2 way active crossover development + troubleshooting

[tanno1]

Hello all, I have been working on an active crossover for the past year and would like to share some progress/gather feedback on my design, possible issues and just talk about it in general. I have attached a few videos from some testing yesterday to show some good and bad things going on with the project. The crossover frequency is 1.8khz.

1. I finished the design after a ton of research, made it into a PCB, bought some from PCBWay (tariffs suck), and then tested them
2. Two op amps had their V+ and V- swapped, oops, so I manually soldered those pins to the correct power pins and got it running
3. The high pass works, the low pass does not, see the attached videos

The V1 (kind of working) PCB:

I am happy to talk about the design, so if anyone is interested, let me know what questions you have. If you have any feedback on my design or can help me identify the error in the low pass filter, that would be helpful as well.

High pass test:

https://drive.google.com/file/d/1rj1S-afhHmbwzez_7ym5usaZm5qQaGgQ/view?usp=drive_link

Low pass oscilliscope:

https://drive.google.com/file/d/1ubzzHtfHHwwOWftyQeR9GNAvBl6zsfmt/view?usp=sharing

High pass oscilliscope:

https://drive.google.com/file/d/17VlyldYoGHkIw46RKg0XvDrnY1Rdxibn/view?usp=sharing

PCB schematic (with U2 and U1 incorrectly powered), attached!

Thanks @Busdriver02 @AnalogSteph for some good advice at the beginning of the project, I appreciate it and have made some great progress since then. Also just an FYI, I studied mechanical engineering so all of this electronic stuff is self taught and I may be saying things incorrectly in the videos. Feedback on that is appreciated as well, I am doing this to learn.

Thanks

 

[tanno1]

+ updated the sharing perms on google drive

 

[DVDdoug]

It's been too long since I've studied filters... And I was blocked from the videos, which I can't really watch right now anyway because I'm at work.*

The feedback, or what looks like feedback, to the + input on the op-amps seems strange to me, but like I said it's been a long time and I've never actually made an active filter.

The low pass does not,

Did you check the output of each stage? All of the op-amps should be resting at 0VDC, and then of course there should be a signal from the output of each stage if you feed-in something less than the crossover frequency.


* OK I got a quick look at one of them without the sound turned-on... Is something oscillating? Of course, that CAN happen with positive feedback, but I could be on the wrong track with that...

 

[tanno1]

It's been too long since I've studied filters... And I was blocked from the videos, which I can't really watch right now anyway because I'm at work.*

The feedback, or what looks like feedback, to the + input on the op-amps seems strange to me, but like I said it's been a long time and I've never actually made an active filter.


Did you check the output of each stage? All of the op-amps should be resting at 0VDC, and then of course there should be a signal from the output of each stage if you feed-in something less than the crossover frequency.


* OK I got a quick look at one of them without the sound turned-on... Is something oscillating? Of course, that CAN happen with positive feedback, but I could be on the wrong track with that...

Could you point out which feedback you are talking about? I am not sure what you are referencing. The filters are Linkwitz riley filters.

I have not checked the output of the resting low pass op-amp. I will do that and let you know what the result is. I really am not sure what was going on, which video are you thinking had oscillation?

Thanks for checking this stuff out @DVDdoug!

 

[Davey]

Your topology looks....okay.
Although I would have located the shelving filter downstream (to the right) of the low-pass filters.
Also, you should have a series resistor on the two outputs to decouple capacitive loads and give you some protection against accidental shorts.

I'm not sure what your issue is with the low-pass. Maybe a bad connection?

Do you have enough local bypassing?

 

[tanno1]

Your topology looks....okay.
Although I would have located the shelving filter downstream (to the right) of the low-pass filters.
Also, you should have a series resistor on the two outputs to decouple capacitive loads and give you some protection against accidental shorts.

I'm not sure what your issue is with the low-pass. Maybe a bad connection?

Do you have enough local bypassing?

When I go back to the lab, ill check the op amps. check the connections and maybe tack on some caps to see if that changes it. Good point though. I desoldered the V+ and V- pins from the board and then took a breakout board, soldered a wire from the input power jack V+ and V- into the breakout board, and then connected the op amp powers directly to the breakout board, so there are no caps on the power supplies to either op amp.

On the PCB, I think my capacitors are close enough to the power inputs of my U1, but it looks like U2 doesn't have a decoupling cap near the V- input, or the V+ input actually. Either way, they were powered from the breadboard during these tests.

Can you explain why you would want to put the shelving filter to the right of the LPF? What difference will it make?

 

[Davey]

With large boosts it's possible to create clipping by over-driving the LP filter. If locating the shelving filter downstream you won't have that issue.
(It depends where your volume control is.)
Also, you should add a 220pF capacitor in the shelving filter feedback loop to reduce the gain to zero above the audio band.

I probably wouldn't have made the boards that small. Easier to work on if they're bigger.......and easier to see if you're an old guy like me.

Keep fiddling and I'm sure you'll find the issue.

 

[tanno1]

With large boosts it's possible to create clipping by over-driving the LP filter. If locating the shelving filter downstream you won't have that issue.
(It depends where your volume control is.)
Also, you should add a 220pF capacitor in the shelving filter feedback loop to reduce the gain to zero above the audio band.

I probably wouldn't have made the boards that small. Easier to work on if they're bigger.......and easier to see if you're an old guy like me.

Keep fiddling and I'm sure you'll find the issue.

Gotcha, good advice. That is another question I had for this forum: How should I handle volume control. I have a 1/4 & XLR audio input , and I was thinking of just taking a 10k ohm potentiometer and putting it between the 1/4 & XLR audio input and the RCA input to the crossover.

I will keep fiddling with the PCB for a bit, and then If I can't solve it here, I am going to breadboard the circuit and make that work before I design another PCB that could possible be a brick. Would you be willing to take a look at my future design for some feedback? It would be great to have you review it.