Randy, LEDWiz question - CHECK IT OUT, youtube embedded video!!!!

[mccoy178]

Okay, I've been trying to figure out how to make a 20 channel x 20 level graphic equalizer out of LED's.  I've found tons of kits that are 3 channel and schematics for 10 channel, but nothing that has really hit me.  I did discover the LM3915 chip that can take variable input and transfer that to the led's.  Here is a sample video on youtube of it in action:

[youtube]http://www.youtube.com/watch?v=WEytbrTY2Ic]http://www.youtube.com/watch?v=WEytbrTY2Ic[/youtube]

I can handle the wiring of the led's.  What I don't know how to do is split sound into 20 frequencies.  My question:
Can software be written to send individual channels to individual inputs on the LEDWiz to replicate the same feature that a bandpass filter would?  I have seen many, many software based graphic equalizers, (winamp among others comes to mind), that look great on the screen, but is there a way to send that to the LEDWiz to power the led's?  I know that there are a few of us out there that would love to have this.  Here is an example of what I'm referring to:

[youtube]http://www.youtube.com/watch?v=yrewCzKJo9I]http://www.youtube.com/watch?v=yrewCzKJo9I[/youtube]

Thanks for any input on this!

[RandyT]

Here's the problem as I see it;

That particular display looks like it has over 150 individually addressable LED's in it.  The LED-Wiz could drive it, with the right software, but it would take 5 of them to get the job done.  It would probably react a lot smoother than the clip shows, but not worth it for what you want to do.

That chip basically does an A/D conversion and then drives the outputs based on the analog levels of the input.  IOW, it needs analog in, not digital.  So what you will most likely need is one chip for each column of LED's and the appropriate capacitors to filter out a specific frequency band of audio as an input for each chip.

And if that's your setup in the video, the specs state that the output of the chip is current limited, so you probably don't need those resistors for the LED's (but double check before taking my word )

*edit*

One other possibility (that I just now considered ) Is whether or not the PWM output of the LED-Wiz would be averaged by the A/D converter and as a result, drive the display accordingly.  If so, the PWM levels of the outputs could indeed drive such a display (again, with as many chips as you want "display bands" using as many outputs on the LED-Wiz)

But I don't have one to try.  And if you do, I'd love to hear whether it works or not

RandyT

[mccoy178]

Well, how does the saw effect work with the LEDWiz?  Couldn't the signal be sent much like that to a single port on the LEDWiz for each channel, let the 3915 use the signal sent to do its thing?  I mean, would it be the same to send a saw type of effect to the LEDWiz as it would be to use a potentiometer to adjust the input to the LM3915 like in the video?  That's what I'm asking.  If so, couldn't software be written to take a graphic equalizer on the computer, and send out the individual channels to individual ports on the LEDWiz?  Much like sending an audio analyzer signal to a betabrite or other form of output?

[RandyT]

Well, how does the saw effect work with the LEDWiz?  Couldn't the signal be sent much like that to a single port on the LEDWiz for each channel, let the 3915 use the signal sent to do its thing?  I mean, would it be the same to send a saw type of effect to the LEDWiz as it would be to use a potentiometer to adjust the input to the LM3915 like in the video?  That's what I'm asking.  If so, couldn't software be written to take a graphic equalizer on the computer, and send out the individual channels to individual ports on the LEDWiz?  Much like sending an audio analyzer signal to a betabrite or other form of output?

See my edit above.  I don't know if it will work without understanding how the chip in question works.  All effects done by the LED-Wiz are digital.  And by that, I mean in the same sense of digital sound.  We're talking manipulated modulation frequencies, not analog voltages.  However, a multimeter on the output of the LED-Wiz shows voltages that correlate pretty well with the PWM frequency (intensity) so apparently the meter is doing some rudimentary sampling and averaging the output into a voltage.  If the chip in question works the same way, then yes, such a display, based on those drivers, can be built and software used to control the the output.

But I don't have one, so I can't try it

RandyT

[mccoy178]

Here is a quote from another website:

LED VU Meter

IC1: LM3915 18 Pin IC socket Does your amplifier have a level indicator? I always envied the fancy amps with LED level bar graph that you find on most stereo systems or amplifiers. This is your chance to build your own LED VU meter that you can link to your amp and watch it in action. The application is based on LM3915 chip. This is a monolithic Dot/Bar Display Driver IC made by National Semiconductor. It takes an analog voltage input on pin 5 then drives 10 LED’s providing a logarithmic 3dB/step analog display.

When measuring power, a 3 dB increase means that the power input has doubled. As the power doubles, an additional LED will be lit until the maximum is reached. The display can be bar or moving dot depending on the connection of pin 9 to the positive supply. The LED drive current is regulated which eliminates the need for current limiting resistors. The supply voltage can be between 3V to 25 V. Full specifications can be downloaded from LM3915 LED VU Meter National Semiconductor site.

http://www.radiolocman.com/shem/schematics.html?id=33991

[mccoy178]

Okay, if your product puts out a digital signal, could something like this be used?

Long link

http://www.analog.com/en/prod/0%2C2877%2CAD5641%2C00.html

[mccoy178]

Randy, what type of output is needed for the LEDWiz to read?  Barcrest is willing to write the software for this, but he needs to know what kind of signal is required.

[RandyT]

Randy, what type of output is needed for the LEDWiz to read?  Barcrest is willing to write the software for this, but he needs to know what kind of signal is required.

Do you have a source for these parts?  We really need to try one.  Do you have one we can check to see how the comparators deal with the PWM output of the LED-Wiz?  Without knowing that, the software is, I'm afraid, pointless.

RandyT

[mccoy178]

Here is some links to each products .pdf:

LM3915

LM3916

I have ordered a few of the LM3915 from digikey yesterday.

[RandyT]

Here is some links to each products .pdf:

Thanks.  I have seen these already.  Just to make sure I'm not misunderstood, what I don't know is what is happening at the input buffer stage of the device.  The schematic shows that this device is pretty much a "resistor ladder".  It works by virtue of the fact that each LED driver has the resistance of it's own resistor, plus the resistance of every other resistor lower in the ladder because they are in series.  This means that the first driver will take very little voltage to cause the LED to light, while the voltage requirements for the drivers that follow become greater the further up the ladder you go.

I've never tried running a PWM output into a resistor ladder, but I'm not convinced it would work.  However, if the buffer at the input of the chip does anything other than just pass it through, there might be a chance. It's going to take some experimentation and other components may be required.

RandyT

[mccoy178]

Okay, I'll use one of the LEDWiz's I have from the jukebox and breadboard a setup with just basic PWM going to an input.  I'll report back next week once I get it up and running.  Thanks for the feedback.  If you happen to know how and where to get bandpass filters, I'll be all over that too.

P.S.  Will wanted me to ask if I could hook up the LEDWiz to 110?!

[leapinlew]

P.S.  Will wanted me to ask if I could hook up the LEDWiz to 110?!

 

[RandyT]

I just wanted to post a quick note to anyone watching this thread stating that I should have an answer on this soon.  I bought a couple of the parts in question, but didn't have the correct value resistors to finish breadboarding the circuit.

After spending some time with it, and thinking about it a little more, the LED-WIz is not very compatible with this particular device in it's normal state.  It may still work, but would probably require the software to invert the output and require a pull up resistor on the input of the driver chip, or the use of an inverting buffer chip.  But even without the extras, I should be able to see what the chip does with the PWM using just the pull-up resistor on the input.

As the chip appears to be expecting the possibility of the input to be modulating, they also show a precision averaging circuit in the datasheet.  It needs a handful of parts, but may do the trick if all else fails. 

I'll post again if I make any headway.

RandyT

[brandon]

so.. how does the LEDwiz vary the intensity of the LEDs?  does it output a square wave of varying frequencies and duty cycles? 
 EDIT:  nevermind.. I didnt know that's what PWM meant 

PWM is also used in efficient voltage regulators. By switching voltage to the load with the appropriate duty cycle, the output will approximate a voltage at the desired level. The switching noise is usually filtered with an inductor and a capacitor.

wow.. learn a lot with this hobby