PAP Trio 15 and DIY - Discussion

For the AMT variant, PAP does a funky 2.5 way design. They use series xovers.

They run the bottom woofer to 250Hz (electrical) which probably is more like 350Hz acoustical given the impedance curve. The top woofer is run with no XO but doing so requires the notch.

So dumb they actually gave their crossover a name. I want to remove their speaker from the Silver Balls award because of this.

 

This is too complicated for DIYers. People should buy the PAP Pass B5 and call it a day:

http://www.pureaudioproject.com/pap-c1-custom-active-analog-crossover-open-baffle/

Adjust the crossover point yourself unlike the DIY B4/LX-mini.

 

No. "Phase problems" have traditionally been overstated.

 

I've seen some pretty craptastic results from software EQ. Like not even a curve, but strange disjointed lines.

If you are running traditional speakers with exotic material cones that require complex crossovers, chances are that the phase response looks pretty messed up.

 

I just took a brief look at this, so take this with a grain of salt.

That being said, if the problem is peaking, I agree with @purr1n in using a Zobel:

First order only:



First order + Zobel (8 ohm + 8 uF) - still a bit of a peak, but the roll off is now steeper:



First order + Zobel (8 ohm + 20 uF) - peak is now mitigated:



First order + Zobel (8 ohm + 33 uF) - peak mitigation is stronger, but note the response is starting to loose it's tilt which may not be desirable since this is a 2.5-way aided from 250 Hz down by the bottom woofer:



Hope this gives some ideas.

This is not a full simulation, so again, take it with a grain of salt.

 

OK. Attached are the zma (impedance) and frd (standard SPL frequency response) files for the W8-1808 and the OB-A15Neo.

 

I have no idea about the Trio15TB crossover and much less the Leonidas (since that one hooks drivers in series). But playing around with the ideas, here is what I came up with for this Trio15TB using the TB driver and their Eminence driver:

All drivers in classic parallel arrangement.

Lower OB-A15Neo driver:



Upper OB-A15Neo driver:



Middle Tang Band W8-1808:



Impedance:



Comments:

Impedance reaches lowest value around 3 ohms. This is due to the 1.5 parallel configuration of the Eminence drivers, each at around 6 ohms impedance. This is a consideration when selecting amplifier. I would avoid large output impedance.

There is a notch filter, but I placed it in the TB driver. I did not like how it seemed to climb up w/o it.

I used only 1st order for the full range driver because it's just that: a full range driver. It may also help the 15" drivers avoid developing a suck out due to their size if a little off-axis. I tend not to like those large (8") full range drivers that much because they tend to be a bit directional in the treble. But if sufficiently away from the driver and if well positioned (things that will have to be done for driver integration anyway), this may not be a problem at all.

Note that for the upper woofer I did use @purr1n idea of using a Zobel to reduce a peak at 1 kHz or so.

Anywho. Hope this helps as an alternative if folks want to try some shit.

And yes, one may get about 96 dB SPL sensitivity with this. Depending on how these things are positioned, maybe one will get 29 Hz low end. Maybe.

I would not push these past 110 or so dB SPL in sub-bass (which is plenty stupid loud anyway), to avoid damage to the woofers due to driver displacement limitations. Maximum modeled SPL is 122 dB though.

EDIT: Note the W8-1808 full range driver is connected in negative polarity with the Eminence drivers. How the full range crosses around 950 Hz with the Eminence can get tricky. It should add.

 

Yes, alignment of drivers and polarity will affect the crossover summing region. Yes, simulation and measurement are recommended (stating the obvious).

My response was in respect to skem's question on LCR notch filters and how the phase changes will affect the sound. Maybe you can hear the phase issues of these. Maybe you can't. If you hear something, it will definitely be an FR change.

-10 points for offering your two cents on the wrong question
-20 for stating the obvious
-30 for proving how smart and also how dumb you are

 

I never understood the concern about phase issues with speakers (other than crossover). I mean, in room, phase just gave the ghost.

@Serious and I had a pretty hot combo some time ago about this. @Serious, please don't derail this thread with such concerns.

If someone has the Heil AMT 1 spec, let me know. With the stuff above I may provide some ideas with the more conventional full range drivers. But @crazychile has the Heil AMT 1, and the more it's known about it the better.

 

f**k!

I think I found it.

 

Ok. So here are the (approximate) zma and frd files for the ESS (assuming they are the right ones).

 

And here are some ideas (very similar to the ones using the TB driver), though the full range does change a bit:

Lower OB-A15Neo woofer (same as with the TB driver):



Upper OB-A15Neo (same as with the TB driver):



ESS Full range:



The handling of the ESS is different. A 4 ohm resistor was added to tame it down and the first order cut off was lowered. The notch filter no longer fixes climb up. It reduces a peak at 5 kHz inherent to the ESS driver. Without it, things look like this:



Which might be fine for some, but can be mitigated.

Impedance (with tweeter notch I would recommend given specs of the ESS):



Again, note the ESS driver is connected in negative polarity. It is possible that things will cancel out due to driver alignment. In such a case the model can correct for this. Worst case, connect in different polarity.

@Serious. Yes. We know exactly what happens when with driver phase, alignment and all that stuff @ crossover region. Which is why I said, use negative polarity for the ESS and TB driver. Believe it or not models can account for that. But I don't know exactly how the PAP kit is put together. Again, worst case, if suck outs develop, use the other polarity.

I feel your concern is likely unfounded though. That is because folks that have this kit have not complained about this issue so far. And the cutoffs and filter orders ideas suggested here are probably similar to the ones provided in the kit (except Leonidas).

 

OK. 1 week off.

EDIT: And yes, the tool can account for all that. In fact, if according to the model, if the crossover is hooked up in positive polarity one gets that horrid suck out:



EDIT 2: Yes, I also accounted for the baffle shape and size. And no, I did not just plopped drivers on the baffle. This is why you piss people off. Your know it all posturing, and assumptions that everyone else around you are stupid.

 

Expanding on this last approach, we can crank up the cap value and lower the resistor value in the Zobel.



Coil: 2.5mH
Zobel: 50uF and 5.1R

---

Another possible approach with the 2.5 way is to use a (simple no Zobel, but brute force) 2nd-order LP on the top woofer. The 1.2kHz peak will be pushed down, but there might be a plateau from 800-1kHz. Then maybe another LP at an even lower point to fill in the area below the peak.

I'd rather do not the notch at the high-frequency driver since it's the woofer making the errors. I'd rather not have the HF driver correct the woofer's non-linearity via summing.