Thanks Jeremy,
Now I'll really have to get my 'A' into gear!
Interesting thing, Ryan was telling me that apparently John B of EC Design found that a -45mV setting on the o/p pins of the 1541A dac chip gave the best sound, which is a bit perplexing as it's well outside the Phillips +/-25mV specifications -
However, if John B mentions something like this, it's certainly worth paying attention to it.
Now, what/how is the best way to get a stable, "slightly" negative voltage on the o/p pins still supplying the required mA from the +5volt supply, hmm?!
The plot thickens!
Now I'll really have to get my 'A' into gear!
Interesting thing, Ryan was telling me that apparently John B of EC Design found that a -45mV setting on the o/p pins of the 1541A dac chip gave the best sound, which is a bit perplexing as it's well outside the Phillips +/-25mV specifications -
However, if John B mentions something like this, it's certainly worth paying attention to it.
Now, what/how is the best way to get a stable, "slightly" negative voltage on the o/p pins still supplying the required mA from the +5volt supply, hmm?!
The plot thickens!
No problem James,
a specially wound transformer comes to mind if that could be pulled off. another way would be to use a LT1010 buffer which will happily reinforce -0.045v at up to 150ma if you can supply it to the input and it will output with max 100nV noise at very low frequency (better at higher frequency). also comes in a more friendly to220 (5 pin) package
you could also do it with a dual opamp like opa1642 fed a low noise reference to both inputs and adjusted in FB to supply 0.045 more (or less) on one side, combined with something like the precision instrumentation amp LT1012
yes my LT bias is showing again, maybe i should get a T-shirt
i'm sure it could be done with Ti, AD, national et al, but i just know the LT catalogue better and this type of thing is their bread and butter.
i'm sure the same thing could be done discretely and with the aforementioned crazy accurate TX; but keeping it stable wont be as easy
a specially wound transformer comes to mind if that could be pulled off. another way would be to use a LT1010 buffer which will happily reinforce -0.045v at up to 150ma if you can supply it to the input and it will output with max 100nV noise at very low frequency (better at higher frequency). also comes in a more friendly to220 (5 pin) package
you could also do it with a dual opamp like opa1642 fed a low noise reference to both inputs and adjusted in FB to supply 0.045 more (or less) on one side, combined with something like the precision instrumentation amp LT1012
Linear technology said:
yes my LT bias is showing again, maybe i should get a T-shirt
i'm sure it could be done with Ti, AD, national et al, but i just know the LT catalogue better and this type of thing is their bread and butter. i'm sure the same thing could be done discretely and with the aforementioned crazy accurate TX; but keeping it stable wont be as easy
