Peak meter for certain frecuencies

[Menticol]

Can a 2N3906 be used for Q1 - Q10? the schematic is below

(the circuit is fed with 15V, according to the power supply draw)

 

[audioguru]

The original Oriental circuit uses the LM3914 with its outputs set to 12.5mA directly driving the cathodes of the LEDs. 2N3906 transistors driving the anodes of the LEDs will be fine with such a low current.

But multiplexing reduces the average current in the LEDs to only 1.25mA so they will appear extremely dim. Many things need to be changed to increase the current.

 

[Menticol]

oh yes I forgot to say that, I'm using LM3915 because I have seen your warnings about the dim light, cannot find LM1916

 

[audioguru]

The LM3916 is a VU meter LED driver. The input voltage step spacings are all different.
The LM3915 is a logarithmic LED driver. Each input voltage step is 3dB more than the LED below it.
The LM3914 is a linear LED driver. Each step is 10% higher voltage than the LED below it as a voltmeter.

All three LED drivers have a max output current of about 25ma which is nice and bright without multiplexing.

 

[Menticol]

Thanks by the data, Audioguru. I got the board etching done, but before go shopping I'm worried about this Eagle's warning (don't laugh)

WARNING: Sheet 1/1: POWER Pin IC2P VSS connected to GND
WARNING: Sheet 1/1: POWER Pin IC2P VDD connected to VCC

Vss must be 0v, so what's the problem connecting it to gnd? should the oriental circuit be redesigned??

WARNING: Sheet 1/1: POWER Pin IC1P V+ connected to VCC
WARNING: Sheet 1/1: POWER Pin IC1P V- connected to GND

another thing, according to datasheets the LM3914 and LM3915 share the same pinout, but could be used indistinctly with the oriental circuit? (I prefer the shame of asking, instead of blowing the money up)

pd: schematics included below

 

[MrDEB]

The orential circuit needs redesigning

I have been working with this circuit as ewell as tghe vellman circuit.
the vellman uses bi lateral switches instead of transistors.
the bargraph driver (3914,15,16) needs some help as the leds will be too dim.
Audioguru and I have been throwing out stuff right and left.
getting down to maybe something that works.
will keep yopu posted or you can add to the discussion, the more the merriery

 

[Menticol]

I lost my money due my bad english!! fortunately yo warned me.

Well, what about using smt leds? I think they are less current - hungry, and no re design is needed

 

[MrDEB]

thats what we are using. the smd leds use about the same as convential leds.
2v forward current and 15-20 ma

 

[audioguru]

The LM3914 is a linear voltmeter. Each step is 10%.
The LM3915 is a logarithmic sound level or audio power meter. Each step is 3dB.
The LM3916 is a VU meter. Each step is different.

They all have the same pins. They are designed to directly drive 10 LEDs, Extra transistors are needed if the outputs are multiplexed.

 

[ronsimpson]

I just wanted to bring this thread up to date so it is easier to find. Some people are looking for it and can not find it because it is old.

 

[canadaelk]

I just posted this circuit on another thread here. If there is interest in the pcb I am just finishing off, I will make the gerber files available:

 

[audioguru]

The average LED current in the Vietnamese project in post #30 is only 1.2mA which will appear very dim.