[sdiy] BBD question

[JH.]

>Any particular reason why the S/N ratio of four 512-stage BBDs in
>series would be worse than that of a single 2048-stage BBD ?
>(Assuming the filters do not contribute appreciably to the noise.)


Not sure. But maybe the longer BBDs have been designed with a higher demand 
on "SNR per stage" in mind - but then, maybe not. (I rather think they were 
all optimized withing a series, and the smaller ones naturally being less 
noisy.)

One thing about cascading short BBDs: The output level is often smaller than 
the input level, so in order not to waste SNR, it's advisible to put some 
amps for makeup-gain between them, rather than just butting one such amp at 
the end of the chain. Thinking of it, that could even be an advantage over 
one long BBD - theoretically.

I haven't made comparisons myself, but the low voltage BBDs have a 
reputation of being more noisy than the old 15V MN3000 series. But if you 
can get a coolaudio version for 1/25 of the going rate of a MN3005, you 
might as well *parallel* 25 of them to improove their SNR. :)
(And build a hell of a driver circuit ... no better forget about that.)

Just some unconnected thoughts. (People keep asking me about a BBD echo PCB 
project, so I at least spend _some_ thoughts on it from time to time. ;) )

JH.