This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

LM3915 cascading amplifier

Other Parts Discussed in Thread: LM3914, TL084, LP5009, LP5024, TPIC6B595

I am working with cascading an LM3916 LED driver with an LM3915. The data sheet states the best practice is to use an op amp to increase the signal to the "first" IC. In my configuration, the 10V signal fed to the 3916 ("second" IC) with a V(REF) = 10V will cause LED 10 (+3 dB) to illuminate, indicating full scale. LED 10 of the "first" IC, an LM3915, is designed to illuminate at -30 dB. With 33 dB between LED 10 of each device, the data sheet implies that an op amp with A = 44.668 is needed.

With the audio signal attenuated to 10.0V under all circumstances, this signal is fed directly to the 3916. But fed through the op-amp with A = 44.7, this means the LM3915 pin 5 (SIG) will see 447 V. The data sheet states the LM3914/15/16 family can accept up to 35V max at pin 5 before damage occurs. Thus, following the data sheet means burning up the first LM39xx chip once volume levels increase.

What am I missing?

  • Hi Hunter

         Could you please provide a block diagram or schematic for your system? I'm a little confused about your description.

    BR

    Sean

  • The attached schematic shows left and right channel for a 100 dB power meter. Upper left is a power supply providing +12, +6 and 0V to the circuit (D1 = 6V2 Zener diode). Upper right shows left above, right below into a "precision half-wave rectifier" as shown on the LM3915 data sheet, p. 18., with ballistics R4/C6 meant to provide peak output (vs rms). This circuit also provides gain of A = 0.1724 with the intent of dropping a 105W maximum input (29.0V) to 5V. V(ref) for the 3915/3916 pair will therefore be (100W / 8 Ohm) 28.28V x 0.1724 = 4.88V, denoted as +3 dB (the amplifier produces 50W rms with 3.2 dB of headroom, or 105W peak). The signal from this junction is split, with this signal fed to the 3916 (driving LEDs representing +3, 0, -5, -10, and -20 dB). The signal also goes to an op-amp for amplification of 33 dB (+3 vs -30 dB) for the 3915, or A = 44.668; output from pins 14 and 8 are sent to the 3915 which drives LEDs representing -30, -40, and -100 dB. The TL084CN is configured per TI's Single Supply Op Amp Design Techniques Application Report SLOA030A with the intent of producing an output that illuminates the -100 dB LED at all times (Vout = A Vin + Vbase, with Vbase being above the threshold needed to illuminate 3915 pin 1) -- this is case 1, p. 8.

    I have not breadboarded this yet. Typically I go straight to prototyping by ordering circuit boards and see if it works, but given the complexity of this project I'd rather get an idea if it's totally wrong before spending $30 on circuit boards.

    Questions:

    Being a mechanical engineer with some electrical engineering coursework way back when, I don't fully understand how to make the op-amps work properly with a single voltage supply, nor do I understand how I would generate a +/-12 V supply that would otherwise be needed to make an op-amp work. So I'm not quite clear on what happens when we feed +6 volts (virtual ground) to pin 4 of the 3915/16 pair or the the non-inverting input of the TL084 (pins 3 and 5).

    I do not want 20 LEDs; as I stated above, I'm looking for eight. I also want dot mode display. It seems to me that by tying the cathodes of an LED that is to be illuminated not only at -5 dB but also -3 dB and -1 dB would do the trick, although I know this is not what the circuit in the LM3916 data sheet p. 11 where pin 1 of IC #2 is unused; because I am using pin 1 of IC #2, I figured that this did not apply to my project.

    Thanks for the insight.

    Schematic_power meter_2020-06-10_08-59-36.pdf

  • Also, and the point of my first question, with A = 44.668, when the amplifier produces, say, 10V, then TL084 will be asked to produce 77 V (10 V x 0.1724 x 44.6) which if my understanding is correct is obviously impossible with a 12V supply. So I suspect the maximum voltage the 3915 will see at pin 5 is 12V, which is within the design specification of the 3915. But then, does this negatively affect the TL084?

  • here's a block diagram of a single channel, might be more clear.block dia.pdf

  • Hi Hunter,

         I try to understand your system and requirement, but it seems very complicated for me to understand what you want, First I think you need to separate your schematic, there are too many wires for me to understand. 

        I recommend that our LP5009, LP5024 for your design.

        And for the amplifier question, I would suggest you to submit another post, and that will be forward to another team to help you, thanks.

    BR

    Sean

  • Sean, here's the block diagram which will be far more clear than the schematic. Sorry for the confusion. I'll check out the LP5009 and LP5024.

    8244.block dia.pdf

  • Sean, took a look at the LP5009 and LP5024. I'm attenuating the +29V signal down to +5 V so that this can be used with an Arduino. It is exceptionally straightforward to use the Arduino and a LED driver (Texas Instruments TPIC6B595 x 2 comes to mind, which I've used in another project, as does the MAXIM MAX7221 which can drive two arrays of 8 LEDS as well). Issue is, I'm trying to make the circuit quasi-passive, rather than spending another $25 on a uC. Because the LM3914/15/16 family were designed specifically for this application, it should be straight forward getting them to work.

  • Hi Hunter,

        From my understanding, LP5024 should be very suitable for this solution which is easier to understand.

    BR

    Sean