VU meter with opamp power help!

Thread Starter

PegasusIII

Joined Nov 12, 2016
18
Hi there,

I've been trying to work out the correct way to wire up the following circuit (a 19LED vu circuit that uses LM3915/3916 and a TL072 op amp on the input):
VU_Meter_2.gif
In the past I have only really worked with DC power that is positive and ground, and I am a little confused about the 12+/-... I understand that opamps like the TL072 require split power like this, and such circuits need both the +12 -12 as well as 0v as a reference?

I have a power supply that can give + and - 12vdc, and i then connected ground to 0v and also connect the signal ground to 0v, but the circuit does not perform as it should.. Is how I am wiring it like this correct? Or am i completely wrong? I have tried to google about -/+ vcc supplies like this without much luck..

I hope this makes sense!
Thank you
 

GopherT

Joined Nov 23, 2012
8,009
Hi there,

I've been trying to work out the correct way to wire up the following circuit (a 19LED vu circuit that uses LM3915/3916 and a TL072 op amp on the input):
View attachment 116475
In the past I have only really worked with DC power that is positive and ground, and I am a little confused about the 12+/-... I understand that opamps like the TL072 require split power like this, and such circuits need both the +12 -12 as well as 0v as a reference?

I have a power supply that can give + and - 12vdc, and i then connected ground to 0v and also connect the signal ground to 0v, but the circuit does not perform as it should.. Is how I am wiring it like this correct? Or am i completely wrong? I have tried to google about -/+ vcc supplies like this without much luck..

I hope this makes sense!
Thank you
What make & model power supply do you have?

Post a photo of your setup (with at least one of the power supply)
 

Thread Starter

PegasusIII

Joined Nov 12, 2016
18
What make & model power supply do you have?

Post a photo of your setup (with at least one of the power supply)
Thank you GopherT for replying!

I had this 12v 3a dc powersupply lying around that has markings for +12 and -12 so I thought it would work for testing purposes..
I will try get an image of actual set up when I get home.

also, here is the pcb I made for this project..
vu pcb.PNG
There could be something wrong with that too but I just wanted to check that I had the power setup correct as it confuses me a bit!

Even if someone can share a link with some explanation of -/+ supplies?
 

Thread Starter

PegasusIII

Joined Nov 12, 2016
18
(blue lines are jumper bridge wires, all the LED anodes are connected together then the end connected to LED+ pad, and the LED cathodes conected via header pins to the IC)
 

GopherT

Joined Nov 23, 2012
8,009
(blue lines are jumper bridge wires, all the LED anodes are connected together then the end connected to LED+ pad, and the LED cathodes conected via header pins to the IC)
I see you have an "input" and 0V connected together. That should be "input common", right? Then where is your signal wire for the input? You need two wires from your audio source. And they shouldn't be connected by a copper trace on the pcb.

Also, use your volt meter on DC to measure voltage across V- to V+. Is it 12V or 24V?

Be careful with this power supply. The mains (wall) power is not enclosed in a chassis. The screws are accessible and could be bumped (this is meant to be enclosed into a bigger chassis).
 

AnalogKid

Joined Aug 1, 2013
10,987
1. You don't indicate which IC is the 3915 and which is the 3916. It matters.
2. One of those two parts drives a linear scale, not logarithmic. You show a log scale for both parts.
3. There are no component values, so there is no way to evaluate it the circuit will do what you want.
4. It looks like you are amplifying the input signal to drive the low end display, then attenuating that to drive the high end display. Not the best approach for accuracy.

Please post a link to where you got the schematic.

ak
 

Thread Starter

PegasusIII

Joined Nov 12, 2016
18
1. You don't indicate which IC is the 3915 and which is the 3916. It matters.
2. One of those two parts drives a linear scale, not logarithmic. You show a log scale for both parts.
3. There are no component values, so there is no way to evaluate it the circuit will do what you want.
4. It looks like you are amplifying the input signal to drive the low end display, then attenuating that to drive the high end display. Not the best approach for accuracy.

Please post a link to where you got the schematic.

ak
Thank you both for your replies.
I got the schematic from here:
http://users.otenet.gr/~athsam/vu_meter_2.htm

Apologies, i forgot the schem didn't have values on it.

I just wanted a Vu display with 19 (or preferably 20) LEDS. I looked around and this seemed to suit my needs the best.
Obviously I have minimal understandings of these things!
 

Thread Starter

PegasusIII

Joined Nov 12, 2016
18
I see you have an "input" and 0V connected together. That should be "input common", right? Then where is your signal wire for the input? You need two wires from your audio source. And they shouldn't be connected by a copper trace on the pcb.

Also, use your volt meter on DC to measure voltage across V- to V+. Is it 12V or 24V?

Be careful with this power supply. The mains (wall) power is not enclosed in a chassis. The screws are accessible and could be bumped (this is meant to be enclosed into a bigger chassis).
Yes I am very careful when I use this supply. I will check that to see when I am back at my desk.

Hooking up the power and input wires is what I am confused about!
at the moment I have the main (center of rca plug) audio signal going to the 'input' and the sleve going to 0V. But now that you say it this does not make sense at all! I was just following the schematic I found.. Can you please advise where I should be putting the audio signal sleeve/ground?
 

GopherT

Joined Nov 23, 2012
8,009
The sleeve of your audio source should be connected to the 0V and the ground from your power supply shoud go to the 0V.

You need to get a utility knife/razor blade and cut the wire from input to 0V. Use a magnifying glass to see that all copper shavings are gone and no connection.

Then put the signal tip from RCA Jack to the input.
Put -12V to -V of power supply and +12 to +V of power supply.
 

Thread Starter

PegasusIII

Joined Nov 12, 2016
18
The sleeve of your audio source should be connected to the 0V and the ground from your power supply shoud go to the 0V.

You need to get a utility knife/razor blade and cut the wire from input to 0V. Use a magnifying glass to see that all copper shavings are gone and no connection.

Then put the signal tip from RCA Jack to the input.
Put -12V to -V of power supply and +12 to +V of power supply.
Ok fantastic!
Thank you. This is how I had it hooked up, but obviously that copper connection between 0V and input is an issue (I hope this is the only issue!) Hmmm, I wonder why this was put in the original schematic? I will check when home tonight.
Thanks again for your help!
 

GopherT

Joined Nov 23, 2012
8,009
Ok fantastic!
Thank you. This is how I had it hooked up, but obviously that copper connection between 0V and input is an issue (I hope this is the only issue!) Hmmm, I wonder why this was put in the original schematic? I will check when home tonight.
Thanks again for your help!
That is how shielded cable connectors are drawn. The sleeve to ground and tip to signal.

They should have been a bit more careful hitting the tip and sleeve circles.
 

dannyf

Joined Sep 13, 2015
2,197
got the schematic from here:
The design is a little bit interesting in that 1 it uses a rectifier stage, and 2 it amplifies the signal and then attenuate it.

The Chip can handle ac signal natively. So unless you want to do averaging or peak, no rectification is needed.

On amplification, just use an opamp to amplify the input for the lower LEDs and feed the input signal directly to the higher LEDs. The amount of gain needed is dependent on the chips used.

Also, config the opamps as ac coupled so you don't have to deal with the dual supply issue.
 

Thread Starter

PegasusIII

Joined Nov 12, 2016
18
The design is a little bit interesting in that 1 it uses a rectifier stage, and 2 it amplifies the signal and then attenuate it.

The Chip can handle ac signal natively. So unless you want to do averaging or peak, no rectification is needed.

On amplification, just use an opamp to amplify the input for the lower LEDs and feed the input signal directly to the higher LEDs. The amount of gain needed is dependent on the chips used.

Also, config the opamps as ac coupled so you don't have to deal with the dual supply issue.
Thanks for that!
When you say "config the opamp as ac coupled" what would this involve? Where would the coupling cap go and would the two AC's just go straight to +12 and -12?
 

GopherT

Joined Nov 23, 2012
8,009
Thanks for that!
When you say "config the opamp as ac coupled" what would this involve? Where would the coupling cap go and would the two AC's just go straight to +12 and -12?

No, no, no!

The A/C Danny is talking about is the wave of the AUDIO signal! I already told you your power supply was sketchy for a beginner. Don't make it worse by connecting that improperly.

First thing you need to do is measure voltage from V- to v+ (should be 24V.
Then measure with black lead on gound (common) to read lead on V- (-12V) and then move only read lead to +V to read (+12V).

If you get significantly different values, there is something wrong and that may not be a split (dual) power supply. That is, your ground screw MAY just be a chassis ground for safety. We don't know yet, please report back.
 

AnalogKid

Joined Aug 1, 2013
10,987
Stick with Gopher the Yinzer. Danny's advice is incorrect. While the 39xx parts do have a form of signal rectification (actually asymmetrical clipping) built in, it works *only* if the incoming signal is centered about GND *and* is DC coupled. The external rectifier circuit you have is a) more accurate; and b) gives you control over the dynamics of the display with the R5-C6 time constant. However, the output of IC1a might distort and clip on loud signals. It has a gain of 30dB when TR1 is centered.

If you find that the display is inaccurate with loud signals, one solution is to rearrange the two opamp circuits. Rectify the incoming audio first and rive the right side display directly. Then amplify that signal by 30 dB to drive the left side display. It doesn't matter if the signal clips in that amp because at that level the left side display already is peaked.

I was working from memory, and forgot that the 3916 has a non-linear dB scale.

ak
 

AnalogKid

Joined Aug 1, 2013
10,987
LOOP:
Single supply opamps will not work in this application.
Single supply opamps will not work in this application.
Single supply opamps will not work in this application.
GOTO LOOP

ak
 

Thread Starter

PegasusIII

Joined Nov 12, 2016
18
No, no, no!

The A/C Danny is talking about is the wave of the AUDIO signal! I already told you your power supply was sketchy for a beginner. Don't make it worse by connecting that improperly.

First thing you need to do is measure voltage from V- to v+ (should be 24V.
Then measure with black lead on gound (common) to read lead on V- (-12V) and then move only read lead to +V to read (+12V).

If you get significantly different values, there is something wrong and that may not be a split (dual) power supply. That is, your ground screw MAY just be a chassis ground for safety. We don't know yet, please report back.
Thanks you all for your input and for clarifying things for me.

I just had a chance to put my multi meter on the power supply unit and it only reads ~12v across the -v and +v terminals, and a negligible mv rating between either point and groud. So I guess it is just a casing ground. Should I instead look at something like this to supply power to this circuit? (looks like a simple rectifier then splits the dc with a lm7912 and 7812)
 
Top