active crossover built, now what

[lmartine3]

I am starting to build a small collection of old but decent power amps so I am experimenting with tri-amping using Rod Eliot's P09 active crossovers. I have built a pair of crossovers and they work fine with test tones from the a signal generator, so now it is on to speaker surgery.

My target speakers (JBL L100t) are quite respectable as they are, soundwise, so it would not surprise me that my relatively uninformed tinkering will have a negative impact on sound, but remember, this is an experiment!

The active crossovers are fixed frequency (800Hz and 4500Hz), 24 db/octave and the speakers have the same crossover frequencies but only 12 db/octave rolloff.

So now it is time to integrate the crossovers and I am trying to figure out how to set the amplifier gains for each band.

Even if I could find sensitivity specs for the individual drivers, these numbers would not take into account the end result when integrated with the passive crossover would they?

I was thinking about running test tones through the existing unmodified speakers, measuring SPL with my ultra-cheap digital SPL meter, then matching the SPL with the active crossover. Would this approach work?

[c2cthomas]

JBL replacement parts for old speakers are hard to find and sometimes expensive. If this is an experiment for learning about active crossovers I would suggest that get some cheap flea-market (craigslist - flea-bay - garage sale - junkyard) 3 way's and experiment on those 1st.

[Pano]

I agree with Thomas, you want to be careful. It's the tweeter that is most at risk.
But anyhow...
Yes, your idea of measurement is a good one. You might also try it with pink noise so that you don't land on some odd frequency where there is a bump or hole. Sweeps can be useful, too.
And just to double check, try measuring the voltage at the speaker terminals.

With pink noise you can probably do it by ear comparing your active filters to the other speaker still passive. Start with the amp volumes at zero and bring them up slowly. It won't be hard.

Let us know how it works out!

[cbdb]

The problem with SPL measurements is the room interaction, this is usually done in an anechoic chamber.You could also run tones into the speakers before going active and measure the levels at the drivers (after the passive xovers) and then try to match these levels when you switch to active.

[lmartine3]

Thanks all for the tips. I will definitely have some caps in line for tweeter protection.

I am always looking for used stuff on the local market and have found some good electronics locally. Speakers I have not been real lucky with, although I have checked a lot of them out. Ttrying to find the combination of something that is cheap, half-ways listenable, can handle some power and does not have foam rot and/or crispy tweeters and/or a disgusting aura of cigarette smoke clinging to it can be tricky.

[c2cthomas]

Ya can make sum cheap
Parts-Express.com:*JAMO 20388 15" Paper Cone Woofer | JAMO 20388 15" Woofer Treated Paper Cone driver speaker bass $25.00 ea
Parts-Express.com:*Peerless 830872 4" HDS Nomex Shielded Woofer | Peerless P830872 HDS Woofer bass mid midbass mtm poly cone nomex cone vifa Tymphany cast frame shielded $29.00 ea
Parts-Express.com:*JAMO 20301 1-1/8" Soft Dome Shielded Tweeter | JAMO 20301 1-1/8" Dome Tweeter waveguide hf high frequency mtm ferrofluid shielded JamoBuyouts111309 $8.00 ea
Free shipping for over $50.00
Get two sets if ya wanna build a pair (recommended - these are buy out parts and may not be around if ya want more of 'em)

R-Tech at Lowe's: 1" x 8' x 4' Insulated Sheathing $13.00 ea Enough here for 2 OB baffles.

Some 2in. x 4in. lumber to cut up for a frame

Some chain - rope - parachute cord - to hang speakers from frame per Stig Erik thread “My open baffle dipole with Beyma TPL-150 My open baffle dipole with Beyma TPL-150 post # 371

PLUS – U get to fool around with Open Baffle design!!

[lmartine3]

I am using the pink noise output from Visual Analyzer as a source to an amplifier and then probing the crossover outputs.

It is pretty hard to get a consistent level measurement with a DVM or a scope, is there a better way to measure the output? I might go with single tones just to get consistent numbers.

A side issue is that 1V into an efficient 8ohm speaker is pretty freaking loud when you are sitting beside the thing on the workbench. You really need everyone out of the house and ear protection!

[Pano]

So you have found out how loud 1 volt is! Good.
If you have an AC voltmeter with lower scales, you can run a lower tone. Many meters have only high voltage AC ranges, like 200 & 600V.

It's going to be hard to get a steady voltage reading with pink noise. But if you do have an SPL meter, set it to "slow" and try that. And how about comparing it to the passive version by ear?

[c2cthomas]

Heh-heh Yeah 1 volt can be pretty loud - depending on the driver - but the real fun starts at 2.83 volts.

Zaph measures a bunch of speaker drivers so I suggest you follow his lead on how he goes about doing it. Zaph|Audio

Many of us here have a lot of respect for his work and his findings.

I can't think of a better place to get your feet wet than visiting his site.

[lmartine3]

I am learning a lot. Perhaps the biggest thing is that proper loudspeaker measurements are not easy! There is interesting stuff on the zaphaudio site for sure.

Without doing any measurements, you have a basic understanding of things like crossover slope and driver sensitivity. Real understanding only starts with real (even just half-*ssed in my case) measurements on the bench.

There are so many things going on with loudspeakers that make an audible difference it is mind-boggling. Forget about 16g zip cord vs. braided cat6 speaker wires; simply moving your head will cause the sound to change!

I was playing a 400Hz test tone with the speaker pointing away from me into a corner, then leaned over the workbench to read the meter. The sound level to my ears changed so much I thought something got disconnected, but no, it was just my small movement relative to the speaker which changed my perception.

How do PA installations handle equalization with multi-amp setups, does the sound guy/gal do it by ear?

Pics with the crossover and bench setup attached.