i am currently using the ti tps71733 ldo voltage regulator (3V3 output) and using 1uF on the inputs and outputs along with a 10nF on the bypass pin as per the ds.

i would like to be able to also use the fairchild fan2503 in the same circuit as a drop in replacement, but its datasheet wants a 2.2uF on the input side with a 1uF on output and 10nF on bypass, so the 2.2uF input cap is the question.

how important is the 2.2uF vs 1uF difference on the input?

ti tps71733 ds - http://focus.ti.com/lit/ds/sbvs068g/sbvs068g.pdf
fairchild fan2503 ds - http://www.fairchildsemi.com/ds/FA/FAN2503.pdf

thanks in advance for your expertise,
bob

 

Page 11 of the TI datasheet says that an input cap is optional and that including one will only make life better. Using 2.5uf should not pose a problem. The Fairchild datasheet spells out the important requirements for that capacitor.

 

since i have the quantity of the 1uF i have on hand, is there any issue putting 2 in parallel? would 2uF be acceptable?

 

You can try not obeying the rules, and maybe it will work. If it doesn't, follow the rules and install the right capacitor.

 

Three would be better than two. Do your caps meet the other requirements?

 

yes, the caps meet the right requirements.

one of my main reasons for asking is it seems sometimes capacitance value is not very important, other times it is and i was not sure if this time it was an important value or one that could be changed. also, the 2.2uF are 6x more expensive than the 1uF and in a one off prototype no big deal, but as quantity goes up i would prefer to spend .05 vs .35, but not at the expense of possibly taking more time chasing an issue if it does not work correctly.

how is it determined if the value needs to be strict? or is this data gathered strictly from the ds?

 

The datasheet is the manufacturers guarantee of specs. Sometimes you can rely on the fact that most chips are better than absolute minimum guaranteed performance. Sometimes you can't.

Op-amps usually have smaller offset voltage than the sheet says. Voltage regulators will usually pass more than the minimum guaranteed current during self limiting mode. Temperature during current is pretty much absolute. Frequency stability is one thing the chip engineers are way ahead of me with. I don't second guess that spec.

If you don't want to pay for caps, use the TI chip that Joe mentioned in post #2.

 

i think that is what i am going to do, thanks for the input.