lm3915operational-amplifier
simulate this circuit – Schematic created using CircuitLab
I have made a LED VU Meter using LM3915 and added a low pass filter too. Then I've tried a basic op amp circuit using LM324 (inverting and non inverting) to amplify the LM3915's input, but it doesn't work. It only lights up the first led, it doesn't work like a VU meter should. Can anybody show me how to use an op amp like the LM324 with the LM3915?
My LM3915:
The two LM324 circuits I tried:
First, you should clarify what your LED is. Single LEDs typically have forward voltages around 1-3V, so 12V LED makes me suspect it is either a part designed for automotive use, or a string of LEDs in series.
If it is an automotive part then it's quite possible it's designed for high current operation and the LM324 can not supply the required current to light the LED. The LM324 might burn itself up in the attempt.
If it is a string of low-power LEDs then you still have a problem because you need current limiting. If the component tolerance variation causes the string of LEDs to have a forward voltage of greater than 12V, then they won't light up; if their tolerance causes the forward voltage to be less than 12V they (or the opamp) might burn up without current limiting.
There is a reason non-inverting 1st order high-pass filters are sometimes used and this is to preserve the dc level in the circuit. It's horses for courses - some designs will want to preserve the DC level, some won't care and some won't want to. It's not a black and white sort of thing.
When it comes to a 2nd order high-pass filter the non-inverting configuration (Sallen Key type) will usually be preferred and this doesn't preserve the dc level: -
Again horses for courses.
Best Answer
You haven't indicated what you did for the low pass filter, but what I've found with these devices when I used them 30+ years ago is that you really can't see the higher LEDs turning on with a straight signal going into them, since they are on for such a short time, they're very dim. For a proper VU meter, you need a peak hold circuit - mechanical VU meters had a fairly quick rise under the moving coil's force but were well damped so returned fairly slowly under the hairspring. The LM3915 datasheet shows a peak detector based around an op-amp, the size of that output cap will determine the response.
The gain of the op-amp circuit can be adjusted to get a reasonable scaling, but you can also reduce the value of the reference voltage on Rhi (pin 6) to increase the sensitivity at the expense of accuracy. The datasheet shows examples of that too.