First DIY speaker project thread!

I have been wanting to do this for years (literally), but every time I started to get things together my plans changed (time, money, etc.). However, I am finally going to build something- no excuses.

I remember you from CAF home audio section, ~1.5 years ago. You wanted to build an NSB line array, then transmission line speaker, etc.

Don't integrate the subs into mains. This cripples your options later if you need to
relocate the subs to another part of the room for best sound. Plus, subs are usually
the first upgrade itch. Your mains may be nice for years but you may drool for more
bass. It easier to keep the sub seperate. Receivers should allow high passing the mains,
look at the receivers you plan to use and see what options they have. That would be
the ideal situation vs. making a passive high pass circuit.

If you want to build something relatively inexpensive but still quality, why not Zaph's BAMTM?

http://www.zaphaudio.com/BAMTM.html

Well, Zaph's uses higher quality drivers (and they obviously cost more), but I think it'd be worth it.

I will sometimes shop from both if I can save a bit of $$$, but you'll be just fine with either PE or Madisound.

Zaph's BAMTM has a much better crossover network. I've heard a good number of people who like them, and I'll probably be building some for a friend soon. I also agree with you on the tweeter. The Seas tweeter will out perform the DC28 by a fair margin.

Wait until you build the rears, because the center will be otherwise useless.

Looks like you already decided to build Zaph's BAMTM. But you still consider his WG TMM? I think the Seas P18RNX poly cone woofers, a pair of which you said you already have, can be a drop-in replacement for the CA18RNX in the design. Their frequency response, impedance curve, and TS parameters are all very, very close to each other. VAS is a bit different, but you can tweak the design's bass response using different tuning options. I don't see any reason you can't use P18RNX instead of CA18RNX for a good result.

I think BAMTM is also a good choice.

It's probably a steel core. Check the sledgehammers at madisound.

That would be my guess. Doesn't Zaph post what kind of inductors he's using and where? For the large values he *may* be using the steel core, then air core for everything else. For those I would look at the Jantzen coils from PE. The Dayton 5% caps are great fort he price also.

Ok I read Zaph's project page. The 4.0 inuctor is a sledgehammer, all the rest are air core. I would use Jantzen from PE of whatever guage gets the resistance right. PE's site is kinda junky, when you look at Jantzen inductors don't use th general inductors list, instead click on Jantzen or select it from the brand menu. If you use the general list if you click on say 2.0mh 20 guage it always takes you to the 2.0mh 18 guage page. It's been like that forever I can't believe they haven't fixed it.

I would lean towards the perfectionist XO. Better offaxis response and phase integration means there is nothing funky going on the power response around the XO. Just my .02

Although the P18 is pretty close I don't think I'd use it unless Zaph said it was possible. If the SPL were fairly flat on both for an octave or so after the XO it would be possible i think. But they both roll off fairly quickly after teh XO with P18 starting 1000 Hz sooner than the CA18. With 2nd order slopes this will cause a more pronounced difference. It may not be that bad witht the minimalist XO but...I would ebay the P18 and do it right. Then again you could build the enclosures, take actual measurments with SW adn see if that XO will work. Could be a good learning experience.

For the inductors, just try to get the resistance as close as possible by using various gauge wire. With the 14 gauge Perfect Lay inductor you can get about .62 Ohms, which is probably close enough, since I doubt .4 ohms will make that huge of a difference. For capacitors the Polypropylene ones will do well.

Brandon, Nice to see you here!

Yes, you're right. Zaph's approval would be a definite answer. But the P18RNX and the CA18RNX are really close to each other. As for your concern, I looked at their FR again. But I think it's the other way around. The P18 can be used with the Perfectionist XO without a problem, but perhaps not with the Minimalist XO. BTW, even the Perfectionist XO is not that complex---it's still very simple, compared to, say, the Modualr MTM!!

Here are the CA18 and P18's FR charts with the CA18's on top:





Although P18's response seems to roll off a bit earlier than CA18's, their FRs are damn close up to 6 kHz. Next look at the woofer and tweeter's individual responses with the Perfectionist XO:




When two overlapped responses have a difference larger than 20 dB at a frequency point, a combined response is larger than the louder one by no less than 1 dB, even if the two responses are perfectly in phase. If you look at the above individual drivers' responses, the 20 dB difference occurs at around 5 kHz, at which the tweeter's SPL already hits the targeted 90 dB SPL. This means if we use the P18 instead of CA18 with this XO, the only difference will be a very small dip (perhaps < .5 dB) between 4 kHz and 5 kHz, which should be negligible.

But for the Minimalist XO, we have:



XO point occurs at around 3 kHz and we have a 20 dB difference between woofer and tweeter SPLs at as high as 9 kHz. This means that there will SOme difference in combined FR in the 4 kHz ~ 8 kHz range. But then again this is Minimalist, anyhow, and perhaps the difference may not be so audible. But I'd prefer the Perfectionist XO, which is simple enough.

Also, another minor difference between CA18 and P18 is that P18's bass and midbass are a bit weaker than the CA18. But this design is 2.5 way and by nature has full baffle step compensation. The speaker will still have more than enough bass/midbass as a 6.5" woofer based floorstander. Tha VAS difference (21 liters vs 36 liters) may need a smaller box volume, which can be implemented by shortening the cabinet depth.

One more thing to add is that looking at the nonlinear distortion plots, as Zaph pointed out before, the P18 has one of the cleanest bass in the group--even better than ScanSpeak 18W8531G!

So, I'd have no problem using the P18 in place of the CA18 for a very close (and, of course, good) result.

What do you think?

Hi Jay. You make a good case. SPLTrace it and model it man Looking a little closer at the SPL and Z plots you probably could swap them. Acoustic center should be the same on both since they seem to use teh same basket/motor so that isn't a concern. I do wonder about the Z though. It only seems slightly different but the scale is 10 Ohms so I wonder...Probably not a problem with the minimalist XO though, or even the perfectionist since it's not very complicated.

Still though the CA18 is known for it's good bass. Depending on whether a sub is involved and the XO frequency this may be important to the OP. I dunno...I agree this is one of the very rare times you could swap woofers, I'm just a little conservative about swapping drivers into the existing network of a decent design. Small, unforseen differences can really jack up the results.

I think this is a good opportunity for the OP to use that SW and measure and model and see what the exact differences are! As you know we can all debate for hours on this stuff. I'm skeptical of everything until I've measured it for myself. One day in the lab is worth thirty in the classroom as they say...

Eh the OP doesn't have SW...that was another thread I posted in it's getting late I should sleep

Also I'm not so sure about the Z differences effect on the minimalist XO now...it might actually be less of an effect on the perfectionist. Either way I would still build the perfectionist. Better results from just a little more money.

I just looked at manufacturer's spec sheets of both drivers. It seems that some of our observations based on Zaph's measurements are simply unit to unit variations. Here're they are:

http://www.seas.no/Prestige%20Basser...X-H1215-08.pdf
http://www.madisound.com/catalog/PDF...P-H1350-08.pdf

Some differences we were concerned about are reversed here: the bass / lower midrange sensitivity is now lower for the CA18 than for the P18, and impedance curve is a bit lower for the CA18 than for the P18. Also, on Zaph's measurement, the voice coil inductance of P18 is somewhat lower than that of CA18; but on the Seas' table, they're virtually the same.

I think we can simply view that these measured differences between the drivers are just sampling variations. They are even closer than I thought.

I figured they used the same motor so I wasn't sure what to make of the Z differences. It would be prudent to do an actual measurement and model with that to be sure, but if they are indeed the same you could probably be ok without doing that.

Yeah, they indeed share the same voice coil height and diameter, cone area, xmax, and basket. I think their acoustic centers will be virtually at the same depth. So I don't think there'll be any problem for good phase tracking.