[CaptBubba]

My order came in today, and I've spent the last couple of hours soldering and troubleshooting, and am completly flummuxed. All my joints are good, and there are no solder bridges and everything is as it should be, but evidently it isn't. I made a few substitutions, the resistor in the feedback loop is a 4.7Kohm, and the one connecting the loop to ground is a 680 ohm, for a gain of about 8. Oh, dam, I just think I figured out what is wrong with it. I didn't put the cap between that resistor and ground, that would cause the whole signal to ground out, right? I feel dumb, but mostly tired, sleep first, or at least surf for a while, then sleep. I'll let everyone know if that was the dumb move that is causing the problem. I need to be more careful.

 

[CaptBubba]

I had "one of those days" yesterday. This morning I was looking at the soldering while I had an OPA134 in my hand to make sure all the pins where right, and was talking to myself. If you were there, this is what you would have heard: "Hmmm, there's pin #1, and then I flip it over and all the pins rever..ohmygodi'msuchanidiot." I soldered on the botom of the board, but like I was looking from above. That didn't work too well. I redid everything right, and it worked, perfectly. Thankfully I didn't kill the op-amp. Now that I know how to do it, I can do the other channel, and just maybe I'll get it right the first time the second time around.

 

[Apheared]

1) you screw up but it's easy to spot.
2) you screw up but it's not fatal to anything.
3) you screw up and it's fixed in a few minutes.

Other than that, it ain't funny.

One thing I wanna bring up, since we're talking about a Szekeres and single supply AC coupled gainstages... that cap you're talking about; the one that keeps AC grounded but lifts DC at the inverting input... that DC lifter coupled with that resistor forms a highpass just like the input section does. So you dropped it down, from 10k to 680 ohm, you're gonna want to recalculate what size cap to stick there if you want anything below 60Hz. I'd guess it would need to be 16-22uF @ 680 ohm. Here, let's check:

Cap = 2*Pi*680*10Hz = 42725.66 = 1M/42725.66 = 23.4uF

If you want to know where you'd end up leaving the 4.7uF with your 680... HEY let's do that one too:

Freq = 2*Pi*680*4.7uF = 20081.06 = 1M/20081.06 = 49.8Hz

Also, and this one I'm not sure why - with a low feedback you'll get alot more offset, but who cares as long as it doesn't oscillate cause we're AC coupled everywhere right? But get this: it'll sound less dynamic.. almost muddy... if the ratios get too far off. Might have something to do with the 4k7 to ground to force conduction, forming a divider with the feedback and being that low robs half the power? Again, I dunno... but if you up those values x10, to 6k8/47k, I bet you'll like the sound better. Someday I'll learn to backup my "because" statements with information other than just testing it myself.

Experimenting is easier than learning though.

 

[CaptBubba]

My resistor picking was based on what was closest at the moment. Here's how I PLAN to do it: 100K in the loop, and the following outside it. 10K (gain=11) 20K (2 seriesed 10Ks gain=6) and 50K (2 par. 100Ks gain=3). That gives me rolloffs (using 1uF cap) of: 15.9Hz, 8Hz, and 3Hz. I have lots of 100K and 10K resistors laying about, so that's why I'm choosing this method.

A cruddy headphone jack confused me for a bit, because I was geting sound out of both right and left drivers, and only had one side hooked up. Knowing I was about to either become very rich or find a fault somewhere, I started trying to find what had happened. I could give the jack a gentle tug and it would go back to the correct driver, so I'm figureing a short somewhere in the jack (there's a reason why its my cruddy testing jack.)

Ah, the joys

of DIY.