injection mold question

Hi,
Currently Im designing a mesh like hole but have problem with the mold direction.
The most similar shape with the design intention I have is google nest wifi point.
It have a holes that seems have mold pull direction radially instead vertically (bottom and up direction)
how do they do it?

Hi Greg,

We cannot see your attached image.

my bad, fixed

A collapsible core would be my guess.
Little animation of it:

Really good question.

I went to iFixit (a shining beacon of rationality btw) in hopes they had a teardown.

Someone else had the same question as the OP, wondering how they dealt with the undercuts and witness lines. Unfortunately there aren’t good images of the interior of the molding.

Our resident plastic molding expert thought they were formed from the outside, in vertical groups. Its hard to get a good image of this part though, he’s just guessing.

This is the issue when designing products that you want to look like products from massive tech companies. They have quarries-full of cash to throw at products that don’t need to be profitable at the product level.

That is true. Most of these speakers from tech companies are loss leaders… they make money on the data… Ironically most of them are also terrible shapes for acoustics and any audio engineer I ever worked with would have slapped me for making such an acoustically opaque enclosure. Typically they want at least 50% open area. Sound waves need to flow with a minimum of obstruction. If they hit too many hard things they start breaking up and bouncing into each other… also a lot of these speakers fire their drivers downward… imagine instead of talking toward someone instead you put your face an inch away from a table and yelled? How would that sound? … to make up for this expensive enclosure and bad circular archetype they tend to add more amplification (even more money) and a lot of DSP work (more time… which means more money)… but a lot of companies try to essentially copy these products without thinking what a better solution might be.

It is cool haw they molded the holes though. If it was slides you would likely see the tool marks… I’m not completely familiar with collapsing cores, but would they be able to essentially actuate that may rods 360° around the part? Seems like that tool would wear out quick! I would buy a few units, take them apart and send one to your manufacturing partner.

I believe the Amazon Echo enclosure is extruded and then the holes are post drilled, so that’s a possibility as well…

Yes. Then they buy more tools, because another $150K in tools is also inconsequential in their grand scheme.

Interesting about optimizing acoustics and how poor those cylindrical forms really are. Its not like the drivers are cylindrical as well!
That new Beosound speaker has some fancy array that ‘knows’ when its next to a wall and adjusts its performance or driver usage accordingly.

Well, if the driver is facing you then it is cylindrical in the wrong direction… plus any good speaker is going to have more than one driver plus tweeters and bass radiators…

The cylindrical form factor only really makes sense as a cheap small can speaker facing you like this:


as soon as you want to do something with good voicing, spacial imaging, and that can actually play both high and low frequency sounds you end up with arrays. You can put an array in a cylinder, but it isn’t typically cylindrical. Might be octagonal. Notice how big that B&O speaker is. That is about right if you want to make a cylinder sound good.

Here are some powered array configurations I’ve worked on to give you a sense. Some are simple… some are not… :slight_smile:

Sorry, way OT, but point being, be careful who you chase, you might catch them!





(edit: realized I’m getting way OT. sorry not sorry. WFH makes everything blend together including forum topics. Topical content - agree on the %age of open area on a loudspeaker being important if you want it to sound good.)

This was one of the first products I designed that went to market and did well. Snell Acoustics XA90. Wish I bought a set. We really emphasized the “skinny+deep” graphic that signified high-end speakers at that time. We did the little ‘bazooka’ bulged-out sides to let the big drivers have a good seat in the baffle, and emphasize the space inside that ‘volume’.

This was with Catalano Design, in Boston.

(Funny thing about that Beosound, even the designer admits that B&O doesn’t seem to have published any images that clearly show the scale of the speaker. Its bigger than it looks in photos.)

nice accentuations of the lower drivers. Where those powered woofers, radiators, or connected to the amp? (sorry, nerdy question) …

I love having a set of these in my living room :slight_smile: Bi-pole front and retiring midrange and tweeters, lower side firing powered subwoofer woofer and dual side firing bass radiators … they crank. So they have drivers on 4 sides, but you can’t see them. No option to remove the grilles… basically it is an MDF box with some milled chamfers and CNCing with CNC aluminum top and cast base… pretty efficient construction.

I seem to recall the woofers had their own power…yes, according to an old spec sheet they had 300W for the dual 10" woofers. They weighed 100lbs apiece! Funny to recall that the tweeter plate height was driven mainly by the requirement of being in-line with the ear holes of some crusty audiophile sitting in his listening chair. Though the sub Hz isn’t directional, these crusty dudes were all about the ‘true’ stereo experience…plus I think Snell wanted to show off more woodwork, which built on a great New England heritage of cabinetry, yadda yadda yadda.

that sounds about right :slight_smile: Though what those guys can hear is pretty amazing. I really loved working with those guys. They taught me so much. This was one of the last systems I touched while at SU. Very cool phase array that actually used noise cancellation tech but in the room to push sounds outward. The ridges in the midrange drivers allowed us to use a thinner material while still maintaining rigidity… low weight + high rigidity equals better reproduction (faster movement of the cone)… The downward firing port pushed long bass waves out in a radial fashion. That little Hershey’s kiss shape in there did a lot of work actually… it takes a lot of wiring to set these up. You need to run left channel cables to the right speaker and vice versa in addition to the usual beaming so you are talking 3 sets of cables going to each speaker… and if you add the ATMOS module its another set. But it is fun to design for these enthusiast users sometimes.

sorry OP! We nerded out.





I pulled similar apart last year to build the internals into a new shell. It had a 6 part tool with top, bottom and 4 sides. The holes were similar to the one you’ve shown, and if you looked really closely the wall of the perforations was actually linear to suit each slide and not radial. It had very faint witness lines at each quarter.

Off topic- the one I took apart had a similar waveform? to that Polk to distribute the sound waves horizontally. What’s the technical name for it?

Maybe we should have an audio thread :slight_smile: I love sound systems!

+1 I bet you are right about this. Could be more than four, but most likely not completely radial.

Sound design forum topic ‘sounds’ good too.

:joy:

It seems like 6 would be on the money to achieve this without being too obvious on a circular form.

dont worry guys, keep them coming! Im learning here :slight_smile:

Maybe each column of holes is made with a slider made of 3 pins in the mold. Or maybe they are all seperate. It would be a complicated mold, but it seems doable. The only witness lines would be small and on the inside of the part.

Beautiful speakers. I wish I had the chance to work on some. I loved heaters because to get a good design required a real understanding of air movement. I imagine speakers would be a similar challenge, just 10 times harder.

basically a speaker moves air around :wink: