Help requested - phase shift circuit

[mojozoom]

Hi guys. I'm hoping you can help me with a phase shifter project that is not a stompbox, but is in a box, so I'm halfway there right? It seems that this site has a ton of good practical people that can figure these sort of circuits out. It's actually a car audio project where I'm trying to alter the phase profile of the right mid, left mid, and subwoofer so that they all pretty much match slopes. Here's what it looks like:



And here's what it does:
1) Right channel gets an all-pass filter at 165 Hz with Q=5.
2) Left channel gets an all-pass filter at 500 Hz with Q=1.7.
3) Sub channel get an all-pass at 65 Hz with Q=1.5 followed by another all-pass at 86 Hz with Q=1.5.

I worked through the circuits in LTSpice and then breadboarded them and verified their operation on the scope. I tested the right channel breadboard in the car and it appeared ok (but it wasn't a very thorough test). After that I transferred everything to pc boards that mimic breadboard layout, installed those in a metal box, and tested it in the car.

I've got a few problems, the most glaring are associated with the right and left channel circuits. There's some pretty heavy distortion on substantially low and long loud bass notes (think dubstep), and there's some clipping going on in the upper ranges (maybe 1-3K Hz) on relatively clean tones. On the scope I've run about 3V sine waves from my laptop through the circuit from 50-10k Hz and don't see anything odd showing up. Generally it doesn't sound too bad, so if I could resolve the distortion issues it'd be a success I think.

Can you guys take a look at this circuit please and let me know if you see anything that might be causing problems? This is my first attempt at a project like this and the more I learn the more I find that I don't know. Thanks!

[robthequiet]

Hm. Since I can't plug a guitar into it it may be a bit out of my range, but one thing to look at is the narrowness of the 165Hz filter -- I would look for an unexpected resonance somewhere in the chain, possibly in your crossover or power amp. You may be getting some peaking from the high Q of the filter as the music can be mixed to emphasize and thus saturate those ranges.

You could make yourself a test recording and put it on usb or cd, or whatever you have to play music in the car, with test tones, like a room frequency analyzer. My tape player has an annoying ring at certain frequencies in the soprano vocal range, maybe you're getting something like that.

Hope this is useful, or at least gets the project moving towards success. GL