Procrastinatus

Why talk about it now?

LM3915 Based Spectrum Analyzer

Posted by aonomus on July 8, 2008

While preliminary, I have schematics mostly drawn up for a LM3915 based spectrum analyzer using Fliege bandpass filters to give me bands at 60Hz, 150Hz, 400Hz, 1kHz, 2.5kHz, 6kHz, and 15kHz. At this point, digikey seems to have run out of LM3915’s, so in about a week when they recieve some in stock I’ll put my order through (I can’t stand their extra $8 handling fee ontop of a second shipping charge for 7 backordered ICs).

I’ll also have schematics finished soon, however last night I threw together a quick opamp + LM3915 circuit just as proof of concept, nothing all too important, but its a start. The song is “Oh What A Night” by Frankie Valli & The 4 Seasons played off of a youtube clip, fed into the circuit.

Read the full post if you want the nitty-gritty electronics details.

While I do have a quad opamp, I only used it since it was the only opamp I had in my parts drawers, I could easily use 2 for a Fliege filter and the last opamp to complete a proper audio amp (ie: push-pull config) since the single opamp can only amplify positive voltage, and the oscilloscope clearly shows that the lower half of the waveform is lost. I suppose I could place the rectification stage prior to the opamp and avoid the use of an extra opamp, however this was just a fast mini-project. On the far right is the LM3915 chip with a 10 LED bar.

The cause for all the silly wires on the far right side of the breadboard is simply due to the need to limit the current of *all* the LEDs. The LM3915 does have its own internal voltage/current dividing resistor network so that the LED’s all draw the same amount of current, however in doing so the chip gets hot (the circuit is run at almost 20V!). All the LED anodes are tied together and connected to a 5W 270ohm resistor in order to limit the current to the bunch. I really should be using one resistor per LED (which I likely will in the final product) however I was simply out of similar resistors in the 1/4w range.

Anyway, my schematics are almost done and to my knowledge it might be one of the few public audio spectrum analyzer schematics out there once its done (unless my google-fu is weak and I couldn’t find any other similar project). Stay tuned and once I have more news I’ll make another post.

Advertisements

2 Responses to “LM3915 Based Spectrum Analyzer”

  1. Matt said

    Have you seen this project?
    http://tchips.com/home/index.php?option=com_content&task=view&id=35&Itemid=1

    He too uses Fliege filters for a 7-band analyzer. I’m attempting to build this, but instead of nixie tubes I’d like to use LM3915’s to drive the LED’s. This is quite similar to what you are attempting.

    Perhaps his schematics will help. I have very little experience with circuit design, so I hope you carry forward with your project. I’m very much looking forward to your next post, best of luck!

  2. aonomus said

    He seems to have used a bunch of comparators for his driver, summing the outputs. It might be cheaper to build it using discrete multi-comparator IC’s but it seems like it would be more error prone.

    I have however noticed some strange behavior with the LM3915’s, my circuit’s dot/bar pin would not function (I have 2 IC’s, 7 on backorder), and some LED functionality was unreliable, requiring the voltage to be up to 18-20V. The datasheet explains it as oscillation occuring, however trying to get rid of it seems to be very hard (plus I can’t seem to view it on my scope at all). I can’t seem to duplicate the oscillation after boosting voltage from 12 to 20V and adding the ‘recommended’ cap from the datasheet. I don’t have any tanatlums handy (well, I have SMD salvaged ones but those aren’t all too reliable/trustworthy), but the electrolytic solves some of the oscillation problems, as well as the undervolt. The oscillation/undervolt is characterized by the peak volume LED being lit the brightest with all the other lower signal LED’s dimming relative to which peak LED is selected.

    Its a scary coincidence that I’m using the LM358 for the fliege filter, and almost the exact same cap values. I might as well use his table since those resistors seem more available. I only picked a 7 band spectrum analyzer cause its the number of LM3915’s I could get under $20 CDN (pre-tax of course).

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

 
%d bloggers like this: