I have two power supplies running in parallel, if one stops working ... the other takes its place.
They both generate 5V (as well as 48V).
I'm using a MCP1703 to generate 3.3V for two microcontrollers

Now, having two microcontrollers I want two separate 3.3V supplies, so I simply use two same LDO:


It would only be partitioned the current of the 5V on the two inputs of the LDOs

At the level of "safety" and design, is this okay?
Are there any design mistakes that I am missing?
Perhaps there are integrated LDOs with two inputs and two outputs isolated from each other and in the same chip?

I hope the question is not stupid.
Thanks.

 

I want two separate 3.3V supplies

Why?

 

That does not give you isolation, if that is what you want.

 

I don't care why you want to use 2 regulators, but I am curious as to why you have chosen zeners?

 

I don't care why you want to use 2 regulators, but I am curious as to why you have chosen zeners?

Zeners? I see Schottkies, and some LDO regulators?

 

Zeners? I see Schottkies, and some LDO regulators?

That is a symbol for a Zener.
Below is one for a Schottky diode.

 

That is a symbol for a Zener.
Below is one for a Schottky diode.
View attachment 348795

Depends who you ask. According to IEC60617 . . .

 

yes, that is another symbol used for Zener. but why do you think that symbol shown in original post was a Schottky?

 

yes, that is another symbol used for Zener. but why do you think that symbol shown in original post was a Schottky?

 

where did that come from? may not be standard use. to me that is symbol for Zener:

 

found it : https://www.geeksforgeeks.org/schottky-diode/

that image is not right, scroll further down and you will see correct symbol...
even there, more mistakes, text and drawing "a" do not agree... drawing is correct, the diode becomes forward biased when transistor is on. Vce+Vd is less than Vbe would be without diode. this results in starving base current and preventing deep saturation so transistor can turn off faster.
sloppy website...?

 

It seems that the International Electrotechnical Commission did not define a symbol for a Schottky. A google image search gets plenty with two double-serifs and quite a lot with two single-serifs, and another group which are badly printed enough to be either.
So there is some logical sense in having a single serif for a zener, two serifs in opposite directions (with either one or two right angle bends) and two serifs in the same direction for a tunnel diode. This avoids the possibility of a badly drawn symbol with two double-serifs being ambiguous. In the case of ambiguity, then the circuit context should determine whether it is a Schottky or a Zener. At a guess, the TS intended his diagram to show Schottkies (and even if he did intend to draw Zeners, he's hardly likely to admit it)

 

LTspice uses the symbol on the left for a Zener and the symbol on the right for a Schottky:
I believe that's fairly standard.

 

Exactly!
The schottky’s cathode’s bar resembles an “S”, with a little imagination.
But I do agree that of all the semiconductor symbols, this is perhaps the most confusing.
Honorable mention: n-chan JFET and the UJT.

 

Why?

That does not give you isolation, if that is what you want.

If LDO 3.3V has a fault, the power supply is interrupted by both microcontrollers (I can't allow that!)

I think using two separate ones decreases this probability since they should both "break", if you have other ideas and considerations I welcome them

Of course I assume 5V has no problem

 

Okay. Assuming each micro performs a function that can go on without the other one, that makes sense.

 

Okay. Assuming each micro performs a function that can go on without the other one, that makes sense.

exactly

 

Okay. Assuming each micro performs a function that can go on without the other one, that makes sense.

Does the solution I proposed in the main post of using two LD0s in "parallel" with each other make sense?
Could it create additional problems that now escape me?

That does not give you isolation, if that is what you want.

Why? Can you explain?
You mean there is no isolation between the two 3.3V outputs in case of a shortcircuit on 5V?
(which then propagates into both idue LDOs)

Thanks