Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › How much DC offset would you consider harmful to HP and after which point would you throw the commercial amp out?
New Posts  All Forums:Forum Nav:

# How much DC offset would you consider harmful to HP and after which point would you throw the... - Page 5

You don't need latex (or wikipedia) to tell someone how to measure an amplifier's dc offset with a dmm...

you seem to have 2 rather vital types of voltage switched around. AC measures rms heating value, DC does not, not even a little bit. otherwise for instance, measuring a fullscale AC signal at the output of an amp, or a dac, or a signal generator with a low leakage cap at the output would still read DC, guess what? it doesnt, it measures 0vdc

What I'm saying is audio signal is fed in to the dmm.  you can set the dmm to DC, but it's pointless as the signal is ac, its an audio signal.  it fluctuates.  Put it in AC, you get the RMS of AC, as the dmm gives you a number, not respresentations of the audio signal fluctuations.  My question is, if you set it to DC, how do you measure the offset?  One idea would be to send in a tone signal of known value. measure rms, do some math and you will get the offset.

many analogue Rms responding voltmeters are AC coupled even. my hp3400A is, so how it could possibly read a few mv of dc i dont know. a large amount would screw with it due to leakage, so you cant have it set to amplify the input too much or you'll kill the input stage, but a small amount wont even make it through

DC causes AC heating reading errors sometimes, but i have never encountered AC causing DC errors

Edited by qusp - 6/4/11 at 7:48am

Yet for most humans who do not dwell in the vaunted halls of Wikipedia copypasta, measuring DC from signal to ground with a DMM with no music playing is adequate, as once you have <10mV any DC contained in the actual signal is negligible and would probably be hard to get rid of anyway. In fact, measuring DC on an amp with a source attached with music playing is a really stupid idea to beging with, as any DC from the source will be passed on through the amp and give you innacurate measurements.

I understand what you're trying to say with your talk of scopes, but for audio it really isn't necessary to get that accurate. Our ears and transducers are only so good, after all.

So when the music is not playing, offset is measureable?, and it is infact the same offset that the voltage swing will be centered around when playing?  I'm doing this not because I'm anal about offsets, but to get a better understanding of measuring.  Every little bit helps.

Quote:
Originally Posted by JoetheArachnid

Yet for most humans who do not dwell in the vaunted halls of Wikipedia copypasta, measuring DC from signal to ground with a DMM with no music playing is adequate, as once you have <10mV any DC contained in the actual signal is negligible and would probably be hard to get rid of anyway. In fact, measuring DC on an amp with a source attached with music playing is a really stupid idea to beging with, as any DC from the source will be passed on through the amp and give you innacurate measurements.

I understand what you're trying to say with your talk of scopes, but for audio it really isn't necessary to get that accurate. Our ears and transducers are only so good, after all.

Edited by High_Q - 6/4/11 at 7:58am
Quote:
Originally Posted by High_Q

So when the music is not playing, offset is measureable?, and it is infact the same offset that the voltage swing will be centered around when playing?  I'm doing this not because I'm anal about offsets, but to get a better understanding of measuring.  Every little bit helps.

Yes. Measure the DC with the amp not playing anything at all. Make sure the amp is turned on and warmed up.

You should either have your source hooked up or ground the input signal line to ground.

If you show offset with your source connected then you should first check the source to see if it has offset. Again without music or test tone playing, just idle and turned on.

Edited by digger945 - 6/4/11 at 10:07am

Hello,

Can Planar Magnetic headphones like the LCD's or Hifiman HE 500 also get damaged by the DC current since they dont have a voice coil? All answers are highly appreciated. Thanks in advance

Quote:
Originally Posted by kasimoa

Hello,

Can Planar Magnetic headphones like the LCD's or Hifiman HE 500 also get damaged by the DC current since they dont have a voice coil? All answers are highly appreciated. Thanks in advance

Yes, the exact same principal applies, they still have a conductor, equivalent to a voice coil, but in a single plane. It will heat up and burn out if there's enough dc across it.

I would expect them to tolerate dc a bit better than the average headphone though (higher power handling, and large surface area for cooling)

Edited by DingoSmuggler - 2/7/13 at 11:16pm

That makes sense thanks for making that clear to me.

There have been a lot of formulas bandied around here to no great effect.

What can we actually do with formulas?

Here are a couple of very basic ones...

1) V = I x R

2) P = I x V

...rearranging (1)...

I = V / R

...and substituting into (2)...

3) P = (V x V) / R

This tells us the voltage required to produce a given power in a given resistance.

Harmful is one thing, desirable is another.

Modern music systems are, for the large part, 16-bit. In the mastering process dither is applied to the least significant bit. Presumably this means that you can hear the effect, otherwise why do it?

I have a pair of Shure e2c IEMs here. They're considered reference-quality by people like Seigfried Linkwitz, and, more importantly, they're representative of the typical impedance and sensitivity readily obtainable at a reasonable price. They have an impedance of 16 ohms and a sensitivity of 100dB/mW.

Taking 100dB as a reasonable listening level, from (3), the voltage to give 1mW in 16 ohms is 0.126mV.

16 bits means 65536 levels.

So the lowest perceptible level in these phones is 0.126 / 65536 = 1.9uV. Yes, that's right, ~2 microvolts.

So, IMO, we would ideally like to see a DC offset <2uV. Then we could reasonably claim it was insignificant.

This is difficult to achieve in a mass produced item. It really requires a DC servo and individual adjustment of each unit. It could be automated using an OTP preset digipot, although measurement at such levels would require reliable contact in the test jig. A low-leakage cap is not really a viable solution. The typical leakage current in a 220uF cap with 1V across it will produce ~35uV in 16 ohms, you need a big cap to maintain the FR, and anyway, nobody wants a cap like that in their system. You probably can't hear it, but if it's not there, then you definitely can't hear it.

Of course this is all nit-picking to the n-th degree, but I don't really consider an offset >1mV acceptable from an aesthetic POV, considering what is achievable, although a few mV of DC is probably of no consequence in these Shure IEMs. I mean, they should tolerate an output of 120dB, which is 1.26 (rms) volts in.

Obviously less sensitive, high-impedance phones can tolerate bigger offsets and systems intended to drive speakers have historically been built to higher tolerances.

But 300mV? For a headphone amp, knowing what is likely to be plugged into it? No way.

w

After all the maths I was waiting to see if somebody would say apart from burning out the voice coil if very high another problem that is basic is that  in Audio Power amps when connected to moving coil loudspeakers --The voice coil is -as long as the amp is switched ON-OFFSET from its NORMAL NEUTRAL position.

There bye causing a DISTORTED reproduction as the loudspeaker designer made the high spec./low distortion taken into   consideration the NEUTRAL position of the voice coil. And yes input which has a DC content when applied to an input that is DC connected  ALL DC connected signal circuit[ NO output capacitor]. will  be amplified in proportion to the gain of the amp.

So the same would apply to MC head-phones.

I too have a HP bench voltmeter -HP 3455A with a no leakage isolating cap at the input.

I thought that I should measure the DC offset of my phone before plugging any amp to it. Here are the results:

Everything was measured between the R/L and ground, all values were absolutely constant, even the ones with music playing.

I think the left channel looks a bit weird, right? What could cause the DC offset when music isn't playing? Is it safe to hook up an amp to it? I'd really appreciate some help

The figures you quote aren't bad and in hi-fi company engineering terms --"are within the specified range".

But that depends on what your amp is -If a twin amp audio chip then those figures could  apply.

A single channel audio chip can have its off-set reduced to zero by adjustments to the off-set tags on the chip.

Some don't. If it is a power amp that is discrete [no chips] then that can be adjusted from the input balance.

If a power audio chip then it depends on the chip. It is also possible  to create a circuit at the output to balance it.

Specialist companies usually try to get it to zero to enhance the engineering qualities of their product.

So yes it probably will be able to achieve -zero -It depends if you think that small amount is bothering you?

Just noticed you are talking about a pre-amp-so it stands a good chance it is a twin audio chip which doesnt have off-set .

But if it is discrete then again can be zeroed by adjusting the input balance. Can you post a link to a circuit diagram?

New Posts  All Forums:Forum Nav:
Return Home
• How much DC offset would you consider harmful to HP and after which point would you throw the commercial amp out?