[sdiy] CV input op-amp circuit

[Brian Willoughby]

Always check the data sheet. It should hopefully spell out the protection diodes as well as any resistance or capacitance on the ADC inputs.

The corollary is true, too: Some parts - especially high-precision op-amps - don't have protection diodes. For those parts, a simple current-limiting resistor is not enough to protect the part from exposure to excessive voltages. You'd have to provide the protection diodes in those cases.

Brian


On Dec 4, 2020, at 12:07, Tom Wiltshire <tom at electricdruid.net> wrote:
> Chris, is the -5V to 7V range a hard limit or just the “expected” input?
> 
> If it’s the expected input, the 3.3V single-supply rail-to-rai op-amp clipping is preferred, since it *can’t* go beyond what the microcontroller input can handle.
> 
> A circuit like David’s that scales and offsets is fine while the inputs are reasonable, but the output is potentially going within a diode-drop or two of the +/-15V rails which is plenty to fry the poor micro.
> 
> Another think to mention is that I’ve had good results just putting a series resistor ahead of the ADC input. The micro will have protection diodes that prevent over-0voltage, but too much current can fry them and leave the chip unprotected (=instant death) but with the resistor, the current can never get high enough to kill the diodes and the protection survives the abuse.
> 
> On the PICs I used, a 4K7 value was enough to save the input from all practical abuse. The higher the value, the slower the ADC can be read (series resistance increases the acquisition time) so be aware of that limitation.
> 
> Tom
>  
> On 4 Dec 2020, at 19:36, David G Dixon <dixon at mail.ubc.ca> wrote:
>> Or, if you prefer, here is the same circuit with a -5V to 7V sinusoid, coming out at 0 to 3.3V (ya gotta love Multisim):
>>  
>> <fart.png>        <CVShiftSmall.png>