ONE MORE P3A QUESTION

Status
Not open for further replies.
Jump to Latest
#10
Bootstrap

Or, to keep the amplifier running as present, you could bootstrap the input as per Self's illustration. Chapter 4 in the current 5th edition of his Audio Power Amp. Design Handbook. I haven't tried it but this should keep internal noise at present level while matching to the higher impedance set by the resistive divider added at the input. ..... BTW, It's also part of the trimodal amplifier, if your edition is earlier.
 
Last edited:
#11
ok ...thanks all for the replies ...will make some tests and get back to u

kind regards sakis.

( idea originates from kenwood 5700 . Had one repaired some days ago and this amp features a total passive "preamp" just a switch input selector obviously amplifier board with some increased gain and an input of 100K ... that actually completes the picture )

as about the tone control ( which is not an issue here ) is standing also passive on the feedback node ....
only active thingy is the phono stage of the amp ...

thanks sakis
 
#12
As the previous poster pointed to, there is a reason amps no longer have very high input impedances... the selfnoise of high ohmage resistors. It may be more forgiving in old (valve) systems etc. Due to the ability to lift the signal amplitude way above the noise floor.
 
#13
digits said:
As the previous poster pointed to, there is a reason amps no longer have very high input impedances... the selfnoise of high ohmage resistors. It may be more forgiving in old (valve) systems etc. Due to the ability to lift the signal amplitude way above the noise floor.
Click to expand...

correct i think but still amps with input of 22-47k if driven with a passive pot obviously have absense of dynamics when used in low power ...( that will be average casual listening )

Testings with B1 Nelson Pass buffer had better behaviour but not actually the best .

Best was so far the Death of Zen but still passive is the cleanest you will ever get .

Kind regards sakis
 
#14
P3A has about 2mA tail current;
If the imput devices have Hfe of 200 , they will drain about 5uA base current
through the 100K biaising resistors , that will see a 500mV voltage drop.

Thus , if the input biaising resistor and the feedback resistor have 5%
tolerance mismatch, this will create a 25mV DC output offset.

The solution is either to use higher gain input transistors with beta of about 500
at least , or to reduce the biaising resistors values , wich is precisely what has
done Rod Elliott , to reduce their voltage drops and consequently input
and feedback tolerance mismatches.
 
#15
Use well selected devices for the input LTP pair and the input offset current will be very low. I would not expect any competent builder to select his input components so that they alone gave a 25mV output offset. A little work can easily get that down below 10mV and with care down below 2mV over a wide range of operating conditions. Some builders can achieve +-0.5mV of output offset.
Who uses unmeasured 5% resistors around an LTP pair?
 
#16
sakis said:
correct i think but still amps with input of 22-47k if driven with a passive pot obviously have absense of dynamics when used in low power ...( that will be average casual listening )

Testings with B1 Nelson Pass buffer had better behaviour but not actually the best .

Best was so far the Death of Zen but still passive is the cleanest you will ever get .

Kind regards sakis
Click to expand...


Lol, I am of the no buffer, no preamp brigade. I ripped up just about every pre I ever made and eventualy just left the potentiometer and input selector in the box....
 
#19
Most builders have at least one resistance meter.
It's easy to compare resistor values in the range 10k to 200k.
There is no excuse for not measuring resistors. Stating 2.2% mismatch is equivalent to the production engineering done to minimise cost vs performance for cheap product coming off the mass production line.
DIYaudio can do 10 times better, if they try.
 
#20
AndrewT said:
Use well selected devices for the input LTP pair and the input offset current will be very low.
Click to expand...

lineup said:
hFE mismatch can cause a difference in input bias current
and so additional offset difference
Click to expand...
Yes.
Neither resistor tolerance nor resistor value are the offset issue.
Input offset current and methods to reduce that are the issue for resource strapped builders.
 
Status
Not open for further replies.
Top Bottom