DC offset on Meta42 w/ bass boost?
Nov 15, 2004 at 7:08 AM Thread Starter Post #1 of 4

patricklang

100+ Head-Fier
Joined
May 7, 2003
Posts
228
Likes
0
I set up a simple modification on my Meta42 to test out bass boost. I put a 100k alps pot (from radioshack) and .1uF poly/metal cap in parallel in the global feedback loop (same as PPA circuit). I'm currently measuring 33 mv dc offset with the bass bost level I want (probably around 10k/.1uF), which worries me a bit. As I recall, when I originally built this amp, I was at 10 mv or so without bass boost.

Would this dc offset level be damaging? I'm using k501's (120ohm).

What should I check into for alleviating this offset?

Here's the specs from what I remember:
hi-z crossfeed
overall gain ~9?
(R1-1k, R3-1k, R4 9.09k, R5 1.58k, R6 191k, R7 68.1k, R8/9 short)
no cascode
OPA2778
2x EL2001cn's per channel
 
Nov 15, 2004 at 12:22 PM Post #2 of 4
By adding the bass boost you have increased the gain of the amp at low frequencies, which extends all the way down to 0Hz (DC). It's not surprising that the output DC offset increases as a result.

There are ways to alleviate this but each also has its own downside:
  1. Insert a large value capacitor in series with the resistor that goes from the inverting input of the opamp to ground. This causes the gain to fall back to 1 as the frequency drops near DC, but that capacitor (being large in value would have to be an electrolytic type) would degrade the sound. You can add a high quality film cap in parallel with that electrolytic cap, which helps restore some of the lost quality.
  2. Insert a large value output coupling capacitor to block the DC offset from being "seen" by the headphone. This also has the same problem as the first solution.
  3. Add a DC servo circuit. This adds complexity to the amp.
  4. Reduce the amount of bass boost.

In your case, 33mV is marginally high for a mid-Z headphone like yours. It's probably safe, but I'd sleep better if I do one of the above to reduce the offset.
 
Nov 15, 2004 at 8:13 PM Post #3 of 4
Oops, in my original post I stated OPA2778, its a 2228
wink.gif


I'm going to take the bass boost back out and remeasure the offset with the AD8620 and OPA2228 to see if there's any difference. If I can't get the offset down after checking on other stuff, I'll probably skip the bass boost until my next amp. I don't feel like hacking in more caps, especially electrolytics.
 
Dec 3, 2004 at 2:01 AM Post #4 of 4
I took the bass boost back out, measured the dc offset, and it was still over 30mv. Swapped in my AD8620, and I was down to 12mV. Same with my AD8066. I need a scope, maybe the cursed 2228 struck again? Regardless, I think the AD8066 sounds better.
 

Users who are viewing this thread

Back
Top