noise floor problem

[johndiy]

I have built this zen circuit as a headphone amp, all works
well except for its noise floor i use sensitive headphones and
these make the problem worse

in the power supply i have used a total of 10000uf capacitance
to tame the ripple but noise floor is still too high

whats the problem?

can you help

john

 

[MikeB]

You might try a resistor to the gate. Eventually the mosfet oscillates.
Shouldn't the mosfet have a small resistor at the source ?

Mike

 

[AndrewT]

Hi,
just guessing, but, have you snubbered the PSU rectifier?

How about RCRC type smoothing after the recifier?

Headphones are very critical of noise and demand a very good PSRR of the amp and/or very low noise from the PSU.

 

[johndiy]

Hi,
just guessing, but, have you snubbered the PSU rectifier?

How about RCRC type smoothing after the recifier?

Headphones are very critical of noise and demand a very good PSRR of the amp and/or very low noise from the PSU.

"snubbered the PSU rectifier" yes

rcrc no i dont want voltage drop at the filter

i have use 10000uf should i go higher?

john

 

[johndiy]

You might try a resistor to the gate. Eventually the mosfet oscillates.
Shouldn't the mosfet have a small resistor at the source ?

Mike

it is not oscillation, small resistor at the source? why?

john

 

[parsecaudio]

.

The source resistor is in order to put a feedback for bias current control and stabilization.
Your amp has the problems of all too much simples circuits.
A low noise circuit is suggested for headphone amp.

 

[richie00boy]

What is your actual problem? Is it actually noise (which is broadband) or is it hum? If it's hum then it may simply be a grounding issue and nothing is wrong with your circuit or PSU. I know headphones can be sensitive but 10000uF should be more than enough by a long way.

 

[gmphadte]

The voltage used is not known but still...

Use an LM317 circuit with cap at adj pin or use a capacitance multiplier with a darlington.

Gajanan Phadte

 

[parsecaudio]

The philosophy of this circuits is the employ of the minimal count of the components that they are in series with the audio signal.
A voltage stabilizer integrated circuit it is not adapted to this philosophy.
This amp have a very low PSRR and the supply noise
(above all without any circuit of ripple rejection) it must to be tolerate....

 

[AndrewT]

Hi,
poor PSRR, simple C PSU (snubbered), and headphones are a conflicting set of requirements.

Something has to give.

 

[parsecaudio]

Hi johndiy
My advice is that a little modification in the power supply you must made.
A suggestion is to supply this circuit with a rechargeable battery.
If the ground are correct,the supply noise will be reduced.

 

[johndiy]

Hi everyone and thanks for your excellent inputs

this circuit has a low psrr and needs a good ps to work
best ,if you have any good ps design please post them
here

john

 

[FastEddy]

richie00boy: " ... If it's hum then it may simply be a grounding issue and nothing is wrong with your circuit or PSU. I know headphones can be sensitive but 10000uF should be more than enough by a long way. ..."

AndrewT: " ... You might try a resistor to the gate. Eventually the mosfet oscillates. ... " " ... poor PSRR, simple C PSU (snubbered), and headphones are a conflicting set of requirements. ..." " ... How about RCRC type smoothing after the recifier? ..."

My suggestions:

In addition to trying a (metal film) resistor (1k ?) in the FET input gate ... AND a poly snubbing cap (about 0.5 uF) across the main PS filter cap ....

... is to add a ferrite bead to the input gate as well, as close to the gate as possible. There may be some HF noise entering the audio input stream, driving the input FET too hard.