I have built a low frequency oscillator using lm324.

the setup is being powered by a dual supply of +5v and -5v , having a ground.


i would like to drive 2 leds , one for the positive part of the cycle and one for the negative part of the cycle without loading the signal. i know i could just connect the output signal to two inverted LEDS in parallel connected to ground but i do not want to load the signal.


can any one of you guys tell me how i can use generic transistors to do such a thing ?


2n2222 ,3906 , 3906



the positive signal led was not so hard to do using 3904, its the negative part that's making me crazy.

 

Is the opamp output a square wave? If not, do you want the negative indicator to turn on before the output is -0.7V?

 

No the output of the opamp is almost a sinewave with some minor bumps on the decay.



if really there is no other option ill take the solution tih the led not turning off before the 0.7v then.....

 

Try this


If you want the negative indicator to turn on closer to ground, make a 0.7V reference and use that instead of ground for the emitter connection.

 

yes sir , did the trick , i dont think i will actually do anything about the 0.7v even though it is noticable.


Thanks alot


marlon

 

I have built a low frequency oscillator using lm324.

the setup is being powered by a dual supply of +5v and -5v , having a ground.


i would like to drive 2 leds , one for the positive part of the cycle and one for the negative part of the cycle without loading the signal. i know i could just connect the output signal to two inverted LEDS in parallel connected to ground but i do not want to load the signal.


can any one of you guys tell me how i can use generic transistors to do such a thing ?


2n2222 ,3906 , 3906



the positive signal led was not so hard to do using 3904, its the negative part that's making me crazy.

If both LEDs are red - you can get ultra-efficient LEDs that give a useable indication with only 2mA, the LM324 should be able to handle that with nothing more than a couple of current limiting resistors.

There are some pretty good blue and/or white LEDs nowadays, but with a Vf of about 3.4V - the supplies you mentioned probably don't have enough headroom.

 

ok on a separate note , is it possible i make another 2 leds , one for when the voltage is increasing and one for when the voltage is decreasing?

 

yes ian , the leds are red but i preferred having a very tiny current to drive a transistor rather than connecting the led directly

 

is it possible i make another 2 leds , one for when the voltage is increasing and one for when the voltage is decreasing?

It will take more than just 2 LEDs. What is the frequency range of your oscillator and how much additional circuitry is acceptable?

 

i dont think i will actually do anything about the 0.7v even though it is noticable.

Did you do anything to make the positive indicator come on before the output was above 0.7V? Or did you let it have a dead region too?

 

yes ian , the leds are red but i preferred having a very tiny current to drive a transistor rather than connecting the led directly

Doing this would result in even more dead zone.

 

ok on a separate note , is it possible i make another 2 leds , one for when the voltage is increasing and one for when the voltage is decreasing?

If you have split rails, you can capacitor couple an inverse parallel pair of LEDs between the output and the 1/2 Vcc divider, there will only be LED current while the output voltage is changing.

If all you have is +5V Vcc, you may find it necessary to hang the capacitor/LEDs between a full-bridge output setup.

 

i dont think i will actually do anything about the 0.7v even though it is noticable.

While I was thinking about how to detect pos/neg slope with a couple of comparators, I saw that this circuit would indicate pos/neg signal. It's a little more complex than your current circuit, but this one has no dead zone.

The LM393 operates from the same supplies as your LM324. You can omit R1-R3 if you don't care about the effect of input bias current.

If you tell me what your oscillator frequency range is and how accurate you want slope detection to be, I can offer suggestions.

 

If you tell me what your oscillator frequency range is and how accurate you want slope detection to be, I can offer suggestions.

Guess I can answer this on my own. Since it's a visual indicator, won't be of much use above 10-20 Hz. In that case, you can use a derivative of the positive/negative indicator in post #13. Put a cap after R2 and connect R2 to the input signal; I'd also make R1-3 smaller.