Thanks for advices . Actually I m very close to put 4x lme49990 instead of 3x . Why? Because I need to have 100R at the output of each one , for stability . So in this moment the output impedance is around 33,3R . It will be so much better with 25R 🙂) . Plus ,less noise , less distorsion , more power to output , 100 ma at least . Now it s 75 ma only . I must have a buffer in the front of the resistive attenuator , because the impedance there is only 1k . Now , about the discrete stage . I really want to see one with the performance of 3xlme49990 . Maybe those op amps are to good for that job , anyway .
Last edited:
And about the power supply stage . There are many reasons to use a filter before the stabiliser . One is that it can be used only one filter and a lot of stabilisers after it . The second one is that those lme49610 are not included in a feedback loop , so the pssr is very poor . I can only hope that the first stage of the supply will improve a bit the pssr at high frequency .
I think it s easier for me to use even 4x lme49990 than to built from scratch a discrete design . And cheaper .
Last edited:
Hi,
well, as You seem to know everything and everything better, I can just wish You good luck. 🙄
jauu
Calvin
well, as You seem to know everything and everything better, I can just wish You good luck. 🙄
jauu
Calvin
calvin , I'm really sorry if you think that . If I'm not agree with your opinion , that's doesn't mean that I know everything . This means that I like more my solution , that's all , nothing more . I would appreciate much more your next ideas instead of a deep silence .
well , good luck to you too .
well , good luck to you too .
Last edited:
If you get paid per omp used, you're on the right path.
If you're looking for good sound, then follow Calvin's advices.
If you're looking for good sound, then follow Calvin's advices.
ok , let s see what I can do . First , lme49610 can not give more than 250ma , so I can not move it on the first stage of power supply , where the current is much bigger . Next , the output impedance of the stabilisers . Usually there is a resistor between the reg and each audio input stage , in the range of 2,2R to 22R . Will be a difference in my case if the reg have a very very low ouput impedance, or 0,1R ? I don t think so . Of course when you have connected a lot of consumers on the same reg , this is a critical problem . I have everything supplied separately . That s what I think .
Now , about the discret op vs integrated one . Of course integrated op are good enough for audio , even the primitive , cheap one , like njm4580 or Ne5532 . A part like lme49990 , ad797 it s far away too good for audio. Once again , this is just what I think .
How can be explained otherwise a design like Aava used in Accuphase , or the use for the input stage of P7100 , a truly exceptional amplifier , a njm4580?
Actually my concern for the sound of lme49990 is close to zero . Of course it will sound well , there is no doubt about it.
I believe that the biggest limitation of lme49990 is the output current and the output impedance . 100 ohms and 30 ma , that s all . But using 4 , the output impedance is down to 25 ohms , and the output current is raised up to more than 120 ma .
That s good enough .
the board have a mistake , I already correct it .
Now , about the discret op vs integrated one . Of course integrated op are good enough for audio , even the primitive , cheap one , like njm4580 or Ne5532 . A part like lme49990 , ad797 it s far away too good for audio. Once again , this is just what I think .
How can be explained otherwise a design like Aava used in Accuphase , or the use for the input stage of P7100 , a truly exceptional amplifier , a njm4580?
Actually my concern for the sound of lme49990 is close to zero . Of course it will sound well , there is no doubt about it.
I believe that the biggest limitation of lme49990 is the output current and the output impedance . 100 ohms and 30 ma , that s all . But using 4 , the output impedance is down to 25 ohms , and the output current is raised up to more than 120 ma .
That s good enough .
the board have a mistake , I already correct it .
Last edited:
- Status
- Not open for further replies.