[moved] Query about VU meters (led bargraph)

[Audioguru]

Correct, get rid of the Jfet and use an opamp and two resistors to set its gain instead.
You need to read and learn a little about opamps. Your TL072 has a voltage gain of typically 200,000 times (1/5th of a million) from DC to about 12Hz because it has no negative feedback. With negative feedback setting its gain to 10 times then its frequency response is flat from DC to about 320kHz. Your input resistors are not needed and do nothing.

 

[c_mitra]

Do not forget to make sure that your input signal is at the line level (1V peak) and the reference voltage is also set correctly. I would use a pot to adjust the reference potential so that the input signal is seen full scale...

 

[Enzy]

Do not forget to make sure that your input signal is at the line level (1V peak) and the reference voltage is also set correctly. I would use a pot to adjust the reference potential so that the input signal is seen full scale...

input voltage from my laptop must be atleast 1v peak? what are you referring to as reference voltage?

- - - Updated - - -

Correct, get rid of the Jfet and use an opamp and two resistors to set its gain instead.
You need to read and learn a little about opamps. Your TL072 has a voltage gain of typically 200,000 times (1/5th of a million) from DC to about 12Hz because it has no negative feedback. With negative feedback setting its gain to 10 times then its frequency response is flat from DC to about 320kHz. Your input resistors are not needed and do nothing.

its hard for me to communicate with you guys to be truthful but im trying to learn and work as I go.

I did another drawing.

- - - Updated - - -

using a 5k pot in the schematic.

 

[c_mitra]

input voltage from my laptop must be atleast 1v peak? what are you referring to as reference voltage?

If it is meant for the headphone, it may be even less. If it is designed as hi-fi, it will reserve lots of headroom and even the headphone will sound low (volume) with the highest volume. On your desktop, you will have line out, line in and headphone connectors. The bar graph driver is designed to show full scale when the reference voltage and the input voltages are equal (not true for log drivers) and they will have some reference input or reference set pin (I have used them long long ago). This reference voltage I am speaking about is on the bar graph driver.

- - - Updated - - -

using a 5k pot in the schematic.

How about adding some gain? Say something like 1 to 10 (adjustable)? You have just made a good buffer!

 

[Enzy]

how is that done?

- - - Updated - - -

and you mean to add to the drawing I did in post #23?

 

[Audioguru]

Your LM3915 has a simple peak detector circuit feeding it. The errors of the simple peak detector are covered up when the reference voltage at pin 7 of the LM3915 is set to about 10V. Then the peaks of your input signal must also reach 10V to light the 10th LED. The opamp we just finished amplifies the input signal so that its output will have peaks at 11V when its input peaks are at 1V.

Use a 20k volume control because a 5k might overload the signal source. I corrected your opamp circuit:

 

[Enzy]

Ok I added the other stuff, check the output cap.

 

[Audioguru]

I corrected the errors:

 

[Enzy]

I have been searching how to calculate the capacitor on the output but i cannot find it, which load resistance would that be? the 14k impedance that we saw on the specs on ebay for the amp?

 

[Audioguru]

I forgot. The opamp with the gain of 11 tiimes feeds the input of the Chinese amp that has an input resistance of about 14k ohms. Then for a -3dB response at about 7Hz the capacitor must be 1 divided by (2 x pi x 14k x 7Hz)= 1.6uF, use 1.5uF or 2.2uF.

 

[Enzy]

also I am using 2 of the chinese amplifiers so I would actually feed the same signal to both amplifiers so would that make the input 28k ohms in that case which would require a different capacitor?

 

[Audioguru]

The 14k ohms inputs of the two amplifiers will be in parallel making 7k, not in series. Then the capacitor needs to be double the 1.6uF which is close to 3.3uF.
Why do you have two amplifiers with the same input signal?

 

[Enzy]

because the amplifier can only go down to 4 ohms and I have 4 8 ohm speakers so I pair them up to each output of the 2 amplifier boards, my aim was to build my own class D amp but I havent made any progress I need help doing that.

So I had to settle for buying 2 of the same Chinese amplifiers, also remember I will be playing bass with it so I dont need stereo so I would just parallel the input signal.

Also if I would be using 7kohm input signal that means the pot on the opamp would be 10k?

 

[Audioguru]

because the amplifier can only go down to 4 ohms and I have 4 8 ohm speakers so I pair them up to each output of the 2 amplifier boards

Will your speakers survive a few hundred watts each?

I would be using 7kohm input signal that means the pot on the opamp would be 10k?

The volume control is on the INPUT of the opamp. The amplifiers are at the OUTPUT of the opamp and do not affect the volume control. The volume control can be 10k or 20k.

 

[Enzy]

Will your speakers survive a few hundred watts each?

yes 4 speakers a pair of them are 600rms and the other pair is 1500rms

- - - Updated - - -

each speaker has that rating so it can manage the power.

So I will build a pcb for the opamp and then I will be able to attack the Lm3915 circuit to the output of the amplifier.

just to verify the schematic again for the Led's also I would use a bc557 for Q1.

 

[Audioguru]

Your peak detector and LM3915 circuit will not work:
1) It has no power supply voltage. 12VDC would be good.
2) Pin 4 of the LM3915 must be connected to 0V.
3) R2 should be 1M not 1m (0.001 ohms).
4) The peak output from the amplifier powered from an 80V center-tapped transformer is close to 56V but the peak detector circuit is missing an attenuator to cut the destructive 56V down to its normal maximum input of 10V.

 

[Enzy]

Ok so is it better for me to connect the input of the peak detector to the signal in of the amplifier?

I added made a change to the schematic I dont know if that would work since the voltage wont always be at that peak of 56v.

 

[Audioguru]

A meter shows levels. If it is at the output of an amplifier then its display depends on how loud is the sound, it will be low or show nothing when the volume control is turned down.
If the meter is at the input to the volume control then it will show the levels of the input signal.
Your simple peak detector needs a signal with a peak level of 10V so the opamp with a gain of 11 should amplify the input signal then it can drive the peak detector and LM3915 and the volume control that feeds the amplifier.

 

[Enzy]

so in other words I can remove the attenuator from the circuit and connect the peak detector to the output of the opamp circuit which would be the input of the amplifiers since it should be receiving 11v?

 

[Audioguru]

Yes, remove the attenuator.
Yes, connect the output of the opamp circuit to the input of the peak detector circuit AND to the volume control that feeds the inputs of the amplifiers. The volume control should be 2k since it is loaded with 7k but your signal source might be overloaded driving 2k so use a 5k volume control. Re-calculate the value of the output capacitor at the output of the opamp circuit since it will be driving the 100k input of the peak detector parallel with the 5k volume control (4.7uF electrolytic with its (-) wire to the volume control).

Now that the volume control is moved then we do not know if the opamp gain is too much causing clipping or is not enough so maybe the opamp should have a gain control or its own volume control as a trimmer pot.