LM339 and a simple 4 segmented VU meter

[Potato Cakes]

Hello, Everyone,

I have been trying to build a four segmented LED VU meter but I am a little bit stuck. I am working off of the attached schematic from a console service manual. The differences are I am using 1N4148s, a BC560, a TL072, an LM339, and a +V/0V power supply. I have seen TL072 used with single source power supplies but I do not know if that can be done in this configuration. I may have to use a virtual ground scheme.

In my test build, all of the V- connections from the schematic are connected to 0V/GND. When I wired everything up, all of the LEDs lit up, so I started taking things apart to do some individual tests. The first problem I an running into is the TL072 buffer/rectifier circuit. It is my understanding that the first amp is the buffer and the second is a gain stage where the AC is rectified by the 1N4148 (BAS16 on the schematic) which is then fed to the LM339. However with no signal, the DC on where the 1n4148 and 1k8 resistor meet is equal to the power supply. When I apply signal to the input, I get less than a 1V of change, even at +18dBu and the input. So I guess the first thing I need to make sure I get is that change in AC voltage to the op amp should show a change in DC on the other end of the diode or am I missing something?


Thanks!

Paul

 

[ruffrecords]

Your problems all stem from converting from a dual to a single supply. This usually means your op amp outputs will sit at half the rail voltage and that also need to be the 0V reference for the resistor ladder that operates the comparators. Post your actual schematic and we should be able to tell you how to fix this.

Cheers

Ian

 

[Potato Cakes]

Thanks for the reply, Ian. I'll have to sketch something up and post it. The only thing I am really modifying is the use of a single LM339 and adjusting the resistor values for the corresponding desired signal levels.

Thanks!

Paul

 

[Potato Cakes]

Here is the modified meter circuit I am using. Once I first powered it up and all of the LEDs came on I began to think that single source power supply was probably causing the issue but I didn't know why. I tested the R? for adjusting the signal level feeding the circuit but it did not change the gain for some reason, so I left it as a jumper as in the original schematic.

Thanks!

Paul

 

[Potato Cakes]

Here is the power section. Just for testing proof of concept I omitted the 100n cap between +V and ground as shown in the original schematic.

 

[Potato Cakes]

Also realized I didn't change the component designations on the 4 segmented meter schematic that I mentioned above. For clarity:

4580 = TL072
BAS16 = 1N4148
BC858 = BC560C

Thanks!

Paul

 

[ruffrecords]

The problem is the output to the meter sits at a dc value of LG2 and you do not show how this is derived. The connection to ground of the comparator resistor chain should instead be connected to LG2.

Cheers

Ian

 

[Mr. Moscode]

Check this IC out LM3915, very simple 10 led VU driver chip with 10 steps 3 db apart. Input signal is referenced to ground. You can block DC at the input with a coupling cap.

The LM3914 is the linear version of this.

 

[Potato Cakes]

LG2 is just ground. I didn't change that from the original and should have for clarity.

 

[Potato Cakes]

The range I am wanting to cover exceeds 30dB. I have looked at those before and had used the LM3916 before it became obsolete.

I built a rail splitter but stupidly hooked up V- and V+ backwards and heard a crackle before I realized what I had done and disconnected power. I connected everything correctly and I still get the same issue, except now I have +4VDC on the Meter OUT connection without any signal, and it fluctuates constantly within 0.5V. Don't know if I fried anything but if I did that really sucks because I will have to build a whole new thing to test.

The incredibly simple, pedestrian simple circuits have been the biggest headache and source of frustration as of late...


Thanks!

Paul

 

[Mr. Moscode]

IC's such as these aren't protected from reverse voltage. My informed guess is it is cooked. Dan't ask how I know.

You can stack the LM3915s and get a 100 led array. It's in the app notes and probably the data sheet.

If you just want 4 lights that cover 30 db what db points the 3915 covers 30db. Pick the points you want.

Unless I missed something. . .

 

[Potato Cakes]

Tried hooking up with a bipolar PSU and same results. Looks like I have to start over. I know the schematic works I just have to figure out how to scale to my own needs.

As I stated previously, the range I am trying to cover exceeds 30dB. It has to be done with one comparator.

Thanks!

Paul

 

[Potato Cakes]

Found something more inline with what I'm after and it uses TL072s for the input, which I have quite a few. I would be interested to know how this could be adapted to be used with a single source PSU.

Thanks!

Paul

 

[Mr. Moscode]

Here's a problem. You either need to have a resistor across C49 so it can discharge quickly enough to respond to the next event or you can take it out of circuit.
It gets charged but it can't discharge. I don't think you even need it unless you want a peak hold function or a slow display. You can experiment. Make the resistor 1K8 and give the buffer amp 6db of gain to make up the 6db loss. Find a cap size that gives you the results you want.

You have an active rectifier so everything is a pos voltage but the cap can't discharge thru the diodes and nearest resistor is 82 k

 

[MidnightArrakis]

Maybe one of these will suit your needs:

























Hope this possibly helps you!!!

/

 

[TJTex]

Found something more inline with what I'm after and it uses TL072s for the input, which I have quite a few. I would be interested to know how this could be adapted to be used with a single source PSU.

Thanks!

Paul

Have a look at the attached PDF. This should solve your problem of a single power supply. I have not had a chance to run it though a simulator. However, there may be some adjustments to be made because you did not specify the following items:

- Supply voltage you are using
- METER_IN signal.
- LED specs

METER_IN will impact the values for R4 and R5 which impacts the gain of U2B.

The Supply voltage and the LED specs will impact the value of R13, R14 and R15 in the current source that drives the LEDs.

Finally, the values of R7, R8, R9, R10 and R11 in conjunction with the gain set in U2B will determine the signal levels for triggering your LEDs

Also, as pointed out by Moscode, the 1uf cap might impact the response. Currently there is a discharge time constant of 150 ms set by the cap and R5 + R6.

 

[MidnightArrakis]

Have a look at the attached PDF. This should solve your problem of a single power supply. I have not had a chance to run it though a simulator. However, there may be some adjustments to be made because you did not specify the following items:

- Supply voltage you are using
- METER_IN signal.
- LED specs

METER_IN will impact the values for R4 and R5 which impacts the gain of U2B.

The Supply voltage and the LED specs will impact the value of R13, R14 and R15 in the current source that drives the LEDs.

Finally, the values of R7, R8, R9, R10 and R11 in conjunction with the gain set in U2B will determine the signal levels for triggering your LEDs

Also, as pointed out by Moscode, the 1uf cap might impact the response. Currently there is a discharge time constant of 150 ms set by the cap and R5 + R6.

Looking at your schematic reminded me to download the KiCAD 7.0 software. THANKS for the reminder!!!

/

 

[ruffrecords]

LG2 is just ground. I didn't change that from the original and should have for clarity.

Which is probably your problem. You cannot operate an op amp from a single rail supply and reference the inputs to 0V. You need to bais them to about half the supply voltage.

Cheers

Ian

 

[ruffrecords]

Have a look at the attached PDF. This should solve your problem of a single power supply. I have not had a chance to run it though a simulator. However, there may be some adjustments to be made because you did not specify the following items:

- Supply voltage you are using
- METER_IN signal.
- LED specs

The other factor to consider is the series combination of the LED forward voltages. These obviously need to be somewhat less than your supply voltage.

Cheers

Ian

 

[TJTex]

The other factor to consider is the series combination of the LED forward voltages. These obviously need to be somewhat less than your supply voltage.

Cheers

Ian

Which is why I need the LED Specs and the supply voltage.