Active Overnight Sensations Anyone?

A buddy of mine has a pair of OS that he uses with the passive cross over, but he recently got a minidsp, so I offered to come up with an active cross over. He lent me his pair. I started by measuring the passive cross over.

Then measured the woofer:



Then the tweeter:



Sorry, kids were a sleep and the fridge was running.

I then did all the steps required to get Jeff Bagby's PCD running. My buddy only has the 2-way PEQ plug in, so complex cross over functions aren't possible right now. So PCD will do just fine.

First up, I threw this together quite quickly this morning. I'm hoping to get feedback here and tweek as I go.



I shot for LR6 at 3khz. Judging by the woofer response, there's a breakup around 4.5khz And the tweeter Fs is at 2khz. So sharp and in between I figured.

That cross over consists of:

Woofer
BW5 @ 2400
delay = 0.10ms
950hz, -4db, Q = 2
1700hz, -7db, Q = 1.5
4500hz, -5db, Q = 3

Tweeter
BW5 @ 3400hz
Level = -6db
5000hz, +2db, Q = 3
10,000, -2db, Q = 1
18,000, -5db, Q = 2

Well, that depends. I'd say it looks like it'll clearly perform better... on paper. But a minidsp is $100, plus you'd need 4x channels of amplification. Not exactly convenient on a mini MT. And a passive xo will cost a lot less. If you already have a minidsp and some amps kicking around, and wanna run these active, it's probably a good way to go.

I haven't measured the settings yet. Just modelled. The raw measurements indicate it should hold together well off axis. Also, the model is minimum phase and is showing good off axis performance. The green line is the modelled power response. Not bad. Wish I could push this tweeter lower, but fat chance of that I think.

The original OS was designed to be full BSC, so I'm leaning that way to hold to the original a little. In this case, I think flat is working out.

Some things I'm not sure about, is the very top of the top octave??? This particular set I have, has a bit of a router ring around the tweeter that may be causing some diffraction. Also, the baffle hasn't been rounded over yet. I dunno. Perhaps my mic has accuracy issues up there. I'm also unsure I have the relative sensitivies of the drivers correct. I measured at the same distance, but the woofer seems to sensitive compared to the spec sheet and 3rd party results. I'll find out when I measure the system hooked up to my minidsp. Hopefully this weekend.

Haha, I wondered why as well. Then I thought about the title. "Active Overnight Sensations Anyone". Kinda sounds dirty doesn't it

What's a Zappa angle?

What is the concern? The 2db dip in the power response at the xo? I don't see where you're coming from. That's hardly a sin.

If you'd like to discuss direct sound vs power response, shall we start a new thread? I'd take the position that direct sound is more important, within certain limitations. But which is more important is, as you mentioned previously, heavily dependant on room type, treatments, placement, listening axis, etc.

Not much can be done to improve that power response with those drivers. The woofer starts to beam (smoothly/gently thankfully) above 1.2khz. And the tweeter is flaring below about 5khz. Other driver limitations are driving the xo point into the 3khz region, so that's the power response we get, like it or not.

In this particular simulation, it has been optimized for listening directly on axis with the tweeter, at around 2.5m listening distance.

Something like this?



Even though I'd take the other sim, I appreciate feedback. Negative feedback is good feedback. So... which would you take?

Those are BW6 slopes. In phase. I could do 90d phase and see what that looks like. But the BW6 slopes are protecting the driver issues less than the LR6 slopes.

It's not nice yet, gotta prove it with measurements. Hopefully this weekend

I'd still take the first sim (without listening to them). A peak is more noticeable than a dip. And there's been studies showing on axis FR to be preferred flat above all else. I do believe power response is a close second. Combined with the added tweeter protection, the LR6 slopes get my vote.

I'll post the settings later, measure both, and let my buddy listen and decide. I'm betting it'll be tough to even tell the difference. One thing I'm please with, is on axis and power response should be much better than the original passive version which has to cross higher and softer to keep parts down.

Thanks for the feedback LTD. I'm debating starting a FR versus PR thread. Seems to be a point of contention around here in a few threads.

Following what LTD02 was suggesting, here is the same configuration with BW6 slopes around 3khz, but this time 90 degrees out of phase. This give a slowly falling power response and quite flat frequency response at the expense of the drivers being out of phase (audibility is arguable).

I still need to note that the BW6 slopes at 3khz may still be somewhat lacking to protect that tweeter. Listening would tell us whether or not it's a problem.

I plan to get to testing the results tonight or tomorrow. I've been very busy working out of town lately.



Not bad.

Oh and if anyone wants the frds or anything to give it a shot themselves, just let me know.

Here's another LR6 but now at 3500hz. If the tweeter is strained with the 3000hz cross over, I'll try this. Problem with this is it's getting close to the cone break up. Paul put his cross over at 4khz iirc, but his notes also indicate he drooped the midrange to hide this. The break up may not be bothersome, but it's darn close to the cross over. Listening is required to know if this one works.

I measured the first xo tonight. I was off on my scaling for the tweeter or something. Listening to them is fun. Very low sensitivity cute things though. I'll post more details and measurements later.

This is what I got using exactly the same settings from the first post EXCEPT that I did end up having my relative sensitivities off for some reason so the tweeter level was -13db not -6db.

On axis


30 degrees off axis (horizontal)


Reverse null


Sorry I didn't have the time to do more than that.

I listened for a short 2 hours this way. They were good. Low senstivity as expected. I didn't have enough time to really criticize them, so I'll leave that up to my friend. My testing has to stop there. I need time for other things, and they're back in his hands. I'm not sure anyone was planning on spending $100 converting these to active anyways I'll work on a few more settings for my friend for him to try, now that I know more about the driver interaction from a completed design.

From this, I would recommend these settings for now:

Woofer
BW5 @ 2400
delay = 0.10ms
950hz, -3db, Q = 2
1700hz, -6db, Q = 1.5
4500hz, -5db, Q = 3

Tweeter
BW5 @ 3400hz
Level = -13db
10,000, -2db, Q = 1
18,000, -5db, Q = 2

Those should perform a little better than whats posted above.