[MaximumEnnui]

regardless of how well you think you installed these tweeters, and set the gains and crossovers, they are blowing for 1 reason, too much heat on the voice coil. Raise the XO point, increase the slope, and turn the gain down.

especially since they're running capacitors.

 

[MaximumEnnui]

Since you're running caps (which is good) which eliminates the chance that low frequencies are somehow hitting the tweets the only option is that the tweeters are getting too much power. Since you've got a lot more amplifier power than what tweeters should be using try setting gains for midbass drivers and then level matching the other drivers using an octave of pink noise in their pass band or some other method to ensure you aren't over driving them.

 

[MaximumEnnui]

You must have missed the part where I checked them with a scope at the setting they were at, about 1/4 up from minimum, and the first time any signs of distortion or clipping showed up was at or a little above 3/4 to max level. I have set up every system that I’ve done for myself or someone else the same way unless the customer specifically wants something set different, but I normally set them back far enough that the volume can be maxed out with useable sq still. Some guys like to get to 75%, I’ve done several for a guy who I know already that he wants his to be maxed at 1/2 volume. He is the one I’ve been trying for 10 years to understand that lower settings sounds better and cranking up the gains don’t add power.

If you are interested in trying to get it to stop: use a multi meter at the output of the amp to make sure the output voltage of a 0db tone is low enough the amp CANT overdrive the tweeter. When you set the gain using an (edit) oscilloscope what was the output voltage of the tone you were using? Was it playing at 0db? And what frequency was that at?

 

[MaximumEnnui]

Related, but I have been wondering this as I explore fail safes in my system. Does have two filters like this create any phase issues?

As long as they're outside of the passband of the crossover: not really. Having cascaded filters as close to one another as OP does will change the acoustic slope of the crossover which will result in improper summing. You primary fail safe should be a capacitor higher than the resonant frequency of the tweeters. 1.5-2x will offer more protection in case you don't remember to turn an XO on. A cap will also protect against an amplifier fault which could send DC to a tweeter and kill it.

 

[MaximumEnnui]

Got it, thanks. I just installed caps in my car after a losing a set of tweeters due to a failed DSP. Was wondering if that was my only means of protection.

It should be sufficient. If you want an additional fail safe set the XO on your amp close to where the caps are as well. As long as the rolloff for these is far away (close to an octave) from the actual crossover point it shouldn't have an audible effect.

 

[MaximumEnnui]

So you’re saying that if the DSP hp is set at 3600, having the amp hp set as close as 500-750hz lower would change the slope in a negative way? But the 2.2khz caps I added would be fine being the fs of the tweeters is 900hz? So no amp crossover, turn it to full range?

Yes. Absolutely. From a SQ perspective the easiest way get symmetric crossovers is to use LR24 on a dsp and then EQ down to target. The 6 db/octave cap at 2.2khz would have some effect on a crossover of 3600 but would be 100 percent workable contrary to what anyone tells you.

If you are running amp crossovers in this system as well as dsp crossovers AND caps i would eliminate the amp crossovers just to eliminate a variable that might be causing you to clip gains or some other type of shenanigans. Idk there's been a lot of helpful info posted here i imagine between all the feedback you could test and report back what was going on.

 

[MaximumEnnui]

Without o-scope readings, eq adjustments, headunit settings, etc we're just guessing.

OP doesn't seem like he's going to provide any of that. Idk i have a distaste for taking the time to troubleshoot when it's taken in bad faith.