Hello everybody !

Im trying to make a comparator with a LM319, and it doesn´t work .

Does anybody know what the problem is ?

 

Your schematic looks OK. Do you have DC blocking caps between your signal sources and the inputs? If so, you need a resistor from each input to ground.

 

Thanks Mr. Ron ! I made the circuit in real life but it didn´t work , any solution ?

 

Thanks Mr. Ron ! I made the circuit in real life but it didn´t work , any solution ?

What do you get on the output?

 

Define "doesn't work".

 

I have to do this kind of pulse, i can make it with a PIC but i need a pulse with +9V , so i need a high speed amplifier or comparator,That´s why i used a LM319.

The circuit in real life is the following:

 

What do you get on the output?

the output voltage is 0V , the LM319 doesn´t wanna work

 

Define "doesn't work".

by " doesn´t work" i mean that the configuration for a comparator do nothing , i don´t if i am clear, sorry for my english

 

The internal bias network is common to both comparators. Floating inputs on the unused comparator will upset the bias on the one you are using. If you tie both unused inputs to ground, the unused comparator may oscillate. Try connecting one of the unused inputs to ground, and the other one to your 3.5V divider.

 

A 1uS pulse is going to take around 10MHz bandwidth or better to pass. How did you lay out your circuit? If it's on a breadboard with long, loopy interconnecting wires and long leads to your caps, that's where your problem lies.

 

A 1uS pulse is going to take around 10MHz bandwidth or better to pass. How did you lay out your circuit? If it's on a breadboard with long, loopy interconnecting wires and long leads to your caps, that's where your problem lies.

I know that working with a pulse width of 1useg is not a good idea working in a breadboard, thats why i used a sinusoidal wave , a triangular wave and a square wave of many frequencies, those signel were compared with a constant voltage value (3.5V) . But i couldn´t get something at the output of the LM319 , just 0V .

 

Did you try what I suggested in post #9?
Also, post a picture of your breadboard.

 

Did you try what I suggested in post #9?
Also, post a picture of your breadboard.

I dont have it right now, and i´m working in other project, on friday i am going to do that and i´ll post a picture of my breadboard. sorry for the delay

Thank you all for your help

 

I'll ask for the third time: Did you try what I suggested in post #9?

 

I'll ask for the third time: Did you try what I suggested in post #9?

i did what you suggest and.... IT FINALLY WORKS.

Thank you so much Mr Ron

 

If the OP wants a zero to 9 volt peak signal, why is he using a +/- 12V supplies? That would give +/- output.

The OPs hand drawing indicates the input and output signals.

 

If the OP wants a zero to 9 volt peak signal, why is he using a +/- 12V supplies? That would give +/- output.

The OPs hand drawing indicates the input and output signals.

According to the datasheet, he needs the negative supply if his input swings to ground (see input voltage range spec in datasheet), although, if the other input pin never goes below 3V, it should work OK without the negative supply.
LM319 has separate negative rail connections (pins 3 and 8) for the output. In his schematic, they are connected to ground, so the output will not go lower than that.
Perhaps you were thinking of LM393?