Yeah I don't want something huge, these have to printed after all. And I want it to work with a finished box, I tend to cut the driver cutouts last because I like to make the baffle slightly oversize, then use a flush trim bit on the sides after I glue it to the box. I do this because the baffle and rear panel always seem to slide when I glue them up.
I'm think about putting holes for #4 screws. Something small. Going to the hardware store tomorrow to see if they have 3/4" ones. I think the finish nailer worked great, but maybe a lot of people won't have those.
I'm think about putting holes for #4 screws. Something small. Going to the hardware store tomorrow to see if they have 3/4" ones. I think the finish nailer worked great, but maybe a lot of people won't have those.
Very early in the project, Erich at diysg contacted me, but my progress was slow and I never heard back from him. I've been thinking about making them myself. If anyone knows how I can get into DIY injection molding or casting, let me know.
I've done about 25 injection molded parts, including one where I made the mold myself and shot parts. These waveguides are large enough that they will need a larger machine that you're likely to be able to host. There are a lot of design guides out there if you google for them, but I've done a lot with Protolabs: https://www.protolabs.com/services/injection-molding/plastic-injection-molding/design-guidelines/
Their automated quoting system can give you feedback on if your part is going to work and where the trouble spots are. Wall thickness uniformity and sink marks on the visible face of the waveguide are the main things I think you'll have trouble with- mostly from the need to have screw bosses for mounting the tweeters. If you want to bounce ideas off of me I don't mind giving feedback.
I'd expect a Protolabs tool to make these waveguides to cost between $5k to $15k, and if you find a china based company and do a small enough waveguide with a long lead time you may get under $5k.
I know that Erich at DIYSG is holding a fairly large number of SEOS-8 waveguides that went obsolete when the dayton tweeters got changed. I'm working on doing some measurements on some of those to see how they'll work with the new version or some other tweeters, but I had some things come up personally that slowed my progress. I figured you might want to know his status with regard to small waveguides. He'd probably be of help in getting these made, but I would expect the point where it 'makes sense' is at least a thousand waveguides.
I also had the 8" design quoted for CNC in wood, and it was a few hundred bucks each for the full baffle. That could come down more for an order in the low hundreds of pieces, so that would be another avenue to explore. Much less setup charge there, and lower risk in terms of being able to check quality before pulling the trigger on the big batch.
Thanks Joe for feedback. So you think these waveguides are too large for a DIY injection molding setup? Not that I know how I would build one of those anyway. I thought this was cool: Easy Composites, but way too slow with hand loading the mold with fibers and wetting them down with resin. It seems like molds are easy to make these days, I just need a material that is quick and easy to inject in them for the actual part.
The kind of "DIY-friendly" single shot injection mold machine that I used is able to do parts about the size and shape of a cassette tape or a drink coaster at most. Plastic injection molding requires plastic injection in the thousands of PSI range, and so every square inch of cross section of the part adds thousands of pounds of clamping force to keep the mold shut, so you're looking for a machine with a capability measured in hundreds of tons of clamping force, and you need a space on slab in your garage of at least ten foot by three foot to host it, plus a significant air supply. Have I seen a guy do it in his garage? Yeah, but I think even his machine would come up short on an 8" waveguide. He probably had $50k in his machine, but he was making the leap to manufacturing stuff in his residential garage that probably should have been in a commercial space.
I think you're on the right track with that, but I don't think you need to lay up fibers. If the 3D printed stuff is stiff enough then surely a slightly thicker wall of casting resin maybe with some short glass fiber blended in will be something you can pour. A machined aluminum mold with mold release could give you good dimensional accuracy for under a grand, and you might be able to make a couple of these per day for not very many minutes of hands-on work and some tens of dollars of material per horn. It's not a great business proposition, but if you wanted to make a few dozen pieces for you and some friends' surround sound systems then I think that's a pretty good deal.
I think you're on the right track with that, but I don't think you need to lay up fibers. If the 3D printed stuff is stiff enough then surely a slightly thicker wall of casting resin maybe with some short glass fiber blended in will be something you can pour. A machined aluminum mold with mold release could give you good dimensional accuracy for under a grand, and you might be able to make a couple of these per day for not very many minutes of hands-on work and some tens of dollars of material per horn. It's not a great business proposition, but if you wanted to make a few dozen pieces for you and some friends' surround sound systems then I think that's a pretty good deal.
I'd go to china. I work at a small company doing audio equipment. we've had a lot of custom stuff done in china including 4 small ABS injection molds. many are SO keen to get your business. They often a real attitude of "lets find a way to make this work". If you're really keen, i'd just go on alibaba, find a company who makes something similar, and just put it to them. "hey I wanna make 100 of these, how much will tooling be, and how much will each horn cost". can go to a few of them untill you find somebody keen to do business. no lie though, I think the tooling for something like this might be 10k USD
What differences are between waveguides for Scanspeak R2904/7000 and Scanspeak D3004/6600? It looks that it may be useable for Vifa XT25TG30 too.
Just made a quick box to test the 5" waveguide with a SB26ADC - works really well.
Sound seems much more clear than my Seas DXT - even though both measurements are very smooth and even. The ADC seems to need way less EQ to work well.... which might give a hint. Awesome work you did with the waveguide - donated
Sound seems much more clear than my Seas DXT - even though both measurements are very smooth and even. The ADC seems to need way less EQ to work well.... which might give a hint. Awesome work you did with the waveguide - donated
I would be curious to hear from the experienced 3d printer operators: recommended number of perimeters, top and bottom layers, and infill settings. Presumably the waveguides are printed face down with supports?
Seems that the best use for a 3d printed template might be to facilitate creation of an adequately large wooden template that could then be clamped to the cabinet or baffle. The 3d print would not have to be large enough to properly support the router, as it could be attached to the bottom of the final template, which could then be routed with a flush trim bit from the opposite side.
That will leave a poor finish on the important side where the supports have to be removed.
Cubic subdivision or similar from 20% upwards makes quite solid parts, gyroid is ideal if you want to fill the voids with resin, the resin can find it's way through but many resins get hot when curing.
Augerpro was there plans for circular sb26 waveguides? would it be easy to create using the same contours or do you need to start from scratch? Thinking more about how a tight CTC is not so critical and that around 1.2x XO wavelength spacing is actually a better case scenario for even sound power (as per V.CAD sims) circular would be an easier option
I tried circular and it didn't measure as well. It's somewhere in this thread, I'll see if can find it.
BTW I think this claim by a single person that 1.2x wavelength is the best distance is gigantic oversimplification that will lead many down the wrong path. This claim needs to be backed up by an actual sim that we can all examine the assumptions so we can understand what we are even talking about.
BTW I think this claim by a single person that 1.2x wavelength is the best distance is gigantic oversimplification that will lead many down the wrong path. This claim needs to be backed up by an actual sim that we can all examine the assumptions so we can understand what we are even talking about.