[proid]

The O2 amp using either AC regulated or battery power supply. As my ear and I have read quite a few post claim that there is no different in sound quality between using adapter or battery or maybe more bass punch from adapter because of higher current capability. But as my knowledge, battery is the most linear power supply and O2 only using simple LM7812/7912 with some capacitors to regulate and even a half-wave recitifier which is more ripple than full-wave recitifier. So is the high-end power supply like the Sigma 22 really worth it when even a simple regulator circuit give the sound no different then battery or because O2 doesn't need really good power supply?

 

[duncan1]

I dont have the O2 circuit diagram but your words are basically right. This amp seems popular and I also am wondering  why only a half-wave rectifier is used with its attendant ripple .It either says we dont need it because our circuit has a large PSRR or they are saving money.
          I realise SMPS are being put forward as "as good" as linear that would only apply on the very expensive SMPS with extensive filtering.
            It all comes down to money. Unless you build it yourself electronic equipment is built to a price- No manufacturer is going to sell at a  
             loss unless they are trying to compete with another manufacturer. Even small businesses need to make profit.    
               Thats why so many people come to Head-Fi  to read the posts as -if they are going to spend their hard earned dollars then they  
                 can find the best buy here. As far as SMPS=LINEAR PS I always come down on the side of linear.for various reasons. Problem  
                  is smps are smaller/lighter/ cheaper to manufacture/ cooler due to better efficiency and thats where the world of electronics is  
                   going wont be long till the only place linear PS  are fitted are in TOTR equipment

 

[00940]

The O2 is using a full-wave voltage doubler. Dual outputs AC wallwarts are difficult to find while single output ones are much easier to get. Ripple is quite easy to get rid off, both the regulators and the opamps have large low frequency rejection. At high frequencies, good pcb layout and decoupling take over. A better supply would only be useful for an amplifier with less psrr.
 
Using a wallwart makes the amp safer and easier to build.
 
Btw, I'm surprised to read on the O2 webpage that it is using an half-wave supply. According to the schematic, it is clearly what is commonly known as a full wave voltage doubler. An oversight of the designer I guess...

 

[tomb]

Last time I looked, he feeds both regulators right after a single diode.  I can see how you can interpret the second set of diodes as a full-wave voltage doubler rectifier, but putting the regulators in-between the first and second set of diodes is awfully strange and possibly can't work as a full-wave voltage doubler.
 
I could be wrong, but it seems to me the designer's description of half-wave rectification prior to the regulators is accurate.  As you describe, he's taking credit for the noise reduction from the regulators and the inherent PSRR of the amp circuit.  (That doesn't work for all circuits, but maybe it does for his.)  He goes on to say the two diodes after the regulators (D1 and D5) are there to block battery drainage in the absence of AC power, not as the other end of a full-wave voltage doubler.
 
Just MHO, but something like the Sigma 22 applied to the O2 would be a gross mis-application - might as well put a Sigma 22 on a CMoy.

 

[00940]

I wasn't thinking about the diodes after the regulators, just the two first ones.
 

 
Top circuit is a half-wave voltage doubler. Middle one is a full-wave voltage doubler. It gives less ripple, sags less under load and the ripple is twice the ac frequency.
 
Bottom one is the one used in the O2. Works like a pair of coupled half-wave rectifiers but since it draws current from either AC cycle, it's still called full wave.

 

[KimLaroux]

Quote:

I wasn't thinking about the diodes after the regulators, just the two first ones.
 

 
Top circuit is a half-wave voltage doubler. Middle one is a full-wave voltage doubler. It gives less ripple, sags less under load and the ripple is twice the ac frequency.
 
Bottom one is the one used in the O2. Works like a pair of coupled half-wave rectifiers but since it draws current from either AC cycle, it's still called full wave.

 
I would like to disagree. The middle one is only two half-wave rectifiers mirrored. The top one receives the positive portion of the sine wave while the bottom one receives the negative side of the sine wave. The total voltage output across both capacitors is the same as if a bridged full-wave rectifier was used. The advantage this circuit offers over a bridged rectifier is the virtual ground, allowing the output to be cleanly split into a bipolar supply.

 

[00940]

What do you disagree on ? Just to make the point, using the same components and AC source (and heavy load to make ripple clearer):
 

 

It might be a matter of semantics... A voltage doubler certainly gives a lot more ripple (for the same caps) than a center tapped or dual winding transformer +bridge would give.
 
All this is quite irrelevant when one considers the noise levels achieved by the O2 even with "kind of half-wave" rectification.
 
edit: I cut the scales when snipping... Green trace is at about 15vdc with 2vpp ripple, red is at about 16vdc with 1vpp ripple and blue is at 8.5vdc

 

[Goobley]

I know Kim already said most of this but I have to reiterate it for the porpoises of what I want to add. The O2 uses a pair of half wave rectifiers of different polarities, meaning that one gets the top half of the ac wave and the other takes the bottom. Wakibaki used this same rectifying scheme on his small lme49600 amp, as it has the advantage of being small, requiring a single secondary (wall wart for example), and can be safely used to make a floating device (ie no ground loops or noise filtering in from ground).
Your point about the ripple the ripple being greater is certainly true, but a half wave rectifier should never be used for any chunky load, luckily the little amplifiers like the O2 are fairly easy on their power supplies. The O2 uses regulated +/-12V supplies from a 12V to 20V transformer, a 12V transformer rectified after half wave rectification will give 16.2V, even with 2V ripple there's still plenty enough clean voltage for the dropout of just about any reg.

To put it in a nutshell, the O2 has a competently designed power supply and I wouldn't go looking there for things to improve.

 

[00940]

This is getting confusing...
 
Goobley: all you said is fine and I agree about it since the beginning. I'm not trying to make a point about high or low ripple here. The graph's intention was to point out that you get twice the voltage with the doubler compared to a bridge rectification without center tap, a point that Kim seemed to dispute.

 

[Goobley]

Sorry, my bad, I didn't get that you were comparing the voltage output between the two schemes, I saw the comment about ripple but not the edit about the scales and thought everything was on the same scale when I looked.
Cheers!

 

[00940]

No biggie, it's my fault for bringing confusing semantics into the thread.

 

[KimLaroux]

Ah, right. I see now... it really does offer twice the output voltage of a bridged rectifier. The real difference is the bridge rectifier makes the negative half positive, doubling the current but not the voltage. The mirrored half-wave rectifier stores the negative half in a capacitor in series with the one storing the positive half. Both of these in series offer twice the voltage and half the current that the bridge rectifier offers. Clever.
 
I got confused with center-tapped transformers, where it's effectively the other way around:

 
 
I'm guessing the choice of capacitors in the mirrored half-wave rectifier is crucial. It has to be large enough not to sag too much when it discharge but small enough to fully charge with half a wave.