Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › cMoy signal VERY dirty, got me puzzled
New Posts  All Forums:Forum Nav:

cMoy signal VERY dirty, got me puzzled

post #1 of 14
2/21/10 at 6:58am
Thread Starter 
Im trying to make my first cMoy altoids amp. Ive got everything wired up in the correct positions according to the tutorial here:

How to Build the CMoy Pocket Amplifier

Without a switch and potentiometer (which im assuming arent necessities) the amp produces a very dirty signal in the both channels. If i place a resistor on the RIGHT out, the LEFT out cleans up, but the RIGHT stays crappy. There are no solder bridges, nothing shorted, continuity is there allt he way. What could this be? is my IC fried? I cant figure out why there is so much signal degradation.

note: I am using 1 9v batt., when a load is present (input and output plugged in), the right channel receives about 7.~ volts, while the left gets ~+1. When no load is present (input but no output plugged in), both sides receive even amounts. (~4.2) This must be the cause of my noise, but i still dont know what could be causing the imbalance in draw.

Any Ideas?
Reply
post #2 of 14
2/21/10 at 11:15am
Virtual Ground Circuits
Reply
post #3 of 14
2/21/10 at 1:58pm
Quote:
Originally Posted by ekliptiko
note: I am using 1 9v batt., when a load is present (input and output plugged in), the right channel receives about 7.~ volts, while the left gets ~+1. When no load is present (input but no output plugged in), both sides receive even amounts. (~4.2) This must be the cause of my noise, but i still dont know what could be causing the imbalance in draw.
Additional to what Tangent posted, the V+ and V- power pins on the opamp are for both channels and are not associated with a single channel. You are seeing the positive rail vs. the negative rail WRT the virtual ground.
Reply
post #4 of 14
2/21/10 at 8:00pm
Thread Starter 
@Tangent: im assuming you wrote this tutorial, however: with 1x9v battery, shouldnt the 4.7k resistors be reduced to 2.4k? that was my understanding, and thats what i did. i still have the same issue. correct if im wrong please, i want this to work.

Quote:
The input and output are both referenced to the same ground point, so nobody cares that the virtual ground isn't precisely centered between the rails. So, why worry about it?
Does this tell me that the imbalance that i am seeing is not the source of my filthy signal? If so, what is?

I dont think i understand this rail splitting buisness correctly...I still dont see where this distortion is coming from.
Reply
post #5 of 14
2/21/10 at 8:05pm
Thread Starter 
Just to make sure, i want both rails to draw evenly, correct?

Is the next step to make this work to add a buffer? thats making things a little too complicated for my understanding atm.
Reply
post #6 of 14
2/21/10 at 8:34pm
Adding a buffer won't fix it, you've made a mistake and you need to find it, post some pics and I'm sure somebody will spot it for you
Reply
post #7 of 14
2/21/10 at 9:34pm
Thread Starter 




there is one more with a flash Here
Reply
post #8 of 14
2/21/10 at 9:40pm
Thread Starter 
Bottom:




thought someone might spot a problem i didnt...
Reply
post #9 of 14
2/21/10 at 9:42pm
Thread Starter 
i think i might have just found my problem: the jumper in the lower left on pic #1 above is supposed to be moved up one hole. where it sits, it doesnt jump the two resistors together. could this be the source of my problems?

ill try making a solder bridge to test this theory.
Reply
post #10 of 14
2/21/10 at 9:58pm
Thread Starter 
FIXEDFIXEDFIXED. that was the problem. i guess only the left chan. was getting resistance to clean up the sound. once there was a solder bridge, it cleaned right up. vast improvement over the ipod hp out.

I still have the 2.4k resistors on the power input, with my limited electronics knowledge, i need those as opposed to 4.7k if im using 1 battery?
Reply
post #11 of 14
2/21/10 at 10:14pm
Quote:
Originally Posted by ekliptiko View Post
@Tangent: im assuming you wrote this tutorial
Yes.

Quote:
with 1x9v battery, shouldnt the 4.7k resistors be reduced to 2.4k?
That will help with the imbalance problem, but it's more common to recommend going to 2x9V if you have a suitably bad virtual ground imbalance that you get bad sound. You get better battery life that way.

If you absolutely must use 1x9V and stick with the stock design, you could lower the rail splitting resistors still further. I wouldn't go lower than 1K. Again, though, this hurts battery life.

By the way, those carbon resistors you're using aren't doing any good for your amp's noise floor. The worst are the carbon comps you're using for the vground divider. This isn't likely the main problem source, they're just not helping. Metal film resistors are quieter.

Quote:
Does this tell me that the imbalance that i am seeing is not the source of my filthy signal? If so, what is?
The imbalance may be causing the amp to clip. The easiest way to check is to try 2x9V. If that fixes it, the amp was almost certainly clipping due to vground imbalance. The beauty of the 2x9V test is that it can both answer the question of what the problem is and provide a fix for it at the same time.

Quote:
Originally Posted by ekliptiko View Post
i want both rails to draw evenly, correct?
With the stock CMoy design, there is no general way to do that. You can hack it to make it even with a particular set of headphones, but it'll shift again on you with different headphones.

Quote:
Is the next step to make this work to add a buffer?
That's one way to fix the imbalance problem, yes. It's not absolutely required. People build CMoys all the time, and stick with essentially the stock design.

The vground article is really the story of the progression from CMoy -> MINT -> META42 -> PIMETA/PPA. Every step along that path fixes some deficiency in the stock CMoy design.

You should clean your board. Truly! Flux can trap conductive junk on the board and hide conductive paths. Burned rosin flux contains carbon, which is conductive.
Reply
post #12 of 14
2/21/10 at 10:16pm
I think you should clean up the flux with some solvent, like ethanol or acetone. That should help a little too.

If you use smaller value resistors your virtual ground divider draws more current, but it should be more stable under heavy load.

That is a good start.
Reply
post #13 of 14
2/22/10 at 1:52am
Thread Starter 
ill be sure to clean it up before it goes in the altoids tin. what does flux look like? is it a yellow/brown?

Thanks for the tip on the carbon vs. metal resistors. the output is vastly better than an ipod, but thats not saying much. ill be sure to use metal film from now on.

next step: enclosure, pot. and switch. thanks for all the help so far.
Reply
post #14 of 14
2/22/10 at 1:18pm
Yes, the flux is the sticky yellow/brown stuff around and on each joint. The color varies depending on how long the joint was heated, and so how much the flux burned.
Reply
New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › cMoy signal VERY dirty, got me puzzled