LM3915 Audio Peak project, not working well

[ashley.hughes]

I am using a diagram given to me by Russell McMahon


Values not given on the diagram are R3 = 10K, C2 = 10uF
Pin 11 on the IC2 (LM324 is connected to 0V)

Now it moves on my LED bar graph very slowly if at all when I take the volume slider on my computer from mid to high and if I go low it turns off at the moment 4 bars is high as I get. I am not getting the bounce with the music that I was after.

Any ideas on how to get this working better? or a revised circuit?

[Kiriakos-GR]

How about lowering the C1 to 10uF or lower, you will get more triggering from the mid and high frequencies,
that usually are more active = more and fast motion on the LED's

[ashley.hughes]

No didn't help, but it lead to my answer I changed C1 back to 100uF and changed C2 to a 2.2uF that I had lying around works well.

Now the next thing is I change the R2/R1 Ratio to change the gain of the op-amp so if I put a pot in place of R1 say a 100k that should allow me to change the gain would it not?

[ashley.hughes]

Haven't looked at the adjustment for the gain yet as I have noticed that it never drops below 4 bars every when mutes or disconnected from the computer.

Any ideas, The only thing that I can think of is that I haven't connected pin 11 to -12V and an op-amp normally would then get half the input voltage if there is not dual rail power would it not? But how do I half the input to pin 2?

[ashley.hughes]

Lol now that I have had a conversation with my self on and off this forum I have worked out from a video that I watched this morning about op-amp and how to connect them up I remembered that if not using dual rails you need to half the input voltage

So I put a voltage divider with 12V---10k----out----10k---0V out going to pin 2 of my op-amp!!!

  if only I worked that out earlier.

Still not sure about playing with R1 as the video I watched mentioned not to play with that value as it changes something so back to the video for me. Just thought I might share how I fixed my problem with out zapping something or smoking up my room!

[ashley.hughes]

This is my circuit diagram that I worked out off the breadboard which is working Don't ask me how though lol. The voltage divider I added on pin 3 of the op-amp is a worry as thats not where I meant to put it but it works



Few jacks in there to get the audio in then out to my speakers and a power jack. Think the diagram is write as far as the circuit and any conventions out there for drawing diagrams I like to mix US and EU symbols as well with cap's (EU) and resistors (US)

[ciccio]

There is something wrong in how you wired the input op-amp.
The schematic in your first post is correct, but the one in your last is incorrect.
The feedback resistor must go to the inverting input (pin2), not to the non-inverting input (pin 3).
The way you wired it will transforms the op-amp in one oscillator, or in something instable
Remove voltage divider on pin  3 (actually a non-sense) , follow the original schematic, and it should work.
A good practice is to connect to ground all unused not inverting inputs (pins 5, 10, 12).
Regards

[Alex]

I am using a diagram given to me by Russell McMahon


Values not given on the diagram are R3 = 10K, C2 = 10uF
Pin 11 on the IC2 (LM324 is connected to 0V)

Now it moves on my LED bar graph very slowly if at all when I take the volume slider on my computer from mid to high and if I go low it turns off at the moment 4 bars is high as I get. I am not getting the bounce with the music that I was after.

Any ideas on how to get this working better? or a revised circuit?

To increase the number of LEDs that light up increase R2 or reduce R1. To increase the response time reduce C2 or R3.
C1 is fine.

You can simplify the circuit with an, irrelevant in this case, impact on linearity by replacing the ideal diode op-amp topology with a diode in series with C1 and then to R3. A schottky diode will minimise the minimum signal needed. A potential divider with two resistors can be used to apply a small DC offset if this is a problem.

Make sure you post a picture/video of the project working, blinking lights are a favourite.

[ashley.hughes]

There is something wrong in how you wired the input op-amp.
The schematic in your first post is correct, but the one in your last is incorrect.
The feedback resistor must go to the inverting input (pin2), not to the non-inverting input (pin 3).
The way you wired it will transforms the op-amp in one oscillator, or in something instable
Remove voltage divider on pin  3 (actually a non-sense) , follow the original schematic, and it should work.
A good practice is to connect to ground all unused not inverting inputs (pins 5, 10, 12).
Regards

I see where I went wrong the part in eagle has the pins up side down I will repost the circuit with the correct connections

[ashley.hughes]

Here is the new revised diagram


Tried to add to last post didn't work

edited new version with pots and changed R4 value to allow for brighter lighting

[ciccio]

I've seen the revised circuit. Does it work? I'm curious...

Resistors R5 and R6 on IC2A and R10 and R11 on IC2B  will have two effects:
1) add a positive DC offset to the op-amps outputs
2) change the gain of the op-amp. In fact the original formula does not work anymore.
It is unusual to have an inverting amp (or a precision rectifier) using such a circuit...

I confirm that the original first post's schematic was correct.
If you need a DC offset, just refer to  that schematic and omit R5 and R11 (use a short circuit in place of them). The offset can be obtained bay changing the value of R6 and R12.

I worked for a Company that sold thousands of audio mixers with circuits equivalent to the one in the first post. Don't know if I can find an old board to test, but if I can, I'll check the schematic.
Please refer to the original LM3915 application note http://www.national.com/ds/LM/LM3915.pdf

Regards

[Alex]

Remove those 10kOhm voltage dividers, there is no need for them and they interfere with the intended operation of the circuit.

Use a split power supply as this circuit requires one. Connect the non-inverting inputs of the op-amps to ground, not to negative supply. The same for the 3915 connect the ground pin to ground,not to the negative supply.

In other words, dont deviate from the original schematic you posted as you are clearly shooting in the dark.

If you want to change the response time and the gain as you asked originally change the component values as I suggested. Do not add or remove components and especially power rails.

[ashley.hughes]

I was shooting in the dark a bit it is working quite well at the moment, running for a day or two. I am stopping work on the project and reading into the theory of op-amp's and going to study Daves video on the LM3914 again.

Hopefully that will allow me to get my head around it a bit more, as for using a dual rail power supply. From what I have read of the theory so far it's not necessary, I might be wrong and I believe you more than I do my self at the moment but I will read into it.

[Alex]

as for using a dual rail power supply. From what I have read of the theory so far it's not necessary,

It is required to reverse bias the diodes in the active rectifier.

[Kartika]

good one