Amplifier DC Offset question

[Bloos]

I have an old Channel Islands VHP-1 amp and when I measure its DC Offset it's very close to 0v, showing 0.0 mV on my multimeter. But when I change the volume, gain setting, or turn off the device the DC Offset can spike up to 30, 80, 175 mV respectively, then slowly make its way back down to 0mV (in 5-10 sec). On the other hand, when I measure my O2 amp it's always 3.2 mV on both channels no matter what I do.
Sound-wise, it sounds great and there doesn't seem to be an issue.

My question is, is this normal behavior for an amp and can this in any way harm a pair of headphones?
and if so, what might be the issue?

[Side-question] Also, is high DC Offset less of a concern with planar magnetic headphones because they don't have a voice coil/inductors?

 

[Rroff]

The O2 has the DC blocking caps after the gain stage which is relatively unusual and why it tends to have more constant levels - 30, even 80mv under normal use isn't really a problem and neither is 175mV if that is just a momentary power on/off transient.

 

[rellik]

Replacement of the capacitors near the dc offset potentiometer should help dc offset settle faster.

Zero dc offset is not always the best setting sonically, use your ear and make the setting appropriate to what your favorite music is. This will effect the settling speed of a certain frequency range. It also may dull fast transients at the position it is set.

 

[Nisbeth]

Zero dc offset is not always the best setting sonically, use your ear and make the setting appropriate to what your favorite music is. This will effect the settling speed of a certain frequency range. It also may dull fast transients at the position it is set.

Ahem, you do know what DC offset is and why it is standard practice to keep it as close to zero as possible - right?

 

[rellik]

Yeah, it is a saturation of the driver which then reduces resolution at, around, or below that voltage level.

With a continuous DC offset, the driver can continue to build charge to a point higher than the initial dc offset voltage potential. Sorry, but slowly dropping in charge continuously into a media will prove to be very difficult to get rid of.

Ever notice why the DC offset can be trended to near zero at idle yet starts to change significantly when playing music? The DC offset circuit is one of the negative feedback loops in the amplifier defined by a simple paralell RC tank circuit.

The capacitor tends to slow the negative feedback so that the amplifier is able to have the previous,current,andfuture sound propogate through the circuit. The capacitor draining and filling above and below the threshold switching state charge(voltage) of the NPN and PNP transistors define the goal steady state of the RC feedback circuit.

The resistor defines how fast the capacitor is able to discharge and thus the speed of the feedback loop.

The factor of time involved in this charge/discharge process inherently defines that the negative feedback subloop will be tuned to a certain frequency. Thus the initial statement of zero offset is not always the best setting. Sometimes even saturation is more in tune with the music.

Dont worry, the "headphones" can handle whatever small amount of DC offset range of whatever circuit you are referring to.

Tell me the next time you drive in you car if you can notice that it is running a pretty hefty DC offset. And hey, we all kno that is the best place to listen.

 

[DutchGFX]

Yeah, it is a saturation of the driver which then reduces resolution at, around, or below that voltage level.

OK, this is true, I suppose.

The factor of time involved in this charge/discharge process inherently defines that the negative feedback subloop will be tuned to a certain frequency. Thus the initial statement of zero offset is not always the best setting. Sometimes even saturation is more in tune with the music.

I'm not an expert by any stretch, but you've missed the point. Your entire post is about frequency response. DC Offset has absolutely nothing to do with frequency response. Your argument is based entirely off of the frequency response of a DC blocking capacitor. DC Offset can be, and often is, eliminated without the use of a DC blocking cap. If there is a coupling cap, you cannot mathematically get ANY DC (unless your capacitor has a very finite resistance, like, less than 100k (it's just a potential divider). I don't understand how the coupling cap becomes part of the feedback loop either.

Tl;Dr

DC Offset is bad. Not trying to get into an argument; maybe there is something I'm missing, but there is a distinct reason why people spend so much time developing advanced DC Blocking circuits, such as the e12 or whatever it is from AMB. The reason is that DC Offset is bad. I'm not sure how much is going to be bad for the headphones (ive never had any issues on any amp I've built), but I can pretty much guarantee that you don't want DC Offset.

 

[rellik]

Yeah Due(d), you should do some reading into what I wrote, never did I refer to the DC coupling cap (DC blocking cap)

The RC servo for the DC offset is different from the DC coupling capacitors.

If you look at a Dynahi, it does not use a capacitor. Just a resistor and an op amp based servo, thus allowing the power supply and the incredible speed of the transistors to color the sound.

DC blocking capacitors serve to allow the DC offset circuit to be faster and more gradual and maybe even include hysteresis for sonic coloring and frequency range eq. This is at the expense of low end attenuation and sonic coloration.

The speed and amount of trending the transistor switch charge level is defined by the current mirrors and or r,l,c network/overall network.

The imparted increased resolution(micro detail) is due to the capacitor's ability to desaturate the transistor network and allow for a more unidirectional propogation(can be boring when using tubes aka high vs low freq collision) without the need for highly matched components, impedance matching pc boards, a common layout, etc.

Guessing that the amp in reference has this, the capacitor often sounds better than the transistors regardless due to the transistor having an inherant self driven oscillation (back clamping emitter to source and emitter to gate using diodes) to avoid propogation of the gate material out of the junction. The cap also helps remove this oscillation and allow for one to saturate the transistors input not their ultimate output, the output cable and headphone.

Saturating the input of the transistor will in effect blur micro level collisions at the media, dependant on the media...

Capacitors also serve the purpose of reapplying (a) specific mix of necessary gate material needed for a transistor to function. The contents of a capacitor varies from the material typically included in wires and the power grid. This material migration is commonly refered to Blacks effect and is what makes old equipment so good due to a more unidirectional crysteline structure.

 

[DutchGFX]

Right. I'm not denying that there are advantages/disadvantages to various methods of DC Offset reduction, but it sounds like you are saying that sometimes adding DC Offset is good, which I am curious why. You are clearly very good at device physics, but I'm not understanding why you would ever want DC on your headphones lol

 

[rellik]

Meh, dunno. It takes someone with little experience and good amount of knowhow (yours truly) to you know break something.

Fixing it is another matter, lets stick to no DC offset.

Edit: If the heatsink is hot, the transistors are hot, and then the wires and headphones are hot.
This is especially true in Class A and especially in OTL, and double especially in non coupled OTL.
Class D is just too fast for your device to measure DC offset (simplification)

 

[DutchGFX]

Glad we are settled

 

[rellik]

No worries, same stuff different day. A good test of DC offset is the band "Third Eye Blind". Increased DC offset can blur crossover errors(?)/confusion. Its kinda kroovy way to open up an audience.

 

[DutchGFX]

Thanks man, I appreciate it

 

[Bloos]

*Now Kiss!*

 

[DutchGFX]

lol no hard feelings at all. Very polite and respectful debate