Did you measure the DC voltage on pin 5?

 

Hi and thanks for your reply, I tried to put a DC capacitor and same..I think this is a LED problem.

Very strange...

 

What could possibly be wrong with the LEDs?

Try shorting pin5 to 0V and see if the LED goes out.

 

What could possibly be wrong with the LEDs?

Try shorting pin5 to 0V and see if the LED goes out.

Hi and thanks for your help I will try it and I let you know.

 

Hi and thanks for your help I will try it and I let you know.

Yes Led goes Out.

 

Then the problem is the DC voltage on pin 5.
Don‘t forget that the LM3914 is a DC Voltmeter. If you want to measure any audio signal with it, you first must turn it into DC.

 

Then the problem is the DC voltage on pin 5.
Don‘t forget that the LM3914 is a DC Voltmeter. If you want to measure any audio signal with it, you first must turn it into DC.

Thanks but I have a Rectifier before and a DC block in serial with the PIN 5 and the problem is till here, I don't have any DC in PIN5 with the DC volt-meter.

I don't have this problem with another 10 LED barmeter.

 

A small resistor between pin 4 and V- would raise the voltage comparator thresholds slightly, which might shut off LED 1 (at the expense of distorting the meter scale slightly).

If you want to measure any audio signal with it, you first must turn it into DC.

.... and may find a peak detector /leaky integrator circuit useful to prolong signal peaks sufficiently for good indication.
.

 

Thanks but I have a Rectifier before and a DC block in serial with the PIN 5 and the problem is till here, I don't have any DC in PIN5 with the DC volt-meter.

I don't have this problem with another 10 LED barmeter.

See the video link enclosed

https://www.dropbox.com/s/fulqkn3savd6ib7/IMG_0049.MOV?dl=0

 

Try connecting a 100K resistor from pin 5 to ground.

 

Try connecting a 100K resistor from pin 5 to ground.

It cut the signal !

 

The long wires all over the place on a solderless breadboard are antennas that pickup mains hum and other interference.
The LEDs show the amount of interference picked up and the lowest level LED is the most sensitive.
The input wire MUST be connected with a shielded audio cable to reduce some of the interference from the other wires

I made an audio VU meter with two LM3915 ICs in series producing a range of 60dB (1 thousandth to 1) and without a messy breadboard its lowest LED showed no interference..

 

It cut the signal !

Really. Show the rest of the schematic of how the audio signal gets to pin 5

 

The long wires all over the place on a solderless breadboard are antennas that pickup mains hum and other interference.
The LEDs show the amount of interference picked up and the lowest level LED is the most sensitive.
The input wire MUST be connected with a shielded audio cable to reduce some of the interference from the other wires

I made an audio VU meter with two LM3915 ICs in series producing a range of 60dB (1 thousandth to 1) and without a messy breadboard its lowest LED showed no interference..

Yes for sure I agree with you, hop I don't will have this fault when the PCB will be finish.

Thanks to you for you message.

 

Really. Show the rest of the schematic of how the audio signal gets to pin 5

The Audio buffer is not ready yet. I let you know once I am finish.

Thanks to you for your help, but it seems to be know problem.

 

Feeding raw audio to the input of an LM3914 or LM3915 produces a blur of many instantaneous levels as said on the LM3915 datasheet. They recommend and show "peak detector circuits" to feed the pin 5 inputs with positive peak voltages stretched to about 30ms in duration so that you can clearly see the levels. Here is my peak detector circuit:

 

If the LM3914 REFOut pin is 1.25v, Q1 Pin will light at when input pin 5 is ~125mV.
What is connected to the CD4053B pins 3, 5, and 9? Can you post that part of the circuit?

BTW-
The schematic shows the LM3914 connected in BAR mode.

 

If the LM3914 REFOut pin is 1.25v, Q1 Pin will light at when input pin 5 is ~125mV.
What is connected to the CD4053B pins 3, 5, and 9? Can you post that part of the circuit?

BTW-
The schematic shows the LM3914 connected in BAR mode.

Hi eetech00, does the message is for me ? if yes the tests I made is without the 4053, with or without it the problem is the same.

I don't have this problème with another 10LED bloc.

Thanks to you.

 

I don't have this problème with another 10LED bloc.

I still think pin 5 requires a DC path to ground for proper operation. I duplicated the circuit you posted and with pin5 not connected or floating I have the #10 LED always on. Connecting a 1meg resistor from pin5 to ground fixed that problem.

 

Hi eetech00, does the message is for me ?

of course.

if yes the tests I made is without the 4053, with or without it the problem is the same.

I don't have this problème with another 10LED bloc.

Thanks to you.

Show a schematic of your LM3914 test circuit so we're not guessing...