[sdiy] Bipolar Rail to Rail Opamp recommendation

René Schmitz synth at schmitzbits.de
Mon Mar 2 20:27:09 CET 2020

Hi Rafael and list,

On 02.03.2020 15:46, Schwarz Raphael wrote:

>> Le 2 mars 2020 à 15:45, Schwarz Raphael <raphschwarz at gmail.com> a écrit :
>>>> I read it so, that you're actually interested in clamping the input voltage. Why not bias the diodes to +5.8 and -5.8v instead?
>>> It is more like +-4.2 actually. :)
>> It’s actually a good idea
>> I realized I overlooked the variation with forward current and temperature,
>> I was a taking for granted the 0,8v drop , at 25° and 10mA it’s actually much less
>> About 0.24v on a bat54s (what I was using)
>> About 0,65v on a 4148.
>> Real life testing of a bunch of diode is consistent with this. My simulator has a very simplified model of diodes
>> I wonder if biasing to about 4.75v is a practical solution with part tolerance , out of a bunch of bat54s I randomly tested it seems to only vary by a few mV
>> temperature variation is not crazy for 10ma datasheet says 0,3v at 0° and 0.2v at 75°. So about 1% of 5V from 0° to 75°
>> So it might be doable

You can derive the bias voltage from a diode also. That would cancel any 
temperature effects. Since you can compensate for the forward drop, you 
don't actually need the schottky.

>> I still think a good rail to rail opamp would be a more elegant solution

I would say this depends on how many inputs need this treatment. The 
bias voltage can be re-used for several inputs, each with its own pair 
of diodes.

And a rail-to-rail opamp might sit away 50mV or so away from the rails. 
Depending on the load.


synth at schmitzbits.de

More information about the Synth-diy mailing list