Site Links
Howdy, Stranger!
It looks like you're new here. If you want to get involved, click one of these buttons!
Quick Links
Categories
In this Discussion
Please review the site Rules, Terms of Service, and Privacy Policy at your convenience. Rules, TOS, Privacy
Get familiar with the reaction system: Introducing the Reaction System
Coaxial shenanigans
Had no plans for these GRS 8PT-8s after they got swapped out in a build. 2" voice coil and vented pole piece, so plenty of room for a tweeter and wires. hmm.
Yup they weren't lying about the copper sleeve... on a $24 8" woofer. If not a perfect FR, it's got "good bones" at least.
First time removing a dust cap. I tried constantly applying acetone for about 10min then lifted the dust cap with a box cutter blade. It still peeled off the very top thin layer of paper. Doesn't appear to be destroyed so success...ish?
Thinking about picking up some ND28Fs tomorrow and shave down the flange to fit. If nothing else a cheap experiment.
Comments
This is awesome! I would consider a waveguide to help the low end.
This should really beg the question why copper isn't standard issue at this point.
Bless your soul to align all of this magic Drew , I hope it pans out !
Current contenders are Dayton ND28 and Lavoce TN131.00
The Lavoce has an overall OD of 1.89" which might be a perfect fit. Would just need to fashion some sort of spacer to flush the tweeter with the edge of the former.
IDK starting to lean the Lavoce direction because it might be easier to mount with least modification. But the Dayton is a bit more tried and true. Which should I try?
Or the Peerless XT25SC40-04.. It is 1.73" dia and almost as deep as the available space. But I don't know if that one would be able to dig down in frequency enough to be satisfactory.
I think that Peerless is going to XO nicely to that concentrically ribbed cone profile at around 2 ~ 2.5KHz and the beaming of both will match pretty well based upon the specs/factory response curves, Drew.
I love coaxial MT's so I am watching . . .
If I remember right the ring radiators tend to have a gradual upward slope to the distortion curve so by ~3khz it is going past 1%. The FR of the 8PT looks like it would be pretty dicy crossing up there.
I think 2.7k was the lowest I could take the XT25SC40/50/60/90. They are all basically the same tweeter. The Fs has to be compensated to make this work though, whether full LCR or R across it.
InDIYana Event Website
That step at the gasket ring will also do bad things to your response.
InDIYana Event Website
Yeah if flush mounting I’d remove front foam gasket. If rear mounting I’d put a roundover on through-hole cutout.
I may end up reusing some cabinets that already have a hole cut for an 8" driver (from the '23 indiyana theme). The recess cut in the baffle is probably a bit larger than the 8PT dia. So tinkering with an idea of having a second baffle that would be pressed up against the gasket.. essentially like a grill board that hugs the driver diameter closely with a chamfer or roundover. But first 'gotta see if I can even get these tweeters mounted succesfully and in a way that doesn't ruin the response.
I decided to stick with the ND28Fs
I removed the dust cap from the other driver to be sure there were no surprizes. I was a little more patient this time, covering the driver a couple times between a few applications of acetone. The glue seemed like it wanted to give up before the paper but ultimately the very top super thin layer of paper tore anyway. So it seems even more patience is needed. Fortunetely the glue they use on the former and spider joints is more resistant to the acetone (though not totally immune).
I found 1.5" PVC pipe is listed as ~1.9" OD and appears to be identical size of the pole piece leaving about the same amount of gap (maybe ~1.5mm). The ID is smaller than the tweeter body so I'll need to use a drum sander bit for the drill to open it up more.
Lopped off some pieces of pipe and shaved about half of the material out of the inside. Fortunately the appropriate dimension to clear the top edge of the former gives just enough space behind to solder on some thin wires and bend the terminals flat.
Next up, glue the face plates to the pipe spacers and trim the face plates down to the pipe OD. Got a simple plan for that. Just depends if I can get to that today.
This is stupid easy. I love these kind of projects. Super fun
Careful when you bend the terminals. They shattered the plastic around them when I did it once.
Keep it up, this is cool...
Oh, and the former rim should likely be trimmed down for best transition results. Not sure that this is doable in your case for clearances.
InDIYana Event Website
Got back from shopping and dinner early enough I could still use the router.
Super glued the face plates to the spacers then trimmed em down.
I used a solid carbide laminate trim bit since it only has one narrow cutting edge. Less aggressive and safer (still used a face shield lol). Placed a half inch scrap of birch ply on the router base to use as a mini table. Then just held the piece with my fingers, going slow. I did have a speed controller to slow it down a tad to keep melting to a minimum. Worked well.
Since I'm this far, ya know I had to plonk one in to get a look
Yeah I thought about trimming the former. The big glue glob where the wires come in would be a headache. I did trim up the glue glob a little bit on each one.. Hopefully I didn't already cut a VC lead [Edit: tested, they are fine] . Seemed like an opportunity for me to screw it up if I continued so I opted to just run with it as-is.
This is pretty much a simple proof of concept for me. I may look to do a version 2 in the future with better drivers, building off what I learned this time.
Hoping this could stoke someone else's fire too. Do better than me!
Soldered on leads. Terminals bent down fine. Hit the sides with a Sharpie. I want to coat the terminals with something since they will be inaccessible once everything is glued in place. Tape seems inappropriate. I haven't used "liquid electrical tape" before so I'll probably pick some of that up over lunch tomorrow.
You got lucky the PVC pipe was actually round. I guess the bigger the diameter, the worse that gets.
I like this! Moding into a coax is good stuff! Obviously there's a lot more to it to make a proper concentric coax, but fun to DIY a working example. 👍🏻
Having made a few coax speakers, I will say nice job getting a working prototype. I would l suggest getting the tweeter farther back, as in closer to flush with the cone edge. You'll notice when your testing starts that the fore and aft positioning makes a good deal of difference in the response, I've had to move tweeters as far as 80 thousandth's, but sometimes as little as 8 to 10 thousandths makes a large difference. Keep going, and play with the positioning once she's up and running, I think you'll learn a gain so valuable insight and have fun doing it.
So I imagine a waveguide that extends forward into the cone space would bypass most of that fiddlyness and effects of the woofer construction/geometry?
Not necessarily
Not in my experience, I have machined an extended wg like you see on some commercial speakers and it was still a bit position dependent, but more importantly it had a coloration no matter what profile I tried. In the end I gave up on it, but you have to be your own judge.
Cool project! I like coaxials, the advantages frequently outweigh the disadvantages. I would suggest listening to some good reference songs early and often as you make changes.
Aaaaallright. Glued the tweeters in and got some quick n dirty measurements. I just placed the drivers on the carpet face up and put the mic about 1ft above. Whether or not it is all that informative IDK. But it is something.
The small lines are 1db increments, so the darker lines are 10db. (1/12db smoothing)
Both tweeters:
Both woofers:
A single woofer/tweeter pair:
Maybe not stellar, but IMO not "orders of magnatude" worse than many other factory coaxial measurements that I've looked at. That 6-7.5k trough.. I suppose is the lack of good transition to the cone? The boosted top end seems kindof normal, even the sourcepoint 8s seem to exhibit this from what I read. Adjusted with amount of toe-in.
Since BSC will drop the woofer down a peg, they seem reasonably matched for sensitivity.
Not a bad place for a trough, however. Reason being that 6 to 8 Khz are where the "ssss" sounds live, if I recall correctly.
Maybe look at 15 degrees off axis ?
At ~6.2 kHz the woofer zigs and the tweeter zags - maybe really workable…
Would that response dip be related to the gap between the tweeter waveguide and the woofer cone? Maybe fill the gap with a 1/16" thick x 2.5" diameter foam rubber donut, then re-measure to see what happens.
The former stuck out probably 1/8" from the cone and the tweeters are mounted roughly flush with it. So i figure that's probably the big reason.
I was talking to guys at work about the little project. They seemed interested so i brought one in today for them to see. The cone was frozen! Only slightly, as a little bit of force broke it free. Yikes. Will check the other one this evening.
Acetone leakage?
Guessing the glue just took way longer than usual to cure and a tiny bit touched the former.