Masterpiece

[Stefanoo]

sorry for taking so long, but I got back from Italy 2 days ago and I still got jet lag so didn't do much.
Hopefully this weekend I will be able to try the new buffer and report listening impression!

 

[Joachim Gerhard]

I listen to it everyday. I have tried different power amps and speakers.
I am still very satisfied, the result is always very good, no matter the combination.

 

[Stefanoo]

I listen to it everyday. I have tried different power amps and speakers.
I am still very satisfied, the result is always very good, no matter the combination.

interesting!

 

[Stefanoo]

is the 150pF needed across CB? Have you tried to remove it to see if it oscillates without?
I will likely build one tonight.
To try it on the differential circuit I will have to build two of them, don't know if I have enough energy tonight 😡🙄

EDIT: Nevermind, just simulated and without caps frequency response gets a bump at about 30MHz

 

[Joachim Gerhard]

I use the 150pF.

 

[RCruz]

What are the RLp1 and L1 for ?

 

[Joachim Gerhard]

That is a zobel to decouple long cables. I use it.

 

[RCruz]

Thank you, what are the main differences between the two circuits ?

You built the left one ?

 

[Joachim Gerhard]

No, i build circuit B.
I promised Calvin to keep the discussion of this buffer in the original thread.
There you find all information.

 

[MiiB]

Difference is the Rail to rail cap, making additional Rc Rail filtering..

 

[FdW]

Difference is the Rail to rail cap, making additional Rc Rail filtering..

I would think that 'SE' vs 'PP' is a hint 🙂 http://www.diyaudio.com/forums/analogue-source/220309-masterpiece-69.html#post3460408

 

[Joachim Gerhard]

Yes, circuit B is push-pull. Here the CCS is driven too. That gives double the output current. I would call that a Taylor buffer.
In circuit A the CCS just delivers constant current.
I found that circuit B sounds better although it did measure the same when 10kOhm is the load.

 

[Joachim Gerhard]

Here you find some driven buffers ( followers ) :
xen-audio
Actually the circuit i build is different from a Taylor.

 

[MiiB]

Ohh I see now how the lower current source is modulated, nice thinking, could be why it sounds so expansive and dynamic.

 

[Joachim Gerhard]

Yes, that modification made the sound much more robust without loosing fine details.

 

[Stefanoo]

Yes, that modification made the sound much more robust without loosing fine details.

yeah I bet it does as the lower side is now active part in the output contribution thus increasing drivability at no cost compared to circuit A.
Very smart indeed!

 

[Joachim Gerhard]

It is also non switching. When the BJts run out of steam there are still the J-Fets.
Signal path is super simple too for the micro details.
This is a sledgehammer with class.

 

[Stefanoo]

All,

after more analisys and listening tests I am trying to determine if the problem I have heard is just due to something being faulty on my board or not.
It seems to be an overaload margin problem.
What I hear is clipping on high peaks. At first I thought it was my cartridge bu then I put an inverse RIAA and iPad and it shows the same problem.

So I put the phono back on the bench and I was unable to find any problem with it.

However I noticed that at high frequency the clipping looks pretty ugly.

Also I have noticed that the two halves are not so symmetric although I am using dual matched packages for the input stage.


Any thoughts on this? I will run some simulation to double check some parameters.

 

[Joachim Gerhard]

Are you using both outputs in balanced or do you use only one output ?

 

[Stefanoo]

Are you using both outputs in balanced or do you use only one output ?

I am running full balance from input to output.
So yes both outputs.

How can we improve the clipping behaviour? It is something I have no clue how to solve.
I am planning on doubling up the input device and running about 35mA on the mirror that drives the RIAA.