Self designed tube amp problem

[Manowar1985]

Yes they are.I will upload the schematic as soon as I draw it

[Manowar1985]

Here I upload the schematic

[eschertron]

Add another decoupling node after the first two preamp stages. start there.

[Helmholtz]

Here I upload the schematic

I don't think you can use point A or B as NFB inputs, because the excessive open loop gain and phase shift involved is likely to cause instability.
IMO it is not advisable to extend the global NFB loop further than the PI in your design as your driver stage should have lots of gain.

[Manowar1985]

Eschertron thanks I ll try it

Helmholtz what is IMO and Pi and how should I apply NFB

[Helmholtz]

Helmholtz what is IMO and Pi?

IMO: "In my opinion".
PI: Phase Inverter, aka Phase Splitter.

[Manowar1985]

What are the proper modifications to apply global nfb?

[Helmholtz]

What are the proper modifications to apply global nfb?

Try to feed NFB to one of the ECC88 driver grids (depending on OT phase). preferably the lower one as it's lower impedance.
Adjust the NFB series resistor to desired feedback level.

[Manowar1985]

Ok.Should I reduce the number of preamp stages?And what about adding decoupling nodes at the preamp stages.Is this responsible for noise or instability.?Thank you

[pdf64]

Stages within the power amp negative feedback (NFB) loop might reasonably be considered part of the power amp. Hence there isn’t a preamp
Have a look at a tremolo oscillator; only 3 RC filters (each eventually reaching 90degrees phase shift, so a max of 270) in a loop turn NFB into PFB. Now consider how many RC & RL filters are within the loop of your schematic; oscillation is pretty much inevitable.
https://www.aikenamps.com/index.php/...ative-feedback
https://www.aikenamps.com/index.php/...emolo-circuits

[Manowar1985]

Just to be clear. Are you talking about the rc filters at the power supply rail for ripple minimization or the rc filters formed by the coupling capacitor and the grid leak resistor of each stage

[Dave H]

Ok.Should I reduce the number of preamp stages?

Yes, I think you should. As it stands the gain could be 100,000 from input to EL84 grids. EL84s only need about 10V rms for full output so that's 0.1mV rms at the input for full output, no wonder it's noisy!

[Manowar1985]

Ok I will remove the μ follower leaving the schematic with one μ follower and one cathode follower with active load before the Splitter. How's that?
Should I remove cathode follower

[Helmholtz]

Ok I will remove the μ follower leaving the schematic with one μ follower and one cathode follower with active load before the Splitter. How's that?
Should I remove cathode follower

Removing one of the µ-followers seems a good idea. The cathode follower doesn't contribute gain so it doesn't hurt but doesn't seem necressary. Why did you design it it in?

What is going to be connected to the amp (signal level)?

Did you already provide another decoupling filter before the phase splitter and try my advice regarding NFB?

[Manowar1985]

I designed it for light loading of the previous stage because the ecc83(previous stage)has high plate impedance.So do I remove it?
I am going to connect audio from smartphone
No I did not do it yet But I am considering it.