Issues around U5A comparator on Uno board

Hi,

Here are some issues I see around U5A comparator on the Uno (and maybe others ?) board :

This comparator is a LMV358IDGKR from TI and according to its specs, some recommended about power supplies by Arduino are wrong :

1) The pin voltage should not exceed 5.7V

Or, if the Uno is powered by a 12V source (which is inside the 7-12V recommended), V+ (voltage at the + pin) will rise to 6V.
0.3V, no big deal ? Maybe (because in fact I tried and it works...) but 5.7V is an absolute maximum rating in the specs, so the component is not supposed to be stressed this way for long.
So why taking the risk ? Why not recommend only 11V max to be sure to stay under max ratings ?

2) The absolute max differential voltage between - and + pins should not exceed 5.5V.

Or we can see for the Uno a maximum (not recommended) power supply of 20V, which makes a V+ of 10V and so a differential voltage of 6.7V !
Why again taking the risk to write a max at 20V ? (btw, with point 1) limiting at 11.4V, there is no reason to write anything above 11V as a max, and maybe 10V as recommended...)

Thank you for reading and I hope some people will have answers or additionnal information.

And maybe that Arduino recommended operating settings will be changed to lower values ? :slight_smile:

Here are some issues I see around U5A comparator on the Uno (and maybe others ?) board :

Something tells me you mean USB (not USA, where I am right now)

V+ (voltage at the + pin)

(AKA Non-Inverting input pin)

OP AMP-LM358

The Arduino board has a 5V voltage regulator. The ATmega chip and other electronics operate at 5V.

Your op-amp/comparator should also be powered at 5V. If you need to "compare" voltages greater than 5V use [u]voltage dividers[/u],

Just like the one used by the comparator on the board.

That is exactly the point. The voltage divider has equal value resistors. So what is 12V/2 ? (6V)
His point is the maximum voltage on any pin is 5.7V (page 4, Absolute Maximum Ratings datasheet

6V>5.7V

(duh)

What the OP failed to mention is why he is talking about the LMV358IDGKR when the UNO schematic shows the LM358D which has a supply range up to 32V.

His comment about the UNO voltage dividers don't apply to the genuine UNO which uses the LM358D.
Are there other "UNOs" that use some other chip ? ( I have no idea)

I understand what the OP is saying but not his reason for saying it.
Unless he can back up his statement with facts , the only thing I could say is DON'T use the LMV358IDGKR op amp for a USB (NOT USA) comparator.

Hi !

Excuse me for the delay.

Thank you all for your replies.

About U5A, it is really the U5A AOP used as a comparator between 3.3V and Vin/2 in the Genuino Uno board.
I just opened the schematic from the Uno page to verify.

And it is a LMV358IDGKR.

So the problem remains :slight_smile:

I have the same UNO, but I can't see a problem ...

Both input diodes of the OPA are the reason why it works even if we are out of specs.

BUT when there is an absolute max given at 5,7V in the component datasheet, why not respect it ?

Same thing for max differential input voltage. Even if this limit is beyond the first one.

What might happen, is that above 5,7V, you are above the Vf of the input protection diode (with Vcc = 5V), and thus you are doing a short circuit between Vin and Vcc. With a 10K resistor, the shortage current will not be an issue until Vin reaches some high voltage value to damage the diode.

But this is not a good way of doing electronics...
There are always side effects to a short circuit, you don't control then all current paths and that can lead to an issue anywhere on the board.

Furthermore, max ratings specs are here to prevent any damage...
My point is still : "why not respect the max ratings ?"

My point is still : "why not respect the max ratings ?"

My point is that the max ratings are being respected. The absolute max of 5.7V won't be reached because because the protection diodes will clamp the voltage at less than this. With Vin @12V, it will probably be near 5.3V, but I haven't measured it. Why not try a multimeter to get an actual reading? Also, the current through the protection diode will be much less than 1mA (actually 6-5/10K=100µA) which is well within its own max rating.

Could even connect mains voltage (240VAC) through a resistor that limits the peak current to less than 1mA and have no problem. Looks like Atmel's not worried. Actually, there many thousands of products out there with sections of electronics non-isolated from mains that meets regulatory safety approvals and is rugged with long life.

Myself, I have no fear of a 6VDC signal, with its current limited to 100µA, connected to an IC pin with diode protection. Connecting 6VDC directly to the pin ... well, that may let the smoke out.

Indeed you are right, V+ won't increase above Vcc+Vf, the exceeding voltage will be around RIN1A and with 10K, the margin in the resistor power and the diode current is quite comfortable.

But, in this case, you have current sinking from Vin to Vcc, and that is something to avoid.

And more, you are making some guessing about internal AOP components caracteristics, because when it is written 5,7V max in input, you can't be sure if they have have considered the fact that if a resistor is present it will take the overvoltage, or if only the fact that they want to avoid a permanent current sinking between power supplies.

Jerjer:
But, in this case, you have current sinking from Vin to Vcc, and that is something to avoid.

And more, you are making some guessing about internal AOP components caracteristics.

  1. Can't avoid that.
    When the Arduino starts up, VCC is 0volt. V-in is there before the chip's VCC.

  2. Blame the chip manufacturer. Max input current should be in the datasheet.
    Without it you can't design a safe circuit.
    Leo..