Need help with 3916 super project

[supes]

So I have big hopes for my LED project. I have been stocking up on high quality LEDs and rechargable  batteries just for this project.

Anyway here is the problem. I am using LM3916 drivers, and alone I can get it to do what I want with the signal (from sound card) connected directly to the input of the IC. I used this as reference.

Now, I wanted to make this wireless so I got a couple electret mics and tried using this as a guide with my 3916 and hoped it would all work out.
http://www.electronics-lab.com/projects/motor_light/009/index.html

However, I just have all LEDs on all the time in bar mode and the last LED on in dot mode. I dont think the microphone part is working or else there is something wrong somewhere. I followed the schematic exactly (minus the 1W resistor) and have been going crazy going over looking for any mistakes. I tried changing the POT to all values and nothing changes. It is also worthy to note that I used a 358N IC for the pre-amp and my signal into pin 5 of the 3916 is 6.9V.

I am asking for some help now since I am stumped.

 

[audioguru2]

Hi Supes,
Welcome to our forum. ;D
The LM3915 and LM3916 ICs are DC voltmeters. In your circuit the input voltage for all 10 LEDs to be lighted is only 1.25V. Your input voltage is 6.9V (DC?) when it should be 0V without signals.
An LM358 opamp should work in my circuit but it has reduced response above about 5kHz.

The 2nd opamp in my circuit has its non-inverting input grounded so it should cause the transistors to have an output of 0V without a signal.

View attachment 39525

 

[supes]

So here are a few points that will hopefully help you understand my situation better.

1. I am using all 3904's
2. I am using the 358N and I know it will be more of a 'beat' meter than a frequency type since the 358 is limited to about 4kHz.
3. At pin 5 of the 3916 I get 6.9DC and 14.x VAC which seems like overkill and is probably the reason all LEDs are on. Therefore I think the problem is in my pre-amp stage.
4. I attached a picture of my electret microphone as I am not sure if it is one that would work with this circuit.

Circuit with all LEDs on

Microphone (using one on left)

I will continue to work and update the situation. Thank you for the help and the warm welcome audioguru!

 

[audioguru2]

At pin 5 of the 3916 I get 6.9DC and 14.x VAC which seems like overkill and is probably the reason all LEDs are on. Therefore I think the problem is in my pre-amp stage.

Your driver opamp circuit has an output level way too high!
The signal at pin 5 of the LM3915 is DC since Q2 is a peak rectifier.
Maybe your preamp is oscillating or picking up mains hum. My microphone's cable is only about 4cm long but is a shielded cable so it doesn't pickup interference.

I attached a picture of my electret microphone as I am not sure if it is one that would work with this circuit.

It looks good.

Circuit with all LEDs on

I don't use a breadboard because the messy wiring picks up mains hum and other interference. I use Veroboard (stripboard) to make neat and tidy circuits with very short wires and soldered connections.

 

[supes]

Okay I found 2 resistors that were sorted together. I am getting somewhere now. I am getting .1VDC at pin 5 of the 3916 and .24V at pin 6. I turned the POT down and I have only a few LEDs light up now but now still move. If I can get the pin 5 voltage up from the microphone (dont even know if the microphone is working yet) I may get the whole thing working as it should.

I have .49VDC at pin 7 of the 358N. What am I doing wrong? Why is the signal only .1V?

 

[audioguru2]

Okay I found 2 resistors that were shorted together. I am getting somewhere now.

Another problem with using a messy breadboard. It is good that you found it. Let us hope that nothing is damaged.

I am getting .1VDC at pin 5 of the 3916

That is close to 0V and maybe just caused by interference picked up by your long wires.

... and .24V at pin 6.

The LM3916 has a 12k ohm ladder at pin 6 while my LM3915 has a 28k ohm ladder. Therefore reduce the value of R15 to 100k ohms to compensate. Then the voltage at pin 6 will be higher.

I turned the POT down and I have only a few LEDs light up now but now still move.

The pot in my circuit is a brightness control and it doesn't affect how many LEDs are lighted.

If I can get the pin 5 voltage up from the microphone (dont even know if the microphone is working yet) I may get the whole thing working as it should.

I have .49VDC at pin 7 of the 358N. What am I doing wrong? Why is the signal only .1V?

The voltage at pin 7 of your LM358 and the voltage of the signal are about correct without an input.
I guess your microphone or preamp doesn't work.

 

[supes]

Yea I was afraid of the microphone not working. I can hope that it might be the pre-amp.

Anway I did change the resistor you mentioned and I get a bout .18VDC at pin 5 and .49VDC now.

Would you be able to recommend another pre-amp that would work with an electret microphone and give a good level to drive the 3916? I have 4 electret microphones that all look like the one i posted. All are new.

I also have the
2904
386
1458
324
just incase I will need to use them

 

[audioguru2]

The mic preamp circuit in my project should work fine with your half of an LM358 dual opamp, it just won't have high audio frequencies.
Measure the DC voltage at pin 1, it should be 2.5V if you used a 5V regulator.
The case of the electret mic is connected to its ground pin and the ouput pin should measure between about 1VDC and 4VDC.
The LEDs will probably indicate when your meter's probe is touching the output of the mic.

The LM3916 doesn't have much range, a voltage range of about 5 times. The LM3915 has a voltage range of 32 times and therefore is much more sensitive.

 

[supes]

Maybe the fact that i dont have a 3915 is also limiting my progess. I checked pin 1 and it was infact 2.44V. The output pin was .66VDC. The electret mic is at 1.94VDC. Pin 5 of the 3916 is at .15VDC and pin 6 of the 3916 is .48VDC.

So now, where is a good place to get the LM3915 for a fair price. I want to order a few. I will continue to work on this on my own in the meantime

Thanks for you help audioguru.

Also on another note, I want to make the 'Knight Rider' lights and remember reading that you did not like the 4017 version b/c it wasnt smooth. Have you come across a circuit that does a better job of make the effect? If so please share.

 

[audioguru2]

I checked pin 1 and it was infact 2.44V.

Close enough to half of 5.0V.

The output pin was .66VDC.

That is a problem. The mic preamp opamp has C2 in series with its negative feedback voltage divider. Therefore its DC voltage gain is 1 and its DC output voltage should be the same as its input: 2.44V.

So now, where is a good place to get the LM3915 for a fair price. I want to order a few.

I bought mine a long time ago at Digikey.

Also on another note, I want to make the 'Knight Rider' lights and remember reading that you did not like the 4017 version b/c it wasnt smooth. Have you come across a circuit that does a better job of make the effect?

Sorry, I haven't seen a smoother circuit.View attachment 39530

 

[supes]

I think I may have been unclear. The pins 1, 2, and 3 of the 358N are all 2.36 VDC. The output pin I was talking about was pin 7 of the 358N.

For the second ampifier in the 358N. Pins 5, 6, and 7 are 0, .9, and .52 VDC respectively. Pin 7 or atleast that set of pins is the problem I think.

 

[audioguru2]

The input voltages of an opamp should be the same. Yours has pin 6 very different from pin 5. Replace it.

 

[supes]

Error on my end.

Pins 5, 6, and 7 are 0, .09, and .52 VDC respectively.

However, I still did replace it. Pins 5, 6, and 7  are 0, .09, .52 VDC respectively. Funny that they are exactly the same.

Sorry for the error.

 

[audioguru2]

I think your LM358 dual opamp works normally. When you touched the meter lead to pin 6, the mains hum picked-up by the lead was amplified by the opamp producing +0.9V.
My LM3915 is 6 times more sensitive than your LM3916 and my project displays a few LEDs if a pin is dropped on the floor on the other side of the room. Fairly low level sounds and louder cause all LEDs to light. With voices, music or TV sound the LEDs are constantly changing and following the level.

 

[supes]

What were your values for pins 5, 6, and 7 of the amplifier? Or maybe a better question, what are the ideal values for those pins of the amplifier?

I remember you saying that pin 7 should be between 1 to 4 VDC but I am lower than that.

 

[audioguru2]

R7 connects the input pin 5 of the opamp to 0V. Without a signal, R8 provides 100% DC negative feedback so that the emitter of Q1 and pin 6 are also at 0V. The output pin 7 is at one base-emitter voltage drop above the voltage of Q1's emitter.
A signal causes Q1, Q2 and Q3 to rectify it and produce a positive DC voltage that is equal to the peak voltage of the input signal. The max voltage is about 7VDC.
C8 holds the peak voltage for a short duration so that your vision can see the associated lighted LED, and C9 holds the peak voltage for a longer time for automatic gain control.

 

[supes]

So after a little while working on it today, here is what I found out. Pin 6 after the R9 is getting virtually nothing (.08VDC). We have ~2.5V at pin 1. When that goes through C4 we end up with the .08VDC that goes into pin 6.

If I wanted to test the microphone alone, would I put the meter on the on the non-gnd side of the electret mic. When I do that I get .22VDC and .2VAC. I adjusted music volume and the DC voltage didnt budge.

Is the microphone not being driven enough to actually work? Could that be the problem?

 

[audioguru2]

So after a little while working on it today, here is what I found out. Pin 6 after the R9 is getting virtually nothing (.08VDC).

That is normal. The opamp has an internal voltage gain of 100,000 or more. So only 10uV between the inputs will cause the output to change 1V or more. R8 provides negative feedback so when the input to R9 swings to -2V, then each transistor's output rises to +3.57V.

If I wanted to test the microphone alone, would I put the meter on the on the non-gnd side of the electret mic. When I do that I get .22VDC and .2VAC.

You said before that the electret mic has 1.94VDC which is normal. Why did it drop to only 0.22VDC?
The output of the mic is low level and high impedance so your meter's lead will pickup mains hum. Measure the AC at the output pin 1 of the 1st opamp which is amplified 101 times and is a very low impedance to avoid hum pickup.

I adjusted music volume and the DC voltage didnt budge.

The opamp's pin 1 DC voltage won't change. The opamp's pin 7 DC voltage will increase when there is sound.

Is the microphone not being driven enough to actually work? Could that be the problem?

Find out why its voltage dropped from 1.94VDC to only 0.22VDC.

 

[supes]

I tried a different electret mic. It is now reading 1.44VDC. The mic I had may be bad but it still isnt the solution. Going through the pins I get :
pin 1: 2.34VDC, 4.6VAC
pin 2: 2.34VDC
pin 3: 2.34VDC
pin 4: 0VDC
pin 5: 0VDC
pin 6: .08VDC
pin 7: .51VDC, .6VAC
pin 8: 8.14VDC