Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › Unequal sine wave from CMoy
New Posts  All Forums:Forum Nav:

Unequal sine wave from CMoy

post #1 of 7
11/13/10 at 11:48am
Thread Starter 

I wanted to measure the output power of my CMoy amp. So I connected it to my oscilloscope and when I started to increase volume in search of the point when it would start clipping, I found that the top of the sine wave goes flat before the bottom of the wave.

phpuvM4xFAM.jpg

 

The input signal comes from SigJenny @ 1kHz (tried it from my PC and laptop)

The load is 22.5ohms.

The top of the wave starts clipping at about 260mV and the bottom at about 340mV.

...and even if they both would clip at 340mV, isn't that too low of a voltage at 22.5ohms?

 

The virtual ground is done with a tle2426 like this, but with a 470uF cap.

tle2426clp.png

 

The aplifier section is standard as seen here http://tangentsoft.net/audio/cmoy-tutorial/misc/cmoy-tangent-sch.pdf

There is no R5, R3 is 5.1k (giving a gain of 2,96) and the input caps are 0.3uF.

Op-amp is BB OPA2132.

 

Any ides on the sine wave being unequal? And how much output voltage should I expect with a 22.5ohm load?

If it's any help I can post the layout and some pictures of the inside, though the latter can be seen in my profile's album.

Reply
post #2 of 7
11/13/10 at 12:31pm

Regarding your "lower than expected" voltage output. 

On the datasheet for the opamp, check out the spec labeled "Output (voltage)" on page 3 of the datasheet from Ti. You can see that the output voltage relative to the rail DECREASES as load increases. With a 10Kohm load the chip can go ~0.7V from the rail. with 2Kohm its more like ~1.3V, and with 600ohms ~2.3V. 

 

You can see that as the load impedance decreases how close to the rail the chip can run also decreases. If you want to verify that things are right try the amp on the 600ohm load, which is given. If they dont list it in the datasheet it, its probably not very good - they didnt specify the chip for low impedance loads.

 

This spec also gives a hint as to your asymmetrical clipping. If you look at the actual values, not my lame averages above you can see that how close the chip can drive the load to the rail is NOT actually equal compared to both power supplies. This causes your asymmetrical clipping.

 

To fix this find an opamp that is rated to run closer to the power rails with a low impedance load.

Reply
post #3 of 7
11/13/10 at 1:52pm
Thread Starter 

Thanks for a fast reply. I tried it with a 600ohm load. The sine still turns asymmetrical. Top of the wave goes flat at about 1,32V and bottom at about 1,74V. I also tried a 1Kohm load, but nothing changed - top flattened at about 1,48V and bottom 1,93V.

Reply
post #4 of 7
11/13/10 at 3:32pm

Asymmetrical clipping is normal with any load and the OPA2132. How close to the rail it can swing is different for the + and - supplies. 

 

What is your battery voltage under load? Turn the amp on and measure the voltage. If its any less than 8V your measurements are not far from the values in the datasheet. 

Reply
post #5 of 7
11/13/10 at 3:32pm
Thread Starter 

The battery voltage under load is 8.45V.

 

I tried the two other dual op-amps that I had at hand - NE5532AN and RG4560IP. Both gave a bit higher output voltage before clipping. The first one had a symmetrical sine wave even with a 22.5ohm load, but it still sounded distorted to my ear, when I listened music through it. Even at low volume. RG4560 on the other hand was as asymmetrical as OPA2132 @ 22.5ohm & 1Kohm but gave a bit higher output voltage than 2132 before the top of the wave flattened. But when I listened music through it, it seemed to start distorting at the same volume as OPA2132. And as the 2132 sounded better to my ear, I left it in.

So is it possible, that the results differ when measured by measuring equipment and by ear? blink.gif


Edited by Reintz - 11/13/10 at 3:38pm
Reply
post #6 of 7
11/13/10 at 10:22pm

Check the power supply voltages V+ to ground and V- to ground.  When virtual ground circuits go unbalanced it is normally the V+ voltage that drops (while the V- supply raises), which would explain why it is clipping high.

Reply
post #7 of 7
11/20/10 at 12:54pm
Thread Starter 

Thanks guys. I measured the TLE2426 and the virtual ground was unbalanced like SteveZ said. So I thought to myself if there was something I could do to get some more power out of it. I didn't want to make a new PCB so I decided to make a little adapter that would let me use the TLE2426 DIP8 package insted of the TO-92. It would have been cool to make it with SMD parts, but I have never tried SMD and the TLE wasn't available as a SMD at the moment so I made it like this...

php6InhuNAM.jpg

 

It just fits and man did this make a difference. With a 22.5ohm load it goes all the way to 440mV before the top of the wave starts flattening (with the TO-92 it was 260mV).

So the power has gone up from 30mW to 86mW. ksc75smile.gif

Reply
New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › Unequal sine wave from CMoy