Upgrading some stage floor monitors. Looking for some input...

[nodestiny]

I've been using these Peavey PV 12M 12-inch floor monitors for my band for years. Over time, tweeters started to go as we constantly had to crank our Crown XLS 2500's to the limit to get them loud enough, especially for vocal output. Decided to finally attempt a real upgrade on these older cabinets.

I went with the Eminence 3012HO for it's sensitivity, low weight, and overall decent response curve for the primary driver. For the horn, I went with the Selenium D220ti for the price/performance and seemingly good fit for the 3012HO output.

For now, I'll be reusing the Peavey PV 12M cabinets, with a full rebuild using 9-ply 12mm baltic birch plywood and proper space and porting in the future. Attached are the driver's raw output with simple swapout of the drivers. So, 3012HO in unstuffed enclosure and D220Ti on the factory horn wave guide. I placed the monitor on the floor in a mostly treated room (since this is where they'll be used, figured it made the most sense). I measured these with the DATS v2.0 and the EMM-6 with the mic at standing ear level, dead on. Took this data and exported it for XSIM and designed the crossover, aiming for the fewest components possible and measured again to verify (with ~1lb of poly stuffing in the enclosure this time). After a single practice of use, wow. The output is considerably higher to the point we had to turn the amps gains down several clicks to make them tolerable! The details on the output were extremely good in comparison to the stock monitor and the mix never sounded better. Vocalists who tried the monitor were able to hear themselves very well, finally. After an hour of constant use, I pulled the crossover out to feel the components for heat issues and everything was near or barely above room temp. The woofer's magnet was the only thing that seemed to have any heat load on it.

There's a few things I was curious on, however...
-Is there any real issue with having such a large impedence spike from the crossover at the 2.1Khz area?
-I've played with the modeling to try and tame some of the 1.5-5K area, but again, looking for minimal components. Any suggestions?

Thanks for any input and eyes on this!

[Billet]

It all looks good to me, nice job! The impedance peak at 2.1 KHz is a natural result of the low pass and high pass filters converging, I see that a lot. I like the minimal crossover idea also, if you are happy with the sound, I'd leave it as is. If it seems to forward in the mids, then I would probably try rolling off the low end of the tweeter a little faster. A smaller cap and/or smaller inductor may help.

[billfitzmaurice]

I never bother to do any correction with pro-sound gear, as a full 31 band EQ isn't optional, it's mandatory. I also never use less than 3rd order filters on HF drivers, especially when crossed over fairly low. More often than not I use 4th order.

[fpitas]

Bill has a good point about a higher order crossover on the horn. The D220 is a nice driver for the price, but you would be wise to keep LF stuff out of it for longevity. You just know that somebody is going to crank it wayyy up

[jhollander]

There's a few things I was curious on, however...
-Is there any real issue with having such a large impedence spike from the crossover at the 2.1Khz area?
-I've played with the modeling to try and tame some of the 1.5-5K area, but again, looking for minimal components. Any suggestions?

Thanks for any input and eyes on this!

No issues with the impedance spike. As you can measure and assuming you don't move the mic, you should take a T+W measurement and find the offset between the drivers. The offset will help you design an x-o that would help tame the 1.5K to 5K range.

[billfitzmaurice]

Or just pull down the sliders in that range. 31 band EQ is SOP in pro-sound, because the acoustics are different in every room, and it's the only way to pull down individual feedback frequencies that are different on every stage. You also aren't necessarily looking to EQ floor monitors for the best response but for the best cutting through the mix with vocals. This is second nature to experienced pro-sound engineers, while seldom seen in home systems.

[Brian Steele]

The impedance curve suggests that those old boxes might produce better results if braced and damped properly.

Passive x-overs should be used to match the output of the various drivers through the x-over region and deal with any gross response issues (in this case, I'd opt to redesign that x-over to deal with the upper midrange bloom, but first I'd try to take some better measurements as that FR graph suggests that there's a bit of the room in that response curve). Bill's spot-on with the use of EQ of course, but I like to get most things right with the response first before having to resort to EQ to deal with the rest.

[nodestiny]

All good info, thank you folks for that.

I'll definitely do more experimentation in the near future. I did some measurements at the time on floor VS on stand and you can definitely see a big difference between the two for the woofer ~500hz and lower (attached). My concentration isn't on the low end for the crossover design, so I didn't spend much time here. Note: Our isolated live room has some decent acoustic treatment done, so it isn't the worst measuring environment I own.

I'll play around with the crossover some more for optimal response and look into a higher order for the tweets. I could do a 1st/3rd order or 2nd/4th for woofer/tweeter to save on cost, since the woofer wouldn't be any more protected from the lowpass. Though, I'd like to keep the filter stronger than not around 1.5-2K on that woofer.

Cabinet design in the future should address any resonance and just bad tuning issues. It's around a cubic foot with a single 2.75" hole in the 1/2" MDF... Once I get my 3D printer going again, I'll attempt to at least make a bigger port with some length to get the tuning down closer to ~70hz, but ultimately, my nice 12mm ply is calling my name! Constant 100*F+ weather and no AC in my shop says I'm not in that big of a hurry, though.

(BTW, any graphs I post are not calibrated decibel wise. I didn't measure amp output, but ran them with some decent power to more closely emulate what they'll do in production)