or Connect
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › Intermittent noise in LM6172 Cmoy
New Posts  All Forums:Forum Nav:

Intermittent noise in LM6172 Cmoy

post #1 of 7
Thread Starter 

Hi all,

 

First time post here. I'm having some interesting issues with my first DIY project, an LM6172-based CMoy design. (I thought, perhaps too audaciously, that the challenge of getting that chip to work would be more rewarding than reproducing the original circuit.) The amp is based on the schematic provided on the Tangent "Working with Cranky Op-Amps" page, with an LM317/TLE2426 power supply regulated down to 14V from a DC wall adapter (currently 24V). I am using a pre-patterned perfboard with pads arranged in the breadboard style, and my layout is as shown on the cartoon image I've linked to below.

 

Schematic: https://www.dropbox.com/s/lv0xyrc8mpb0hrv/hs-opamp-schem.png

My layout: https://www.dropbox.com/s/rqu1oedi2xsda4m/cmoy_layout.png (black lines are insulated wires, blue lines are jumpers/leads, purple lines are out to the headphone jack. The power section is not complete as shown, but the IC's are in those places in the power supply on the actual PCB.)

 

My problem is an intermittent noise in only the left channel (this channel is actually on the left in the layout pictured). It is rather "static-y," sounding like a more mellow version of a piece of paper being ripped in half. It's quiet, unnoticeable except during quiet passages of music (and of course when no input signal is provided). It comes and goes without any apparent pattern, typically in small pulses that range from a fraction of a second to a few seconds in length. The circuit may go minutes without making any noise at all. I don't believe the amplifier is oscillating badly, as music still sounds good, the chip is not getting hot, and current through the circuit is well below its maximum.

 

I have tried a couple things to get rid of this, including increasing the gain (R4 to 620K, rather than 470K) and grounding the unused PCB pads near the offending channel. Neither of these helped. Since the amp sounds fine with an OPA2134 chip, I'm hesitant to try soldering this chip to the board until I understand better where the noise signal is coming from. Any ideas as to what is causing the noise and/or why the circuit is picking it up would be very welcome. Perhaps some type of interference? I am kind of a big noob as far as circuits go, so many thanks in advance in case anyone has an idea.

 

(I'd also be interested in any feedback on the layout of the circuit as well---though I do understand that a patterned board like this might be problematic with a high-speed op-amp!)

post #2 of 7
Have you tried another of the same kind of op amp?

Can you get (adjust to...) a 18V supply?
post #3 of 7

I had not previously looked at your layout. 

Not a fan. 

 

There are too many parts going in too many different directions. 

post #4 of 7
Thread Starter 
Quote:
Originally Posted by nikongod View Post

Have you tried another of the same kind of op amp?

Can you get (adjust to...) a 18V supply?

 

I haven't tried another LM6172, so perhaps I'll pick one up next time I order some parts. I will try increasing the supply voltage tonight.

 

As a side note, this noise is almost identical to what I'm hearing in a DOA Schiit Modi that I'm sending back to the factory for repair (or hopefully replacement...). Among other problems (audible distortion, channel imbalance), the DAC makes the same type of intermittent noise in one channel in all testing conditions (different computers, amps, etc). I know the Modi uses an op-amp for the output stage, so I'm wondering if this is a typical faulty chip problem.

 

Thanks also for the feedback re: the layout. It's a little sloppier than I intended, since I foolishly didn't work directly from my designed layout while I was soldering parts down. In general, is it preferable to keep parts aligned on the same axis? 

post #5 of 7

Too many parts going in different directions was not exactly correct. 

I'm not so concerned with alignment as I am with spacing. 

Such little stuff is butted up against the socket. There is tons of empty space. That empty space lets problems in (poetic licence). You are already pushing your luck with an op amp like that in a socket... Your best bet is to keep a circuit like this physically TIGHT. 

 

Although it does need to be tweaked SLIGHTLY to accommodate the extra caps, the classic Tangent Cmoy layout is pretty good start. 

post #6 of 7
Thread Starter 
Quote:
Originally Posted by nikongod View Post
 

Too many parts going in different directions was not exactly correct. 

I'm not so concerned with alignment as I am with spacing. 

Such little stuff is butted up against the socket. There is tons of empty space. That empty space lets problems in (poetic licence). You are already pushing your luck with an op amp like that in a socket... Your best bet is to keep a circuit like this physically TIGHT. 

 

Thanks for elaborating. I was concerned that the all of the resistor leads sticking up out of the board in the Tangent layout might make the amp more likely to oscillate, but in retrospect that idea was totally unfounded. My goal in the design was actually to keep it as tight as possible without orienting the resistors vertically...

 

I tried increasing the supply voltage (now 17.1V at the op-amp) as well as grounding the traces that run underneath the socket, unfortunately to no avail. I suppose my main question now is, what am I actually hearing here? I don't have a scope, but this doesn't seem like the oscillating behavior I've read about on the forums. And the similarity to the problem with the Modi makes me all the more curious!

post #7 of 7
Thread Starter 

So, I eventually ordered more parts, including another LM6172. Sticking the new chip in the socket got rid of the aberrant noise in the right channel---but instead produced a similar effect in the opposite channel! Switching the chips back and forth demonstrated that they were definitely not the same. For me, the LM6172 was definitely more trouble than it was worth.

 

Fortunately I also ordered an OPA2228, which appears to be well-regarded and could take advantages of the above modifications to the standard Cmoy circuit. This chip sounded all right at first pass, and so I boxed up the amp---only to find, after a few minutes, that the sound had a nasty, compressed feel (both dynamics- and imaging-wise) and gave me a headache. It was also rather noisy, which I did not mind too much, and which I attributed to the high absolute values of the feedback resistors used (120k and 620k).

 

I decided to try reducing the values of the feedback resistors in hopes of improving the sound. The datasheet and Tangent's DC offset calculator indicated that the 120R and 620R I had on hand would work, so I switched them in. And voila! The sound is now more rich, spacious, and dynamic (and most importantly, doesn't make my head hurt), on par with what I'd expect from an entry-level headphone amp. Background noise is also gone.

 

As an introduction to circuits, this has been a most gratifying experience. The layout planning, troubleshooting---I blew a 1W power supply resistor and TLE2426 in a glorious plume of magenta smoke trying to find a short circuit, which ended up being a busted opamp bypass cap---and adaptation of the circuit were all great learning experiences, and this for sure won't be my last DIY audio project. I'm thinking a Nelson Pass JFET or MOSFET preamp next, as an introduction to finding the "sweet spot" in discrete circuits...

New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › Intermittent noise in LM6172 Cmoy