I testing using battery. I predict to be using unregulated power supply in final product.

Could be the DMM being inaccurate. It's running in low power warning after all.

No rail to ground caps after TLE2426. Tangentsoft tells not to do so.

I'm sorry but what 3 channel are you referring to?

Assuming circuit simplicity is still in priority, would rail to ground cap be considered as a good alternative to provide more current?

Is it sonically possible to build a prototype amplifier on breadboard? I just had a quick try with simple OPA2132. Ordinary jumper wires were used. Audio jacks would not fit on breadboard hence I used crocodile clips for input and putput. I know it sounds very wrong but the performance was way below expectation.

Constant hiss. Sounds like white noise. Terrible low frequency roll off; bass filtered even without any input capacitors. When the song goes through bassy passages, the entire volume dips down. Vocals are excessively wide soundstage. Singer sounds like singing in a hall. At complex passages, the roll off would affet the vocal as well i.e. Voice gets filtered like minus one track.

DC offset is normal (3mV) on both channel. Rail splitter appears balanced and stable during operation. No heat on battery, op-amp and TLE2426 rail splitter.

Could the breadboard and interconnects caused so much degradation or the problem is within the circuit itself?

This post has been edited by mengsuan: Dec 28 2010, 03:21 AM

Voltage swing is increased. Useful for high impedance headphones, which you think is too soft with low gain. Higher gain would also amplify the noise, so keep it low if you can.

Culprit found for the hiss. My metal film resistors are new and they come with tapes. The adhesive material remain stuck on the leads. Used my finger to scrape of the adhesive and it's fine. I also added rail to rail capacitors (cheap 100uF electrolytics). DC offset is very healthy@0.1mV. No hiss at all.

Now the hard part comes, how to improve the amp to ensure that the output of the amp is definitely better than unamped.

The objective of the amp is to improvise CMOY-based amp into something bigger, bolder, better.

Fucuk, the buzz suddenly came back by itself. Arrgh

No pots implemented yet. I plan to control digitally after I am done with the microcontroller.

It is indeed poor connection. Took out the resistors and cleaned the leads, then back to normal. Sigh

Perhaps I should really solder them onto PCB.

---

Added on December 28, 2010, 11:19 pmOoh interesting! Tell me more about it

This post has been edited by mengsuan: Dec 28 2010, 11:19 PM

Alright I will check it out first.

Goodness, the breadboard is already so messy. http://twitpic.com/3kk7k8

Haha, the components are expected to be on PCB later on so I didn't cut them off yet.

All tests was done on a lousy Edifier clip on headphone. After confirming a DC offset of only 0.6mV, I'm hooking up to my Senn HD555. Honestly, I think it is better than my X-Fi Titanium unamped for the vocal clarity.

---

Added on December 29, 2010, 2:18 amNow the amp hums when the input are left hanging. Humming stops when plugged into a source or when the input are grounded. I haven't exactly noticed that the amp would hum when not connected to source earlier. Not normal I guess?

This post has been edited by mengsuan: Dec 29 2010, 02:19 AM

The hum is quite loud. As loud as music

3.5mm jack have contact switches, I could ground the input whenever interconnects are not inserted. But problem comes when the interconnect is inserted but not into the source. I cannot be having such an embarassing amp.

I feel the only way to tell, is to solder it at least to a perfboard or stripboard. The connection is too wanky on breadboard

There's a lot which lectures don't teach. I'm just learning as well.

Is there any 3.5mm or 1/4" jack that mounts on the casing itself?

LittleGhost, when can I come to use your drill and get those jack/pots?

For a CMOY amp, is there any worthiness to Class-A bias using just resistor?

Thanks to littleghost's guidance, I have my first complete amplifier with a case. Many thanks for his afternoon on Saturday.



Quite a looker isn't it. Thanks to littleghost's Jujubes box.



Work log, schematic and other details: http://mengsuan.blogspot.com/2011/02/project-ubercmoy.html

Thanks for your compliments. Truly encouraging.

Hopefully one day soon I could progress as far as you and littleghost are.

p.s. littleghost even helped me punch holes through the case.

This post has been edited by mengsuan: Feb 1 2011, 09:50 PM

One of the chip is OPA2132. Another chip is TLE2426, a virtual rail splitter.

You just reminded me that, I should also try dual op-amp like A47 design. Well, maybe I should progress further than that

I put it in tombstone orientation because I want to make the board as compact as possible. The laid-flat resistor, is however, out of the view in the photo.

Hard to reply. None of the components costs a lost. Some components were from my toolbox, some were from littleghost. Nevertheless, it's the effort that costs. I took several visits to Jalan Pasar to get the materials for breadboard prototype testing.
I did not spend the money very efficiently. Moderate amount of wastage.