No, it only came up after switching my headphones from the tube amp jack to the RME jack. I follow carefully the advised procedure for tube amps (mute system, volume 0, unplug headphone, turn off amp, then plug in to RME).

Thank you very much for your responses.

Your answer was clear, but at the risk of insistence, you are sure nothing could be happening via the RCA cables from the tube amp to the RME ADI-2? Could current somehow be leaking that way either while the tube amp is running or after I turn it off? I know that RCA cables are clearly not designed for that, but technically it could happen?

I am only double-checking because this has ONLY happened on the two instances (months apart) when I decided my tube listening session was good for the day and it was time to switch to the RME amp.

As for warranty, I can't believe it is only 6 months from the manufacturer. Even with the extension of that to two years in my country, I am out of warranty since I bought it in December 2018. But the unit has been running flawlessly other than this.

Thanks for the replies. The source is optical in via Asus Essence STX II card. No constant level in Line Out mode in the lowest band.

I guess I will just use it without the tube amp to be safe, or I will always wait a good 30 minutes between switching my headphones from the tube amp to the ADI-2 amp.

I have had this exact problem today with my brand new RME ADI-2 DAC FS that arrived today. I received the message "DC detected. Phones deactivated, Pull out Phones plug to reset output state" and I thought how could this happen. I use a Beyerdynamic DT880 at 600ohm impedance and Hi Power setting on the dac with about -8dBr and windows sound to max.

As the error occured more frequently during my testing I thought that I need to do a research to find what was causing this error. I experienced the error at most when I played a song in VLC Player and then pressed "Stop" button in the player. As Matthias said above I noticed that on the lowest band in the spectral analyzer the green bar did not move down to zero but instead kept holding it's value. I unplugged several cables to see when the bar was going down. After I unplugged the SPDIF optical cable the bar immediately went down to zero.

Can SPDIF Optical transport a DC to an external device?

To test this issue further I unplugged the SPDIF Optical from my soundcard which is a Xonar Essence STX that has been running for quite some years in the PC and plugged the optical cable into the SPDIF socket of my mainboard instead. After I did so I have not yet received DC inputs to the Dac. I would assume that my xonar essence is slowly dying?? Could this be the case?

Also last time I checked the soundcard's headphone amp the sound was really heavily distorted. It was horrible...

Hopefully my research has fixed this error finally and I hope that my ADI-2 did not get any damage from the DC input.

Since I am not too deep into the technical aspect of sound enginerring, could anybody confirm that my assumption was right? And the soundcard was the one to blame here?

Take a look at the left most bar in your spectrum analyzer. If you see it having a high value, then listening with a high volume will trigger dc protection.

Sometimes music are not mastered correctly. So it’s not ADI-2’s fault. The design works as intended to protect your headphones. Your music is the problem.

ADI-2 is not perfect though. The DC protection does not work in balanced phone mode at all if the music has DC. I never tried it but based on my understanding of the circuit that should be the case. I hope MC could put a warning in the manual if that is the case.

Right, that's the same conclusion I reached after looking at the circuit. It just happened that DC from music will be doubled in the balanced headphone output architecture but get canceled out in the detection circuit. So it's not protected against ill-mastered music.

I hope MC either document that in the manual, or implement this protection in software.

Even “ill-mastered“ music does not contain enough DC to be of any concern, or to be capable of breaking headphones.
At least I’ve never seen such that is triggering ADI-2’s DC protection.

The case described by liquid_mercury is clearly hardware/driver related, as the same music played through another path does work.

I think not because different speakers of different headphones have different levels of efficiency and thus loudness to the ear.

Ramses and MC. There surely are albums that is mixed very poorly. Just Google and you will find quite a few notorious ones. See for instance this post:https://www.audiosciencereview.com/foru … ost-776805 and this follow up https://www.audiosciencereview.com/foru … ost-776892

This track can successfully put ADI-2 into protection mode for single ended headphone output. It’s easy to guard against these with software in balanced mode and it won’t take too much effort — as you mentioned here, simple calculation is enough.

Oh sorry ramses I misread.  was supposed to reply KaiS and MC. My apologies . I was reading using my smartphone and the screen is too small.

To cope with DC in recordings one could simply add a 20Hz HighPass in the EQ section, DC and useless ultra low frequency removed.

The track has a huge (and varying) amount of of DC, up to -9 dBFS, even streamed at Tidal,.
Seems you can find all types of stupidity if you only search enough.

Why built a compensation for stupidity?

The implementation of DC removal is by no means trivial (aside from a 2 Hz highpass filter).
I.e. on the mentioned track at 0’30” -13dBFS DC kicks in, at 0’51” DC changes to -9 dBFS,  2’30” the DC suddenly disappears.
How to handle this?

That's not correct. The left bar in the Analyzer is the whole band from 0 to 25 Hz. Mentioned in the manual. It's on near full level on mine most of the time (I like music with a lot of subbass).

The 'DC' in the example catastrophy (unbelievable) is part of the used bass synth. It's indeed a DC error as long as the bass is active.

> that occupies one band of eq.

Sure. But High Pass filter in the PEQ has been added especially for such use cases, so that is no argument in my view.

This flat top audio wave is (steady) DC, but only for a short moment, until it sways back to the opposite polarity.

Same is true for EVERY half-wave of AC (audio).
So in fact it’s not, as you have to watch longer periods of signal to detect DC, at least 0.5-1s for the audio band.

Asymetrical (one sided) clipping causes some amounts of DC with asymmetrical waveforms, typically voice or brass instruments.

But you have to keep in mind, clipping is an untypical state for audio.

> So, too loud a volume which creates/prolongs the cliping is not a healthy thing, (or in other words, a steady tone which is too high and last more than 0.5-1s is no good) right?

Clipping will make the signal distorted so it is not good --- it does not sound good in the first place.

> Or, may be the RME DC protection does provide further function not only for clipping?

Clipping has nothing to do with DC. They are two different things... Having a too high steady tone (let's say 50kz sine wave with no DC offset) that last long will not produce DC --- you get square wave instead of DC. The DC offset of that square wave is still 0.

Fried headphone amplifier opamp will also produce DC. Depending on which op amp is broken the result may be different. There is one 1688 driving the other 5. If it's one of the 5, the others can correct it back (but produce a lot of heat, which will trigger overcurrent protection discussed in the next paragraph). but if it's the first one, the whole circuit will produce full DC output and the protection circuit will shut off the output immediately.

RME does offer over current protection. There are thermistors inside the unit (one per rail, so two per channel) which measure the temperature and voltage in real time. There are 8 (in Pro) or 4 (in DAC) channels of comparators working simultaneously, and will either lower your volume or stop the output when you try to play too high level for low impedance headphones.

You haven’t got me right.
You asked, so my point was to explain how DC can be generated from clipping an audio signal.

Clipping per se hits at a level limit.
It will not damage ADI-2, but being loud enough it can destroy headphones, speakers, and you valuable hearing, with or without extra DC.
And hey - it doesn’t sound good usually!

High value of DC is a problem for headphones and speakers, as the voice coil is heated by the current, but not moving, so there is less air flow cooling happening.

ADI-2 Pro’s DC Protection:
According to my measurements ADI-2’s DC protection protects connected headphones from internal component failure, but not 100% from signals like the stupid track mentioned.

I did NOT measure by connecting headphones , but using load resistors and the music track - BLOC PARTY Blue Light (Engineers ‘Anti-Gravity’ Mix) - from 1’00”.
https://tidal.com/track/86125589

Measurement of ADI-2 Pro’s DC Protection characteristics:
Load/Ohm •  DC/V • Time/s •  dBr • Power/mW

No Load • 0.9V • 15s • -14,5 dBr • 0.00mW

47R • 2.4V •  2 s • -6.0 dBr • 122mW
47R • 1.5V •  4 s • -10.0 dBr • 48 mW
47R • 0.9V • 15s • -14.5 dBr • 17 mW
47R • 0.8V • 52s • -15.5 dBr
47R • 0.7V • >100s • -16.5 dBr • 10 mW

16R • 2.4V •  2 s •  -6.0 dBr • 360 mW
16R • 1.5V •  4 s • -10.0 dBr • 140 mW
16R • 0.9V • 15s • -14.5 dBr
16R • 0.8V • 52s • -15.5 dBr

Conclusion:
Most 47 Ohms headphones can stand 122 mW for 2 s, but some more delicate, i.e. 16 Ohms IEMs, can’t stand 360 mW for 2s.
ADI-2 Pro’s DC Protection is DC-Voltage/Time dependent only, the connected headphones impedance has no influence.

KaiS is right in “ ADI-2 Pro’s DC Protection is DC-Voltage/Time dependent only, the connected headphones impedance has no influence.”

It works by charging a capacitor. When it has enough voltage the three bjts (to make sure it works for both positive and negative) next to it will trigger the protection.  So takes time and enough voltage. The circuit is not connected to the headphone attached so headphone impedance does not matter.

Also, for testing, you don't need that track mentioned earlier in the post. Any DAW or audio programming platform can give you pure DC easily. For instance, in SuperCollider, "{Pulse.ar(0)}.play;" can give you full amplitude DC.
in addition to that, "{Pulse.ar(0.0)*[1,-1]}.play;" should give you full positive amplitude DC in the left channel and full negative amplitude DC in the right channel => A signal that specifically crafted to make DC protection not able to handle.

Haha calm down, KaiS probably just want to  toy with the device:)

I should have written a disclaimer:

As I demonstrated, ADI-2 does a good job protecting typical headphones from DC.

So we have:
• DC protection.
• There is a warning screen when plugging ‘phones with Hi-Power active.
• The level is slowly ramped up once ‘phones are plugged in.
So the user has an option to unplug or stop playback before the level becomes dangerous - for the headphones or the ears!

A lot of protection measures, and much more than other headphone amps do offer.

But - there is such a great variety of headphones out there that it’s not possible baby all of them.

Not to forget, one can simply break headphones or speakers by playing them too loud.

Finally:
I could imagine, an optional 2 Hz (or tuneable) 1st order low cut would be a nice addition.
As ADI-2 is used for mastering, or directly connected to a speaker amp this could i.e. fix tracks like the one above, or protect speakers where no other option exists.

Dear all,
I have had an ADI-2 DAC FS (ESS chip) for a few weeks now. Last night while listening to music I saw the following message appear: DC detected Phones deactivated Pull out phones plug to reset output state.
Since then the ADI has not worked. Whichever headphone I plug in, I always get the same message, even if I am not listening to music and the volume is at 0.
I tried restarting the computer and the ADI, but nothing worked.
I have 2 headphones, HIFIMAN Arya Stealth and a Sennheiser momentum. The problem occurs with both headphones. In fact I will tell you that the problem does not depend on the headphones. All I must do is insert a 3.5 to 6.5 conversion jack into the phones socket and I get the same error.

What can I do to be able to listen to music again?

Thanks for your help.

Gianluca

Thank you for your quick reply.
it's really sad that such an amplifier breaks down after not even 2 weeks doing what it's supposed to do, amplify.
PS: even line out connection with an external amplifier (ROTEL) is not working.
I'm trying to contact the German seller.