Having found dual op amp packages around a fifth the price of comparable single packages from online distributors, I think i'll go for the double. (BTW, I'm using supply of +16, and ground).
I don't want the unused channel drawing current unnecessarily though, or bouncing the output around.. - If I tie the non-inverting input to +3.3V, and have a high resistance feedback resistor to the inverting input (with no other input to that pin), would that be as good as anything? I figure the output won't have to 'try' very hard to modify the voltage of an input pin with no other voltages on it..
> Having found dual op amp packages around a fifth the price of > comparable single packages from online distributors, I think i'll go > for the double. > (BTW, I'm using supply of +16, and ground).
> I don't want the unused channel drawing current unnecessarily though, > or bouncing the output around.. - If I tie the non-inverting input to > +3.3V, and have a high resistance feedback resistor to the inverting > input (with no other input to that pin), would that be as good as > anything? I figure the output won't have to 'try' very hard to modify > the voltage of an input pin with no other voltages on it..
> Am I right?
> Thanks..
Just tie the inverting input to the output, unity gain. Three point three volts on the noninverting input is OK. You can usually just ground this point, in this case connecting it to the negative rail. Ideally the common mode range of the input should include the negative rail but that may not be necessary depending on the particular amp.
> Having found dual op amp packages around a fifth the price of > comparable single packages from online distributors, I think i'll go > for the double. > (BTW, I'm using supply of +16, and ground).
> I don't want the unused channel drawing current unnecessarily though, > or bouncing the output around.. - If I tie the non-inverting input to > +3.3V, and have a high resistance feedback resistor to the inverting > input (with no other input to that pin), would that be as good as > anything? I figure the output won't have to 'try' very hard to modify > the voltage of an input pin with no other voltages on it..
> Am I right?
> Thanks..
As long as you've chosen a single supply op amp (like the LM358) with input range extending to the negative rail, just grounding the non- inverting input on the unused op amp/voltage follower should be OK. But even if you've got an op amp like the LM1458, 3.3V should be fine. If you've got an LM358, GND will be lower current, because it doesn't have the noise of another voltage input.
The "high resistance feedback resistor" won't help. At worst (for high resistance values), it can encourage oscillations, which will dramaticaly increase quiescent current (charging and discharging the internal compensation cap). Just connect the inverting input directly to the output to make a voltage follower, and save the price of the resistor.
By the way, make sure you change the circuit layout for the new op amp pinout.
On Fri, 4 Jul 2008 01:20:49 -0700 (PDT), angus.oli...@gmail.com wrote: >Hi All,
>Having found dual op amp packages around a fifth the price of >comparable single packages from online distributors, I think i'll go >for the double. >(BTW, I'm using supply of +16, and ground).
>I don't want the unused channel drawing current unnecessarily though, >or bouncing the output around.. - If I tie the non-inverting input to >+3.3V, and have a high resistance feedback resistor to the inverting >input (with no other input to that pin), would that be as good as >anything? I figure the output won't have to 'try' very hard to modify >the voltage of an input pin with no other voltages on it..
>Am I right?
>Thanks..
Yup, keep it closed-loop, away from the rails. You don't need the resistor, just connect the output to the IN- input. 3.3 volts on the IN+ input should be fine for any opamp.
Some, maybe most, opamps draw a lot of extra current if railed, so don't tie the ni input to V+ or V-/ground. Some will even mess up the other amp sections if one section is run open-loop.
If you don't have a suitable dummy voltage handy, and the signal isn't really fast, you can use the extra opamp section as a follower of the section that you do use; that keeps it happy, too.
> Having found dual op amp packages around a fifth the price of > comparable single packages from online distributors, I think i'll go > for the double. > (BTW, I'm using supply of +16, and ground).
> I don't want the unused channel drawing current unnecessarily though, > or bouncing the output around.. - If I tie the non-inverting input to > +3.3V, and have a high resistance feedback resistor to the inverting > input (with no other input to that pin), would that be as good as > anything? I figure the output won't have to 'try' very hard to modify > the voltage of an input pin with no other voltages on it..
> Am I right?
** Use the spare op-amp to do something - like buffer the output of other one.
I've pretty much settled on a LM2904, with: ” Input common-mode voltage range includes negative rail ” Differential input voltage range equal to the power supply voltage ” Large output voltage swing 0V to (V CC+ -1.5V)
Taking into account all your comments, I think I'll turn it into a(n unnecessary) voltage follower (my op amp is only to scale a voltage for an A/D conversion). I'll have the output of the used amp as input into the non-inverting input, and tie the output of this second amp back to the inverting input for a unity gain voltage follower with "No extra current draw involved" (Phil).
