I'm working on a speaker system that uses this crossover and am having issues getting the mid or highs to work. I'm circumventing all previous circuitry and only feeding into the two crossover that feed a mixer and 5 amps. The amps work fine with test signals.

The lows work all the time. The mids work a little noisy if I reinsert the path after I have started the circuitry. If I turn it all on, I get a noisy static thumping noise from the speakers from very loud noise. This noise sometimes pops in during normal operation.

I'm actually about to go postal from this. I've spent several all nighters working on this project and it is problem after problem.

 

Are you using TL07x opamps?

What are you using for the power supplies? You're using dual rails, correct?

 

TL074ACD for the opamps, +-12V rails. Current from the power supply shoots up from ~250mA to 1A on each rail (limits) when any speaker but the lows are added in. It also sounds very painful when this happens, like a record scratching. I just tried an ammeter on the resistor I put in but it doesn't look like any of the current is going down that path. At least no signifigant DC current.

 

And you're feeding three separate amplifiers from this crossover, right?

One thing I notice is that the outputs are DC coupled, not AC coupled. I would put decoupling caps between the crossover outputs and the amp inputs.

 

There is a decoupling cap on the amps page that is 1u. They actually feed 2.5 amps. We have 2 of these crossovers and the lows are fed into a mixer which then feeds into the amps.

I've discovered the highs don't complain too much except during switching. The mids are the real screamers. I'd prefer it to perform more reliably.

 

How do you have this circuit wired?

Can you take some photos and post them?

 

This is the setup. I apologize for the lab light glare.

The orange wire coming is a signal from a function generator.

It seems we have a noise problem. We still can't connect the mids due to them making that awful noise. It seems the highs only do it on connect and shut down.

 

Did you use bypass caps on the supplies? I don't see any caps where the power is coming on the board? I can't tell if you used bypass caps on the opamp, because I'm not sure where it is. I can't read the IC part numbers, and you have a lot more on the board than just the one circuit.

Did you use the bottom as a ground plane? I can't be sure, but I don't think so.

Do you have power rail planes in different board layers, or are they mixed in with the traces running around the top of the board?

I'm wondering why you have a USB cable coming on to a sound mixer board? That's sort of like inviting a punk band to play in a library.

 

There are caps on the chips but none on the power supplies entry. The two 14 pin ICs in L-shapes are the 3-way crossovers (They have a weird symmetrical thing going on).

The bottom is a ground plane except for on the amp side we have power rails running under and a few other places have small lines where feedback paths or other various circuit paths had to go under. The minimum trace width is typically 15 mil and 50 mil is used for the power traces.

The school budget is only enough for a 2 layer board so see the above comments regarding signals and power lines.

The USB cable is because the system has an iPod dock and a 3.5mm input. The dock charges an iPod through USB. The USB circuit is the one that Ladyada uses in her updated mintyboost I think.