Ultimate Case Mod HELP!! Electrical Eng. needed

[nheinric]

Ok I have an idea for a case mod, but I think I need some help on it. I am an electrical engineering student, so this should be well within my abilities, but it would be nice to get some input on the idea, or if anyone knows of anyone else having done this, let me know.

I was thinking about my SB Live! Platinum, when something hit me. I am using the digital connection for my speakers which leaves me a pair of analog out spots on my sound card. So I thought wouldn't it be neat to make some sort of spectrum analyzer/vu meter with the analog outputs on my SB live. Here is what I am thinking of doing.

First I will have to amplify the signal from the outputs. Probably using a 741 opamp. Then I need to run the signal through 3 filters. High pass, low pass, and mid range, to split the signal. Then probably another amplification on each of the three resulting signals, and finally feed the remaining signals into three bar graph leds. This way anytime my sound card was outputing sound, you would get the neat moving lights. I have a rough schematic drawn up for it, but I would like someone who knows what they are doing to help with the design. The bar graph led's would of course be mounted in a 5.25" drive bay. I was thinking two identical setups. One for the front dignal and one for the rear. So 6 bargraphs in all. Of course if you only have a single analog output it could be done just singley. If anyone is interested in this project, email me at nheinric@d.umn.edu. I would really like to get started, but need some help. This would be the altimate addition to your case!!

------------------
ATX Full Tower w/300W PWS
Asus P3V4X
PIII 667 @ 700 (Slot 1)
128 MB PC133 RAM
WD 20.5 GB 7200RPM HD
Pioneer 16X/40X Slot Load DVD
Pletor 12X10X32X CD-RW
SB Live! Platinum w/LiveDrive 2
Voodoo 3 3500 TV AGP
CS FPS2000 Digital Spkrs
Logitech cordless Desktop
Viewsonic 19" PF795

[S_Klass]

It's been done and I've seen the website of a person who did it. He had a little how-to there that could be really useful to you. The unfortunate thing is.. that site has been offline now for a while. I know he was running it off his own spare PC so that probably explains why. It was called Project Si or Sj and had 3 parts to it. any one with an updated link please post it. Thx!

------------------
-=S_Klass=- I tweak... therefore I am.

[D|sElMiNk]

http://www.ocaddiction.com/reviews/m...gold/newq.html

Cool, but you'd break into quit a sweat to replicate that

[nheinric]

Not looking for anything quite that complicated. I don't need to adjust it like a graphic equalizer, I just want to see the output level of three-four frequencies of the signal. I think it could look nice with just some bargraph LED's and a switch or something. It wouldn't really be all that hard I don't think. Plus a lot cheaper than 100.00 they want for that thing!! Anyone with some ideas?

-Nater21

[Storfe]

nheinric: i have though of that myself.....

Gonna check Virtual Hideout now to see if i saw something like that there....

------------------
----> http://badsector.mainpage.net/ <----

[Storfe]

ummm, nothing there.

Gotta do the planning then
I'm interested in doing this thing, say we cooperate then ?

Fire off and mail at: kristian.frogner@hm.telia.no

Got "some" eletrcial skills, (i kinda like digital stuff better), but analog stuff isnt that big beef either.....

just a quik though:

soundcard-amp-filter-amp(?)-resistor circuit to get the led to light up at different voltages-bar.g/leds

Anyway, mail me! (anyone interested in helping for this project)

we could also include a "line in" and a switch for turning on a mic, and run it all of the +5V Standby, so it would work when the PC is off!!!

------------------
----> http://badsector.mainpage.net/ <----

[nheinric]

Storfe - YGM


Yeah let's do this. I like the mic input idea. Just so we are thinking the same thing here, I don't want this to adjust any thing with the sound output, just show you visually what the sound is doing, like on some stereos.

Anybody else interested in this project??? Email me or storfe. My email is nheinric@d.umn.edu

Come on guys it will look great!!

:-)

[Philly]

if you figure that out, email me

grnranger311@aol.com

i don't know too much about all that, i work on car stereos (anything DC actually) and i know a mcp (sp) certified installer locally and if you have questions i might be able to ask him, but i doubt he can help

sounds like a badass idea...

------------------
computers dont have speed limits - NOT ANYMORE

[D|sElMiNk]

Hey Storfe - how about integrating that digital crossfader you and Ferguson where working on ? You got it to work almost right didn't you? Only problem was that it didn't feel like fading back right ?

You still got that thingy btw? I never got to hear it

[Storfe]

Diselmink:
Ferguson got it... (i think)
It never really worked like it was supposed to.... wasnt quite "stable" after all, and it was made out of 12 ic's or something, and a cable mess from hell!!! The diagrams werent that good either, we kept changing it without adding the new stuff to the diagram, so its a hell to rebuilt.... and one of the digital fader circuits were broken too, need a new one from Elfa.... With a little more work, it would have work, but our teacher, woudnt give us more time.....

Anyway: Its only a analyzer, it isnt gonna change the sound output, just view it

[nheinric]

Storfe-

I got an email from a guy in England (mark radon) who had a few suggestions, that make tons of sense. First off, he has done projects like this and is glad to offer advice. He pointed out a few flaws in our design already! Here's what he suggests.

Instead of having the second amplifier stage (after the filters), we need to have a peak voltage detector. HE said the circuit was pretty simple, just a couple of diodes, capacitors and operational amplifiers. BTW, I think the word you are looking for when describing components of a filter is an operational amplifier, which I will refer to as an op-amp. The most common is the 741 which I plan to use.

Here is his reasoning behind the peak voltage detector. The signal will be constantly going up and down, at a very high frequency, if we just amplify it, we will not be able to see the changes well, all we will see is all of the leds very faintly lit up. IT won't be bright and there won't be much change in it. What a peak voltage detector does, is it holds the peak voltage for a small amout of time with capacitors I believe, giving us a stronger signal to send to the LEDs. Also he said a resistor array was a bad idea, because as the voltage increases, our LEDs will become dimmer and dimmer. He suggested using an Integrated circuit called a bar graph controller, which takes in an analogue signal, and based on a few reference voltages, drives up to 10 LEDs. He said it is also on a logarithmic scale which means it will be more accurate as sound is measured in decibals which is logarithmic. He sent a link to where we can get the IC's, but I don't have it here at work. I will send you a copy of both his emails when I get home tonight. The IC's are not very expensive if you order 25+ of them. I believe they are about 1.82 USD each. So I figured I could order some, and then send you half, and you could just paypal me or get me money somehow.

Yes obviously the more filters, the more frequencies we can monitor. We need a good band pass active filter circuit. If you can find one in a book, email me post the circuit so I know what we are looking at. Then we can just change capacitor values to get different frequency ranges. As far as the bar graph LED is concerned, all it consists of is 10 square shaped LEDs put together in a bar. So yes it would be the same thing to substitute in regular LED's, I just think the Bargraph ones look better and more proffesional.

As far as simulation software goes, get a copy of PSPICE. I have the MicroSim Version, if you want I can burn it as it is shareware, and Send it to you next week.


[This message has been edited by nheinric (edited 07-21-2000).]

[nheinric]

check out this link also - it's another thread I posted, but someone posted a picture of a variation of what we are doing, although his is only a VU whereas ours is in reality a spectrum analyzer.
http://pub19.ezboard.com/fgideontech...picID=10.topic

[nheinric]

I guess I do have his emails, I will post the big one, in it is a link to the IC's.

I see what you are trying to do. However, while it may work, it certainly
won't work well. All your LEDs will light up at different brightnesses, and
you will struggle to make a logarithmic scale (essential for a spectrum
analyser - just look at Winamp to see how naff a linear spectrum analyser
is.) You'll also find that your op-amps won't be able to power the LEDs
well, so that as more light up, they all get dimmer.

Because it's a pain to make LED bar graphs work well, some enterprising
semiconductor companies make ICs that do the job. The only problem is that
they aren't particularly cheap - I doubt you could get them for less than
about $4 each. (they are cheaper, check out the website below. For 25+ they are only 1.87 each! That sounds cheap enough to me!)

I propose the following basic plan:

1. Buffer amplifier
2. Bandpass filter (ideally active, but a simple passive filter will do)
3. Signal rectifier with peak detection
4. LED Bargraph driver IC
5. LED bargraph
(Obviously parts 2-5 are replicated for each channel)

1. Buffer amplifier to provide a high impedence to the sound card.
2. Using an active filter rather than a purely passive one has the advantage
of having a very low output impedence, so you don't run into problems when
coupling the filter to the next stage. All it will cost you is an extra
op-amp per filter. Look in a decent electronics textbook for a bandpass
active filter circuit.
3. You need to rectify the signal, and then catch the peaks and hold them
for long enough so that they can register on the bargraph - otherwise you'll
just get a blur as the voltage rises and falls with the waveform.
4. Feed the signal into the LED driver IC, and attach your LEDs directly to
the 10 output pins. No worrying about current limiting resistors, and
resistor chains - it's all built into the IC.
5. Use whatever you want for this stage, it'll be fine.

Go get the datasheet from here: http://www.national.com/pf/LM/LM3916.html It
tells you all about how to work the LED driver ICs, and also includes useful
circuits like rectifiers with peak detection.

Cheers,

Mark

[Storfe]

I have found some band-pass filters, ill scan them later this weekeend when i get the time.... they are from some 2 years old schoolbooks, i assume they will work. They are made out of a couple of C's and R's. A little wiring required, but the components are cheapcheapcheap. BTW: how big is Pspice (Microsim version) ? Anything below 30mb is downloadble for me.....

Does the Pspice evaluation version contain all the circuits we need ?

------------------
----> http://badsector.mainpage.net/ <----

[Storfe]

The picture at the other messageboard is the circuit i was talking about, but as said, its just a VU meter for right/left channel, were gonna do it better ! =D


------------------
----> http://badsector.mainpage.net/ <----