... refurbished stuff from the attic box in a modern computer? Why not! :-) ...


 
Mounting some different colored LEDs for different purposes onto a usual beige computer case is not a bad idea to "kill" a boring look of it; if they are somehow "animated" even better. One of the ways is to mount a LED VU meter into an empty 5,25 place in a box cover to have a nice display showing the audio signal level/power when listening to your favorite CDs & Mp3s. You can buy it or make it at your own.
 
Since I bought a really lot of time ago /see the year on Pcb / for totally different purpose a Kit one with displaying linearly the input voltage values (the chip used was a linear Led display driver!), I had to re-engineered it a bit to show values logarithmically with the same chip involved (here is that small innovation) & to rise its sensitivity by adding an amplifier pre-stage to use it with lower level signals, rather than originally ment to be directly connected to already amplified speakers wires.
 
After some experimenting & studying the behavior of that chip, the final schematics came out as can be seen on lower slide; I also replaced some LEDs, so now are there all opaque ones: 7 normal green & 3 red ones, one darker & lighter green one and one yellow bright one that shows overload. Since there on slides & pics everything is IMHO quite well color coded & seen, about the assembly of this circuit is not much more to be said & can be assembled also using universal already perforated proto_board ...
 
LED audio peak meter SCH
 
Well this is how it works [DivX movie, 300kB DL] : the signal from the PC Audio card (taken from SPK out parallel with wires for those amplified PC speakers) is sent to the gadget input, where is than amplified in transistor stage & than full wave rectified to feed the pin 17 - input of IC, which takes care of switching LEDs on & off in a bar (line) according to the signal level of the music. This way the LEDs are lit linearly in a small range & not logarithmically as they should to be with accordance to a human ear subjective perception of the loudness of a sound. That was the case with original schematics I had to resolve & I did it my way ...

Since the voltage on a pin 3 is a threshold that determines how many LEDs are lit w/o input signal and has the opposite behavior than the input pin, the idea was to also supply to that pin some of the rectified signal to achieve "some kind of compression" of a driving signal inside the chip that it can show with LEDs much wider range of a signal. That is done thru that 18k resistor & a high (radio) frequency germanium (AM demodulating glass housing type) diode. The proper LED range (compression) is achieved by the ratio of the signal on that two pins I determined for my needs long time ago & it works like a charm.

 
(To see detailed those 3 pics down, clicking on any will kick you to the Computer section)
 
outside View parts View
 
PCB side Clicking on both slides enlarges to details; left printed circuit board (PCB) & the right one pops up parts mounting & measures - mounting the gadget inside PC 5,25 bay on the plastic cover is turned with PCB upside & LEDs alone can hold it mounted in a place ... parts layout
 
. . . Some notes : the first LED on the board is mounted flipped than all other ones & swapped with its serial resistor & 2M resistor is composed of two serially 1M ones. This very old IC on schematics, since IMHO is out of production, is very difficult to find, but maybe is some equivalent still around, who knows; maybe can be found in some electronics junk yard if someone tries to make this stuff ... & there could be also some problem with the pre-amp transistor (Fairchild datasheet ) to find: it has a very high gain about 1500; but there is also a Motorola (MMPS A18) version of it, which has even much higher gain & tends to self-oscillate, so it is not suitable ! But IMHO fortunately can be used for example a common BC109C one there instead, BUT than have to be some gain adjustment done: lowering the value of resistor 2M to 1,5M & a 150 one to 100 ohms IMHO would be Ok to rise that stage gain a bit ... To rise a bit that also can that 2 10k input resistors be lowered a bit, but not less than to 1k not to "kill" the stereo effect on the audio card (shorting L+D to much & so making it Monophonic).

P.s.: got a report of someone starting to make this circuitry that he actually bought the IC in the store for about a dollar or two (new) ; so it looks like that this IC can still be found around ... ... nice . There were also some reports from others who made this my circuit, that (with newer production ICs than mine) can happen, that except the first LED also another 2 can be lit w/o any input signal. If this happen, lower the 22k resistor a bit to 20k or so by adding parallel to it another big resistor like 270k one. If the second Led is not switched off yet, lower that resistor a bit more (but not less than 18k in total ! ).

 
. . . Maybe someone would like to have a stereo version of it. In that case not to make 2 identical ones could be better to take some another schematic for that, but IMHO, here is no need for a stereo version . Why? The modern music audio signal peaks are mostly determined by a bass guitar & bass kick drum signal peak levels & ... those two are mainly recorded believe it or not, mono balanced in the middle of the sound panorama & mostly also main vocals, so mostly their signal is equally present on both channels . The bigger difference would show with ordinary VU meters only when those signals would not be present &/or leading instruments "floating" between 2 channels ...

 

 

© Zdenko Jerman-Spajky