[dBel84]

sadly ( or thankfully ) tubes do not behave like chips. The SOHA has a transient spike of 12V at either power up or down - hence the warnings to unplug hp's or add a dc sensing relay. I have played with el cheapo iems and they have not been scorched by the transients but some fine headphones have fallen casualty to the tube transients , so .... it may not be in your head but only you can tell that for sure..dB

 

[Paragon]

ah.. tube amp.. I have been away from this stuff way too long.

 

[dBel84]

 

[fortney]

I may be wrong here, of course--

Is not the damage from DC offset to headphones a matter of overheating of internal elements due to the diaphragm (or whatever) being held in one direction from the DC? If so damage would be dependent on time of DC flow in addition to level of DC.

The on/off voltage spike while not instantaneous is a matter of a few seconds of decaying voltage. It can be loud, but that is not a headphone issue.

F

 

[fran]

I would think thats its a combination of the actual voltage, the current too and the time involved.

In my case I think the current is really low so I just wonder if it's really any harm at all since although the voltage spikes quite high, it seems to be really low current.

I would like to put some figures on the current, but lack the knowledge how. Would anyone here know how best to test the current available during this spike?

Fran

EDIT: Just to be clear, there are 2 types of DC offset being discussed here at random.

1. the DC spike that occurs on turn-on of some tube amp, DC can rise to 5v and drop to low levels within say 15secs.
2. Constant DC offset that is there all the time the amp is playing.

No. 2 is not an issue for me (right now), but no. 1 does concern me - even though its transient could it damage my cans?

 

[majkel]

You need pretty high DC offset to cause coil overheating. To blow 32 Ohm Grado's rated say 150mW max, you need (without music signal applied) 2.2V of DC which you'd realize quickly by hearing loud click while plugging in/out or turning on the device, as well as distortotions would occur if the music played.

 

[dBel84]

Yes, we somehow got onto dc transients which has very little to do with the offset behaviour of the amp.

As for damaging cans? I know of one HD600 driver that went up in smoke but I am guessing it has a lot to do with the time, specific drivers, and many more. If this helps the case at all I will admit to being a daft fool and applying 12V to my grado SR60's for at least 1 min - I had breadboarded a PPA and somehow got the output shorted to the pos rail ! Noticed the sound was off immediately but it took for the coil to become hot enough to burn my ear before I yanked the power. The headphones survived and I think that particular driver sounded better than the stock

I have tested the transient spike on various el cheapo cans but I have ( as yet ) not exposed my decent cans to such abuse..dB

 

[kvant]

Quote:

Originally Posted by majkel /img/forum/go_quote.gif ... 2.2V of DC which you'd realize quickly by hearing loud click while plugging in/out or turning on the device,

During turning on the device, not necessarily. The transients can be slow, subsonic. You could hear just about nothing, while your phones are still taking abuse. I have experienced this with some experiments with a servo. This is one of the dangers of designs with DC coupled outputs. SOHA is AC coupled in between the tube and the buffer, but I guess the time constant (C times the input impedance of the buffer) there is pretty high, so that a lot of low frequency mess passes through. And the powerful buffer will happily supply as much current as the voltage requires. With a cap on output, the time constant is typically smaller, so that slow transients associated with heating up the tube are blocked and everybody is happy.

 

[Paragon]

If it makes anyone feel any better.. I have been using my Grados through my computer speaker headphone out (Klipsch Promedia) and while I had all my equipment out, I decided to test the DC output. one channel was around 25mV.. the other was around 35mV and goes up to around 50mV with loud volume. Lately, I usually have my cans plugges in all the time but turn the volume down at night but other than that, pretty much all of my listening has been through that jack.

Cans sound great still. Have never noticed any problems.

I'm stringing my pocket amp in to block the DC for now just because. Might open the control pod up and put in a few caps.. if I can find some that fit.

 

[tomb]

Quote:

Originally Posted by kvant /img/forum/go_quote.gif During turning on the device, not necessarily. The transients can be slow, subsonic. You could hear just about nothing, while your phones are still taking abuse. I have experienced this with some experiments with a servo. This is one of the dangers of designs with DC coupled outputs. SOHA is AC coupled in between the tube and the buffer, but I guess the time constant (C times the input impedance of the buffer) there is pretty high, so that a lot of low frequency mess passes through. And the powerful buffer will happily supply as much current as the voltage requires. With a cap on output, the time constant is typically smaller, so that slow transients associated with heating up the tube are blocked and everybody is happy.

Interesting ... and I once got slammed for claiming "the coupling caps on the SOHA weren't big enough to block all of the DC on startup." I probably should've clarified it by stating headphones - or even some DMM's - don't much care if the transient is 2Hz AC instead of 0Hz DC, but you've stated it much better. The point is, either one will potentially damage headphones on startup with the SOHA - and it lasts quite a few seconds longer than one would think it takes to charge a couple of very small 0.1uf film caps. As you also state, the buffer is amplifying those transients all the while.

In your reference about DC coupling: on the Millett and MAX, the output sits right at the halfway point of the supply voltage difference (24-27VDC, typically). Without the large electrolytic caps, the headphones will see a full 12VDC or more. On startup, the critical time is in fully charging the large caps, which takes some 5-10 seconds (closer to 5 seconds, thankfully), depending on various factors. So if they're plugged in, the headphones are exposed to that full voltage on startup, then it steadily decreases as the caps charge.

Note that I am not stating a preference either way - there are advantages and disadvantages with either method. It's just best to have eyes wide open about the potential consequences of each.

(Plus, I saw an opportunity to redeem myself slightly on an earlier posting incident.

)