Class AA Opamp Schematic

[Dave S]

Originally Posted by jcx

Jung's multiloop amps can give much better performance with the same components but input amp and buffer are in series and you have very large loop gain in the signal path

Interesting! I built a META42 using AD8066 and EL2001 buffer, this uses the Jung multiloop idea. It sounded pretty bland to me, maybe I should rig it up again to compare to the classAA board? I reckon I prefer the classAA circuit (although it's a few years since I listened to the META). Maybe it was the poor pcb layout (poor grounding and decoupling) that limited the META and not the topology?

[Bonsai]

Thanks.

I think I'll just stick to a standard op-amp configuration and output loads of 600 Ohms or more!

[Doug Kim]

Hi There,

I recently purchased this board not because it was class AA but because I wanted to build a versatile dual buffer amp using OPAMPs. This one is already soldered, there are 2 opamps per channel, so I was going to put a volume between NE5532 and NE5534 circuit so that it act as a high impedance input and low impedance output buffer with minimal effect from volume. I was going to short or remove some resistors to make this circuit a buffer.

But anyway, I got curious about the circuit and saw this thread. I attached the circuit since someone said he couldn't see it.

I understand principles of opamp but I have no experience in practically using opamps for audio. I have a few questions.

1. Would I need input or output caps? I saw a few circuits that use caps, also a few circuits that don't.

2. If I remove R3 of NE5534 (R8 of 5532), and then short R4 of 5534 (R12 of 5532), then their gain will be unity. Is this correct?

3. What will be the result if I use 15V rail?

4. Will the second half of 5532 still be OK to be left there for buffer?

Thanks,

Doug

[Mooly]

Caps on the input and output are a good idea generally. A circuit such as this has a response down to DC. That means if the source has a DC offset, then that offset will be amplified by the gain of the circuit. A cap on the input prevents that. A cap on the output prevents DC offset from the circuit itself appearing at the output. The slightest DC across a volume control will make the action "crackly" as it is rotated. The 5534/2 are very poor as regards low DC offsets, to get around that you need either a FET opamp or a more modern device with much lower input bias currents.

The resistors will do as you describe but there may be stability issues with the 5534 unless it is compensated correctly for unity gain.

Supplies can be anywhere between the chosen devices recommended limits.

Second half of 5532 can be used for whatever you want... it's independent.

[Doug Kim]

Thanks a lot for the answers.
I would add caps for 5534 input. Any suggestions: cap value, model, etc? And do I still need input resistor when I use input caps? As I understand, the input resistor is to restrict any dc current so it won't be needed for Ac coupling. I might be wrong though.

[Mooly]

You would just add a cap in series with R1 or R7. The value depends on the lower cut off frequency you desire. Fig 1 here.
High-pass filter - Wikipedia, the free encyclopedia

You woudn't normally need R1 or R7 or the 100pf caps. These resistors have no practical effect on input currents but could be used to configure a passive filter with the 100pf cap.

Unless you need an HF filter like that to address a specific issue or part of the design then I would suggest omitting it.

There are a lot of subtleties to a circuit like this. The opamp DC input currents have to be kept equal for minimum DC offset. That means the DC path on - and + inputs has to be the same resistance to achieve that. Use a FET opamp and that problem disappears. You could a TL072 and TL071 as direct swaps for these.

Also see,
Swapping Op-Amps... you have checked to see it's stable haven't you ?

[Doug Kim]

Mooly,

In my country, mooly means physics, I appreciate your mooly answers.

Thank you,

Doug

[Dave S]

The circuit works fine fully dc coupled. The 5534 has offset trimmig and the 5532 section seems to have lowish dc offset (obviously depends on the individual opamp). All the circuit impedances are reasonably low.

It is important to change the elec caps to something reasonable e.g. Pana FA etc. The Suscons fitted to the board can only loosely be described as capacitors.

I wonder what is would sound like with LM4562s?

[Mooly]

Originally Posted by Doug Kim

Mooly,

In my country, mooly means physics, I appreciate your mooly answers.

Thank you,

Doug

You're welcome