LM3915 LED VU meter problem - pictures included

[ericgibbs]

PNP E = 1.47v
PNP B = 0.8v
PNP C = 0v

Those voltages are not correct.

The emitter voltage is too high thats why you have 1.08v or so on pin 5.

What is the type of the transistor.??

 

[transistor495]

It can be, I thought this guy has done some decent measurements regarding peak voltage measurements.
Connect your Ipod to a Stereo

Anyway I don't see any problem to this circuit to not function with an iPod input. I really suspect OP's transistor,diode and mainly the breadboard. He has to do some flexible experiments his own like re-building on another board etc.

 

[Gaichuke]

2N3906 PNP transistor, as said in the original instruction part list.

 

[audioguru]

The article "Connect Your ipod To A Stereo" is wrong.
You never match impedances anymore. Old vacuum tube amplifiers matched the output transformer to the speaker which caused the speaker to have almost no damping of its resonances. Their output power was low so they matched impedances for the most output power.

Today you almost always have a very low inpedance (the output of an opamp) driving a much higher inpedance for low voltage loss and very low distortion. The output impedance of most power amplifiers is 0.04 ohms or less to drive an 8 ohm speaker. The damping is excellent.
If you match impedances then you throw away half the signal's level and increase the distortion.

 

[Gaichuke]

I tried to move the components to opposite end of the breadboard, and all the voltage values were the same. I don't think the breadboard is to blame here.

Maybe I'll have to get a few 2N3906 from the store. I guess it's possible that I've burned them somehow during testing the circuit.

 

[ericgibbs]

I tried to move the components to opposite end of the breadboard, and all the voltage values were the same. I don't think the breadboard is to blame here.

Maybe I'll have to get a few 2N3906 from the store. I guess it's possible that I've burned them somehow during testing the circuit.

hi,
I get the lowest offset with a BC214 at 0.29v

If the output of an MP3 is 300mVppk you need a preamp.
I would consider using a dual OPA, something like a CA3240.
Use the 1st half as a preamplifier and the 2nd half as a full wave rectifier, in that we we can remove the crude pnp rectifier.

 

[audioguru]

I posted my single opamp preamp and peak detector circuit in post #57 here: https://www.electro-tech-online.com...-problem-pictures-included.106176/#post868689

 

[ericgibbs]

I posted my single opamp preamp and peak detector circuit in post #57 here: https://www.electro-tech-online.com...-problem-pictures-included.106176/#post868689

hi agu,
The post #57 showing the peak detector has a gain of 1, no pre-amp shown, unless I am missing something.

 

[audioguru]

hi agu,
The post #57 showing the peak detector has a gain of 1, no pre-amp shown, unless I am missing something.

I mentioned in the text "Reduce the value of R9 a little or increase the value of R8 for more gain".
The opamp inverts so a negative input is cancelled by a positive output, which prevents the input pin from going more negative than ground.
So you don't want the gain to be so high that the circuit clips its output as high as possible that will cause the input pin of the opamp to go below ground which is bad for it.

My circuit used another opamp as a mic preamp and when it is over-driven then the low gain of the peak detector opamp prevents its input pin from going below ground.

 

[ericgibbs]

hi, agu,
We have sorted out the basic LM3915 project board problems.

I have advised the OP that a pre-amp and a quality precision rectifier is required.
Not having a MP3 player to test, I am working with some web notes which state that a MP3 output is approx 300mVrms into 100~200 ohms.
The LM3915 on my test rig requires approx 8~10Vppk to drive the full chain of LED's when using the existing crude rectifier.

 

[Gaichuke]

Hoo boy... I'm so confused now. Information overload!

I agree that the mp3 player output will be low, and I've seen similar 300mV references elsewhere. So, I guess that preamp will be necessary. I thought that the 5K pot was for that problem, but I guess the output voltage is way too low for that to help.

So I'll still need a few parts from the store, couple of opamps and some other parts as well. But I have absolutely no idea how to get everything I need without parts list from a ready circuit. Audioguru, those circuits that you posted mention electret mic, +6 to +9V, I will have to use 12V for this one, so I guess the resistors and such won't do as in the pics.

The AG circuits look good, I think I can build them, but how should they be modified for my purposes?

 

[audioguru]

If you use my single opamp circuit then you must use a single-supply opamp like the MC33171 that I used or an MC34071. Half of an LM358 can be used but it is poor at high frequencies.
Set the LM3915 like it was before with an input of +1.25V to light the 10th LED. Then the gain of the opamp needs to be only 1.25V/.4V= 3.125.
Increase R8 to 300k or 330k.

If you use a 12V supply then the LM3915 might get very hot. You can add a power resistor in series with the voltage to the LEDs to share the heat.

The 5k pot isn't needed with my opamp circuit. It was used to turn off the first LED when there was no signal in the Mickey Mouse transistor circuit.

 

[Gaichuke]

So the circuit in this picture would be one correct way to do it? And this gain of 3.125 should work with low output voltage sources like mp3 players?

I plan to use 2 LEDS in series per output to reduce the load on the chip.

 

[audioguru]

It looks good. Try it.

 

[TechnoGilles]

According to what I saw on the Web, the output power of most MP3 player is in the range of 5-40mW per channel. For instance, the IPod has an output power of 30mW per channel and Apple supply them with 32Ω earphones so we can deduce a 1Vrms at the output or about 3Vppk.

If you can find the output power of your MP3 player and the impedance of the earphones supplied with it, you might be enough to find out the output voltage characteristic.

But no matter what, I'd say that a pre-amp with an ajustable gain between 3 and 6 would most probably do the job, right ?

 

[audioguru]

You don't need adjustable gain when your are fed from the headphones outputs of an ipod that has its own volume control.
But you might want to turn down the sound level and turn up the gain of the display so it is near max rthen an adjustable gain control is needed.
The 330k resistor could be a 1M pot connected as a variable resistor. Keep its wires short.

 

[TechnoGilles]

What I had in mind with an ajustable gain is to be able to accomodate the circuit to various MP3 players with different output powers (hence different output voltage profiles).

I was also wondering if the OP wants to listen to his MP3 player at the same time he uses his circuit. If so, he might want to adjust the MP3 volume to his liking and then boost the gain of his circuit so it is still full scale.

 

[Gaichuke]

Good morning everyone,

I think that the 1M pot might be a good idea just in case, or to have several different resistors to test what the gain should be. The mp3 output will be at fixed level(max volume), so there isn't room for adjustment there.

I will also need to hear the music at the same time from the mp3 source. I was planning to use just a simple 2€ headphone splitter, one end going to the amplifier and the other to the VU meter circuit. If this is somehow a bad idea, I could use suggestions for proper solution. Could there be signal attenuation?

 

[ericgibbs]

Good morning everyone,

I think that the 1M pot might be a good idea just in case, or to have several different resistors to test what the gain should be. The mp3 output will be at fixed level(max volume), so there isn't room for adjustment there.

I will also need to hear the music at the same time from the mp3 source. I was planning to use just a simple 2€ headphone splitter, one end going to the amplifier and the other to the VU meter circuit. If this is somehow a bad idea, I could use suggestions for proper solution. Could there be signal attenuation?

Morning Gaichuke,
I ran a LTspice simulation on that last circuit you posted, its got a 'strange' output response.

I will run 'agu's as a comparison later today.

 

[transistor495]

Peak detector in both the circuits looks like twins.