Project 7 - Pointy Heads, Seas P17RC - Alcone tweeter

This project was an experiment. I wanted to see how hard it is to make those pointy shaped speakers like the Sophia and such (I am tempted to call these the Sopapia). I also started feeling that my ACI Sapphires were a waste on the HT since I usually listen in the bedroom. I had a pair of the Seas P17RC on the shelf and some weird Alcone tweeters which look like Accuton copies. The Seas were a good candidate because in test boxes I could not get much bass out of them. Since this is HT, I only need to get down to about 80Hz. I also wanted to use a sealed box. The boxes turned out to be about 12 liters and the f3 about 90Hz so I hit pretty close.

What I learned:

1. Those pointy boxes are not that hard and kind of neat. I had planned on making all sorts of jigs to get the angles but found in test cuts if I just dial in the fence and go it works pretty well. Remember once you have the fence dialed in to make all the cuts because once you change the fence back if will never match the original alignment.

2. If you have a driver without much bass, build an HT speaker. In this case the Seas magnet is pretty small and the Alcone is shielded.

3. Sealed really is a little cleaner sounding. :)

4. Not all drivers that initially measure poorly are hard to work with. The Seas had a rising gain as it went above the mid range, pretty ugly. Once I had it in a sealed box with a simple XO, it cleaned up nicely.

The Alcone tweeters had a nasty spike between 15k and 20k. I decided to just put a low value inductor to bring the spike down (and the rest of the material between 15k and 20k). The other choice is to add a bunch of components in a notch filter. I am not very good with notch filters so I used an inductor. I think it worked pretty well.

First glue

Ready for trim cut
This is amid construction with the sides all hanging out and ready for the flush trim bit. Usually I only overlap 1/4" but with the slanty sides I over did it. The bottom board overhang is just a board it is sitting on and not part of the speaker.

ready for poly
They are stained and ready for poly. The discolorations in the wood are poor camera work and not the actual colors. I tested with Tung oil and went back to poly. The key to poly is using fresh cans and putting it on very thin.

in place
Finished product sitting on the ACI Sub-1. I completed the grills but don't have a pic. I plan on restaining the Sub-1s to match the tops. I never liked the color anyway.

The left side is closer to the correct colors. Our basement is being finished now and I hope to have these speakers a lot wider apart in a couple of months :)

Not a bad MLS, eh? Note I cut off the 10k to 20k section to hide the nasty part.

MLS rev
This is the MLS with the tweeter reversed. You can see the junk I had to do in the upper range to kill the spike. Note that my chart is set at 2db intervals so that dip is only 6db (only!!).

raw Seas P17RC
Just for fun, this is the Seas in the box with no XO. That bump at 125-280Hz went away when I moved it away from a wall. I did not save the repeat measure. But still a tough looking MLS, not flat at all.

These are the comments from DIY 2002 in Dayton. I knew going into this the tweeters need a little more padding. Also note that many people liked the bass and these have an f3 in the 85Hz range.
Very clean and smooth
A bit bright
Good clean bass for a small sealed box
Just a little hot on highs
Good transient performance


(9/6/02) I have not listened to it enough yet to give a listener's evaluation and I still have an issue to work out. The subs have a huge set of caps to roll off the upper speaker so the sub can fill in the bass. They were designed for a different speaker than the pointy heads. So I think the overlap is not right. There is a little boomyness at about 90Hz (I think). It is not there if I run the tops alone. I either have to change the caps or switch to the HT bass management. The caps are inside the sub so I have to find a control method.

This is interesting because it really drives home how difficult it is to measure the low end. I am trying to find a hump between two drivers in the 90Hz range. They appear to measure well separately but together the sound is not right. It is impossible to check them together near field. And the fact that the sub is down firing makes it worse. Some of my measurements are getting a null at 90Hz, suggesting phase. But not all measurements. I flipped it over for the last try. I think the next step is to take the whole rig outside and try again. Other than a little rumble the tops sound nice and clear. I have not played much music yet. The movies sound great, better than my good music speakers I had on top of them before.

Also, I have been thinking about the sloped baffle. While the majority of talk is on time alignment, I think the simple fact that we are off axis is more important. If I tune the xo for a flat MLS at 15o off axis vertical (the slope of the Pointy Head), then the variance in a horizontal movement may be less noticeable. Since the shape of a speaker cone is in essence a horn, the most variation in sound will be in the first few degrees of axis. By sloping the baffle we move out of that narrow sweet spot and into a larger sweet spot. Since the XO is tuned for the off axis it sounds right. I am probably the last one to learn this.

speaker icon

Copyright: Peter J. Smith 2005 Return to