The Objective2 (O2) Headphone Amp DIY Project

[Nevod]

Here
It's a compilation of a series of small articles from 2001, that. A guy was designing an discrete headphone amplifier with goal of putting all key elements under constant dissipated power conditions so there would be no drift of DC conditions and associated shift of amp's parameters.


Here is a method which he used to measure 'memory' - integration of DC offset at amp's output after signal has been cut off.

He noticed improvements from that, though also he has found an improvement using high(~100dB) feedback and in general admitted that high feedback is an easiest method to combat it, and O2 as a chip amp, already has that, I suspect.

[lorriman]

I have quiet rustling on turning the pot in one channel when using my etymotics (low gain).

Any clues as to what might cause that? Everything seems normal apart from that. Measurements are fine.

Peculiarly it doesn't happen when playing a generated silent mp3.

[Rullknufs]

Originally Posted by lorriman

I have quiet rustling on turning the pot in one channel when using my etymotics (low gain).

Any clues as to what might cause that? Everything seems normal apart from that. Measurements are fine.

Peculiarly it doesn't happen when playing a generated silent mp3.

I have the same issue with my DIY one from headnhifi.

[agdr]

lorriman and Rullknufs: just curious - which channel has the rustling and does your O2 build use the capacitors in the BOM for C13 and C14 (the square 2.2uF caps)?

According to RocketScientist/NwAvGuy's writings rustling is DC through the wiper in the pot. There should be essentially 0Vdc across the pot since the output of U1 should stay more-or-less centered about zero unless clipping occurs. But it will drift a little with op amp input offset bias and offset caused by input bias current differentials through the input resistors.

If the rustle is there with large signal audio but not with small signal or zero signal, seems like the DC bias out of U1 is shifting slightly, which in turn charges/discharges C13 and C14 until it equalizes again. Or something like that. Just pondering out loud.

[Rullknufs]

Originally Posted by agdr

lorriman and Rullknufs: just curious - which channel has the rustling and does your O2 build use the capacitors in the BOM for C13 and C14 (the square 2.2uF caps)?

According to RocketScientist/NwAvGuy's writings rustling is DC through the wiper in the pot. There should be essentially 0Vdc across the pot since the output of U1 should stay more-or-less centered about zero unless clipping occurs. But it will drift a little with op amp input offset bias and offset caused by input bias current differentials through the input resistors.

If the rustle is there with large signal audio but not with small signal or zero signal, seems like the DC bias out of U1 is shifting slightly, which in turn charges/discharges C13 and C14 until it equalizes again. Or something like that. Just pondering out loud.

Yes, I'm using the square caps.
Here's a picture of my board
The gain resistors are removed now though, but otherwise it's the same.
Tried it quickly and with no signal there's no rustling when turning the volume knob.

Is there any quick and easy fix for this?

[agdr]

Originally Posted by Rullknufs

Is there any quick and easy fix for this?

If you happen to have a LME49720 = LM4562 on hand, a really interesting test would be swapping the LME49720 for the NJM2068 and see if the problem gets better. The 49720 has about 1/3 the input offset voltage and about 1/10 the input bias current of the 2068. The AC parameters (distortion, etc) are the same or better. So if there is an improvement it would be largely due to the DC performance (DC output drift across the pot).

[Rullknufs]

Originally Posted by agdr

If you happen to have a LME49720 = LM4562 on hand, a really interesting test would be swapping the LME49720 for the NJM2068 and see if the problem gets better. The 49720 has about 1/3 the input offset voltage and about 1/10 the input bias current of the 2068. The AC parameters (distortion, etc) are the same or better. So if there is an improvement it would be largely due to the DC performance (DC output drift across the pot).

Unfortunately I have no components at hand But maybe somebody else here does?

[agdr]

Other than trying the LME49720 I'm kind of stumped on this one. Maybe one of the forum circuit wizards here will have some ideas.

[Speedskater]

Originally Posted by Zoide

Anyone? Is NwAvGuy OK?

It seems strange that no-one knows where NwAvGuy has gone for the last 4 months (maybe it's 7 months).

[agdr]

Yeah, he seems to have just suddenly disappeared. Last comments on his blog were back in August. Hope he is OK! Must have either been medical or a sudden loss of interest.

Reminds me of the "Dr Xin" situation over on the H-F forum. He apparently just disappeared one day in the middle of taking orders for a headphone amp.