LM3915 LED Brightness Adjustment

[Depson]

Hello.I would like to do VU meter tower. I found a LM3915 diagram on the internet. I changed a few things but I would like to add LED brightness adjustment and I don't know how to do it (actually I did some brightness adjustment which I found in the datasheet but I got only a 6-15mA scale).
After all I would like to get bigger scale (1-20mA) but I don't have any idea how to do this.
Probably I don't have required parts and will need to order them (I want to do components list and order it all at once, not one by one). These two potentiometers on the diagram are a dual potentiometer but I didn't know how to mark it. Sorry for my English, I hope you will understand me and thank you for any help

[MarkF]

Read the datasheet....

   

[themadhippy]

Read the datasheet....

They did

actually I did some brightness adjustment which I found in the datasheet but I got only a 6-15mA scale

But for a beginner  the tables of numbers and hieroglyphic symbols  can be intimidating ,and its easy to head straight to the pretty pictures  skimming  past things like

A feature not completely illustrated by the block diagram is the LED brightness control.The current drawn out of the reference voltage pin (pin7) determines LED current. Approximately 10 times this current will be drawn through each lighted LED,

which ties in with the graph (fig9),so you need to get around 2-2.5ma out of pin 7 .

[schmitt trigger]

To get a wide dimming range, the best would be to PWM the LED’s V+ supply.

That will also reduce the internal power dissipation of the LM3915 at the highest currents. And more so if you plan in using the bargraph mode.

[Zero999]

To get a wide dimming range, the best would be to PWM the LED’s V+ supply.

That will also reduce the internal power dissipation of the LM3915 at the highest currents. And more so if you plan in using the bargraph mode.

How will PWM reduce the internal power dissipation of the LM3915? Higher currents will always require a higher duty cycle and therefore more power dissipation.

[schmitt trigger]

I don't feel like arguing today. I have been advised that I am losing my job.

So please simply ignore my comment. I have actually modified my reply.

[Depson]

I tried this diagram above, I even changed resistors and potentiometer values but the best I got was 2.6-21mA scale(values on picture below).
Also I need something to adjust two LM3915 at the same time and get the same value. I also tried second diagram(attached below) but I didn't have enough parts and I used TL081 instead of LM386. There was 4.6-18mA, but maybe that's because I didn't built two the same configurations?

[Zero999]

I don't feel like arguing today. I have been advised that I am losing my job.

So please simply ignore my comment. I have actually modified my reply.

Sorry to hear that.

I wasn't looking for an argument. I did presume you were mistaken but was open to the possibility I had missed something. A common misconception is PWM is more efficient, but this is only true if  there's an inductor to store energy, rather than burn it as heat, as in the case of a resistor or semiconductor biased into the linear region. The good thing about PWM is it will enable a wider dimming range, without actually altering the forward current, which can sometimes affect the wavelength, so it's a nice idea.

I tried this diagram above, I even changed resistors and potentiometer values but the best I got was 2.6-21mA scale(values on picture below).
Also I need something to adjust two LM3915 at the same time and get the same value. I also tried second diagram(attached below) but I didn't have enough parts and I used TL081 instead of LM386. There was 4.6-18mA, but maybe that's because I didn't built two the same configurations?

Connect the +V side of the LEDs to the output of a 555 PWM circuit driving a PNP transistor.

[Depson]

Should I connect it like this(picture attached)?
Is this diagram only for dimming?
Can I use 1N4001 for D1 and D2?
Could you give me examples of Tr1 which can I use ?
What is meant by "Load"?

[Zero999]

Should I connect it like this(picture attached)?
Is this diagram only for dimming?
Can I use 1N4001 for D1 and D2?

Yes to all three questions, although there's no need for such a high current diode. The 1N4148 will do.

Could you give me examples of Tr1 which can I use ?

Basically anything which can carry sufficient current for the LEDs. The BC327 will do.

What is meant by "Load"?

The device the dimmer circuit is controlling which is the LEDs in this case.

[Depson]

I ordered parts few days ago and I got it today. I tried to do this on two breadboards before I will do this on prototype board but it worked poorly(One time it worked a little, I touched something and it stopped working) Can it be because of low quality breadboards and a lot of cables? Should this diagram work when I will do it correctly on prototype board?

[Zero999]

I ordered parts few days ago and I got it today. I tried to do this on two breadboards before I will do this on prototype board but it worked poorly(One time it worked a little, I touched something and it stopped working) Can it be because of low quality breadboards and a lot of cables? Should this diagram work when I will do it correctly on prototype board?

Breadboard is notoriously bad for connections, so that's quite likely.

Have you tried the PWM circuit separately, with a resistor and LED on the output?

[Depson]

I tried PWM circuit separately with one LED and it worked. When I was trying to do this with one LM3915 it was very hard because when I only touched something it stopped working,but I did this one time for a minute and it worked. It's impossible to do this with two LM's on breadboards.It is probably because of bad quality breadboards, isn't it?

[Zero999]

Yes, it does sound like the breadboard is to blame. You could try building the circuits on stripboard instead. At first build each circuit separately and test it, before connecting them together.

[Depson]

I had a lot of problems with 50k and 100k dual potentiometers, they didn't work. I connected it in a lot of ways but still nothing. Now I connected them like on the picture bellow (I did it also on the breadboard to show connections). When I am connecting it to Fiio K3 one channel is working, in second one every diode produce light all the time. I also tried on my tablet and there one channel work a little (few diodes produce light all the time, few move) but second channel didn't work (all diodes produce light all the time)

[Zero999]

The circuit I posted is not designed to work with two potentiometers, but one. It works by using one half of the potentiometer to charge the capacitor and the other half to discharge it.

[Depson]

But I didn't use two potentiometers for dimming. I used it for sensitivity adjustment because I would like to have more precise adjustment and there's something wrong there. How should I connect these potentiometers?

[Zero999]

But I didn't use two potentiometers for dimming. I used it for sensitivity adjustment because I would like to have more precise adjustment and there's something wrong there. How should I connect these potentiometers?

It's not clear how they're connected to the rest of the schematic. The potentiometers need to be connected like a volume control, as per the LM386 data sheet.
http://www.ti.com/lit/ds/symlink/lm386.pdf

Ideally you want a potentiometer with a tapered track, as it's an audio application.

[Depson]

I finished a smaller version and it works, but I have one more question. I did everything like on a diagram below but I added a rotary switch which is used to connect different value capacitors (In place of C3 - marked below) I used 100nF, 1uF, 10uF and 33 uF capacitors but everything works fine with only two of them (probably 10uF and 33uF).When I use other capacitors diodes produce light without any signal. Does anyone have any idea why?

[Zero999]

The amplifier is oscillating with the capacitive load?

C3 is far too higher value and should have a series resistor. The values given on the data sheet are 0.05µF in series with 10R, but 47nF will do. It's a Zobel network and can be omitted if the amplifier is just driving the VU meter, without a speaker.
https://www.ti.com/lit/ds/symlink/lm386.pdf?HQS=TI-null-null-sf-df-ds-null-wwe

[Depson]

I can't understand what you mean. Could you rewrite what should I do to fix that? Thank you.

[Zero999]

I can't understand what you mean. Could you rewrite what should I do to fix that? Thank you.

Try removing C3.