 

[nwavguy]

 
Quote:

Ok, here are my findings:
 
The diaphragm of my K-701 deformed similarly to the video in the thread.  I measured the right channel only.
 
I used one of my Fluke meters set to trap min & max potentials.  After visually testing the K701 diaphragm movement, I measured electrically using a pair of LCD-2s, so the load was 50 ohms resistive, again, right channel only.
 
turn off
-2.2 VDC peak
0.584 VAC peak
 
turn on
1.112vdc peak
0.332vac peak
 
I repeated measurements several times to verify results.
 
The voltage measurements represent just under 100mW of power for less than 0.25 seconds.  That's not enough energy to harm any full size headphone I know, but I wouldn't recommend this with sensitive IEM's.
 

 

Thanks kwkarth. That's very helpful and confirms my suspicions that it had to be far higher than a few tenths of a volt. I do, however, want to stress the issue here isn't thermal damage but rather mechanical damage. Because the power off transient is so prolonged it's effectively acting like DC to the headphone driver and, at least visually, that's severely stressing the driver's diaphragm and suspension.
 
Headphones usually have both thermal and mechanical limits to how much signal they can handle. In this case, 100 mW for 0.25 seconds isn't a problem for the AKG's thermal limits, but the DC-like nature of the transient is an apparent problem for its mechanical limits.
 
It would still be best for a headphone manufacture or two to weigh in on this issue. My personal opinion is this is probably OK for some headphones but may be risky for many including the full size AKG K 702s. As I said at the start, the damage done to the diaphragm may not be readily apparent in listening to them.
 
My concern is something like the cardboard like you find on the back of a notepad. Unmolested it has a fair amount of rigidity. But if you crease and fold it, it now has weak areas and much less rigidity. It wants to bend at the creases even after you try to straighten it back out.
 
As I said, I know some tweeter domes can be weakened in a similar way and then measure differently. I don't know if it's also true of headphone diaphragms. But I suspect the answer will be specific to different headphones, the material of the diaphragm, the nature of the suspension, etc. So there's likely no easy answer here about what's safe and what isn't.

 

[kwkarth]

Quote:

Thanks kwkarth. That's very helpful and confirms my suspicions that it had to be far higher than a few tenths of a volt. I do, however, want to stress the issue here isn't thermal damage but rather mechanical damage. Because the power off transient is so prolonged it's effectively acting like DC to the headphone driver and, at least visually, that's severely stressing the driver's diaphragm and suspension.
 
Headphones usually have both thermal and mechanical limits to how much signal they can handle. In this case, 100 mW for 0.25 seconds isn't a problem for the AKG's thermal limits, but the DC-like nature of the transient is an apparent problem for its mechanical limits.
 
It would still be best for a headphone manufacture or two to weigh in on this issue. My personal opinion is this is probably OK for some headphones but may be risky for many including the full size AKG K 702s. As I said at the start, the damage done to the diaphragm may not be readily apparent in listening to them.
 
My concern is something like the cardboard like you find on the back of a notepad. Unmolested it has a fair amount of rigidity. But if you crease and fold it, it now has weak areas and much less rigidity. It wants to bend at the creases even after you try to straighten it back out.
 
As I said, I know some tweeter domes can be weakened in a similar way and then measure differently. I don't know if it's also true of headphone diaphragms. But I suspect the answer will be specific to different headphones, the material of the diaphragm, the nature of the suspension, etc. So there's likely no easy answer here about what's safe and what isn't.

 
In the case of the K-701, I do not feel there is any damage done to the driver.  If one were to view the driver in action using laser interferometry, one would likely see the same non linearity exhibited.  The driver excursion travel would be the same distance, regardless of the period of the signal applied.  97mW is 97mW regardless of the frequency or duty cycle, within thermal constraints of course, with which we seem to be in agreement.  The driver by design has a flexible diaphragm whose stiffness and mass are entirely dictated by its thickness which varies from edge to center of the driver.  This is the nature of the AKG varimotion driver.
 

 

[kwkarth]

Quote:

Wow, thanks for doing the test, the results are very interesting.
 
Fluke's are pretty good, I don't recall if they were immune to overshoot readings though, not many multimeters are, we really need to get one of these under a scope.

I don't have a working scope at home right now, so that will have to wait for someone else.  One can be pretty sure that the peaks I measured, are pretty much worst case.  
 
 

 

[nwavguy]

Quote:

In the case of the K-701, I do not feel there is any damage done to the driver.  If one were to view the driver in action using laser interferometry, one would likely see the same non linearity exhibited.  The driver excursion travel would be the same distance, regardless of the period of the signal applied.  97mW is 97mW regardless of the frequency or duty cycle, within thermal constraints of course, with which we seem to be in agreement.  The driver by design has a flexible diaphragm whose stiffness and mass are entirely dictated by its thickness which varies from edge to center of the driver.  This is the nature of the AKG varimotion driver.
 

You have the advantage of having the K 701 in front of you and I can only go by what I've seen in the video with the K 702. But the "crinkling" is a lot like wrinkling a piece of paper, or say a mylar balloon--it's very conceivable it might create weak points in the material.
 
As for the driver travel, the excursion increases as frequency decreases to produce the same power level. Air provides an acoustic "load" on the driver's diaphragm that varies with frequency. You can easily demonstrate this operating a driver (like a small woofer or full range driver) in free air and doing a frequency sweep using a stationary object to measure the excursion. It's highest at the lowest frequencies, and highest of all, at DC.
 
So 2.2 volts peak at 1 Khz is very different from 2.2 volts peak at 10 hertz or DC in terms of driver excursion, and hence, mechanical damage. For anyone who's skeptical, here's an online calculator for excursion. Just lower the frequency and watch the Xmax number go up to produce the same acoustic power:
 
http://www.baudline.com/erik/bass/xmaxer.html
 
 

 

[kwkarth]

Quote:

You have the advantage of having the K 701 in front of you and I can only go by what I've seen in the video with the K 702. But the "crinkling" is a lot like wrinkling a piece of paper, or say a mylar balloon--it's very conceivable it might create weak points in the material.
 
As for the driver travel, the excursion increases as frequency decreases to produce the same power level. Air provides an acoustic "load" on the driver's diaphragm that varies with frequency. You can easily demonstrate this operating a driver (like a small woofer or full range driver) in free air and doing a frequency sweep using a stationary object to measure the excursion. It's highest at the lowest frequencies, and highest of all, at DC.
 
So 2.2 volts peak at 1 Khz is very different from 2.2 volts peak at 10 hertz or DC in terms of driver excursion, and hence, mechanical damage. For anyone who's skeptical, here's an online calculator for excursion. Just lower the frequency and watch the Xmax number go up to produce the same acoustic power:
 
http://www.baudline.com/erik/bass/xmaxer.html

Please, spare me the primer on excursion.  I understand all of that, but let's not confuse the masses.  This situation we have here is worst case.  The driver did not "krinkle," it was a soft deformation throughout the process.  That's why I said the diaphragm was flexible. No hard edges.  I appreciate you saying to want to be a harbinger and protector of the truth.  Now I think it's time for you to demonstrate more doing and less opinionated words.  BTW, your little demo link does not lend itself to this discussion because it is used to calculate x max at a particular SPL.  That does not enter into our equation here.  It's not a factor at all.  Your example is non sequitur, and actually quite misleading in this case.
 

 

[Jason Stoddard]

Hey all,
 
I jumped on this and did our own measurements after KW did his. And yes, I can confirm: I was mistaken about the magnitude of the turn-off transient. That's what I get for looking at engineering notebooks over the weekend, rather than going into the office and actually testing. Our peak was about 1.7V at 32 ohms, or about 90 mW. 
 
We have never had anyone complain about driver failure or sonic degradation of their headphones with the Asgard amp since we started shipping. We believe it is generally safe for most headphones. We do not recommend any of our amps with IEMs, and have always stated this. 
 
However, we do understand if you're concerned. So here's what we'll do:
 
1. We'll add a foreward to the owner's manual recommending to connect sensitive headphones after turning the amp on, and unplugging the headphones before turning the amp off. We'll also add an addendum to the warranty to cover shipping, parts, and labor for any headphone jack failures during the warranty period, to address any possible wear and tear on the jack.
 
2. We will investigate adding a relay mute to the Asgard, despite my misgivings about relay contacts.
 
And, to address any other concerns our owners might have:
 
3. If you purchased an Asgard from us at any time, from June 1, 2010 to the current date, and feel uncomfortable about its performance, contact us and we will take your Asgard back and offer you a full refund of the retail price. 
 
All the best,
Jason

 

[JamesMcProgger]

soft or not I'll panic if i see the drivers of my beloved K601 doing that ****. now, is this standard operation in the asgard or a malfuntion?
 
BTW great and promt replies from The Chiit Audio guys, that is not very common and nice to see.

 

[nwavguy]

Quote:

 
Please, spare me the primer on excursion.  I understand all of that, but let's not confuse the masses.  This situation we have here is worst case.  The driver did not "krinkle," it was a soft deformation throughout the process.  That's why I said the diaphragm was flexible. No hard edges.  I appreciate you saying to want to be a harbinger and protector of the truth.  Now I think it's time for you to demonstrate more doing and less opinionated words.  BTW, your little demo link does not lend itself to this discussion because it is used to calculate x max at a particular SPL.  That does not enter into our equation here.  It's not a factor at all.  Your example is non sequitur, and actually quite misleading in this case.
 

 
I'm not trying to confuse the masses. I'm trying to be clear about potentially misleading statements you have made:
 
 "The driver excursion travel would be the same distance, regardless of the period of the signal applied."
 
The above is simply not true. That's why I provided the information on excursion vs frequency. My "little demo link" is valid and applicable. With a reasonably flat frequency response a constant input power gives a constant output power regardless of frequency. So SPL is directly proportional to voltage in that calculation.
 
And this generalization of yours is also misleading:
 
"100mW of power for less than 0.25 seconds.  That's not enough energy to harm any full size headphone I know."
 
It's not the energy in terms of thermal heating that's the problem here. It's how much voltage does it take to reach the driver's safe excursion limits (technically known as Xmech). And that's going to be very headphone specific so it's rather impossible to say it's not enough to cause harm across a wide range of headphones.
 

 

[Shike]

I hate to add bad news, but I feel I have a duty to say this:
 
Heard back from AKG, and they said the 250mv DC I was measuring during on/off was already considered potentially damaging (unfortunately, it seems it was much higher than I could catch with my bare eyes on a DMM).  As such for those considering the Asgard make sure you're willing to perform the power on/off cycle religiously.
 
@Jason:
 
Thanks for looking into this and being understanding.  I'll definitely be interested into looking at future Schiit amps if a relay or another form of headphone protection is added.  Unfortunately the amp was primarily for the K702's, had it been for other headphones I may not have even noticed.
 
I must say I'm impressed how you're handling it at this point

 
@JamesMcProgger:
 
It's normal it seems, though AKG technical/support didn't like the numbers I sent them as is (which were lower than reality).  If you consider the current Asgard it would be best to pay attention to what Jason listed under one.

 

[kwkarth]

Quote:

I'm not trying to confuse the masses. I'm trying to be clear about potentially misleading statements you have made:
 
 "The driver excursion travel would be the same distance, regardless of the period of the signal applied."
 
The above is simply not true. That's why I provided the information on excursion vs frequency. My "little demo link" is valid and applicable. With a reasonably flat frequency response a constant input power gives a constant output power regardless of frequency. So SPL is directly proportional to voltage in that calculation.
 
And this generalization of yours is also misleading:
 
"100mW of power for less than 0.25 seconds.  That's not enough energy to harm any full size headphone I know."
 
It's not the energy in terms of thermal heating that's the problem here. It's how much voltage does it take to reach the driver's safe excursion limits (technically known as Xmech). And that's going to be very headphone specific so it's rather impossible to say it's not enough to cause harm across a wide range of headphones.

I can't believe it, you seem to be unable to understand the function of your own provided link.  Since you seem to be unwilling or unable to understand what I'm saying, or even what you're saying for that matter, we'll have to agree to disagree.  I will call you out for any more misleading things you say, however.