Diodes & Rectification - best practice or superstition?

[juz_ad]

Hello,

I'm reassessing some decisions I made a while back regarding IO best practice - particularly with regard to my half-wave rectification circuit and specifically that diodes should be kept away from external inputs.

A) shows my current set-up where IC1A is an input at 100K impedance for an external source that could be AC or DC and anywhere between approx. +/-10V at anything from 20KHz down to LFO/Gate on a +/-12V supply. This feeds a half-wave rectifier which feeds a buffer. Nice stout input, dedicated half-wave rectifier, nice stout output (at 1K impedance).

So - ignoring the inversion path - is there any reason why I shouldn't just use B) with R5 setting the input impedance and
R9 setting the output impedance?

And a follow up question - any reason why I shouldn't swap the 914s for 4148s? Seems like 4148s come in many more package sizes than the 914.

I've got it breadboarded here in isolation and it seems to be working just fine...

Thanks,

J

 

[kripacharya]

i can't see any obvious reason not to. Though you could use schottkys instead, and so (i think) ease up on the opamp specs a bit. The regular 4148 7 914's with ~0.7 would require a higher slew on the opamp ... would they not ?

 

[zorro]

The output impedance of the rectifier stage (i.e. at the node of cathode of the output diode) is very low in the conducting half-wave but high (the value of the feedback resistor) in the other one.
The output buffer avoids that the load or the following stage "raises" the output voltage.
On the other side, I consider that (except for inversion and gain) the input stage could be avoided.
Regards

Z