• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

DIY speakers with Waveguide

617

Major Contributor
Forum Donor
Joined
Mar 18, 2019
Messages
2,403
Likes
5,296
Location
Somerville, MA
After some delays I have decided to post information about my latest speaker build in progress. My hope is that others can learn from my process, making speaker building seem more approachable, and that I can learn from the knowledge of the ASR userbase.

This project started with the impulse buy of four Rival carbon pulp woofers.
https://rhythmaudiodesign.com/collections/rival-acoustics/products/r176-p-08-carbon-pulp
I bought them because I was told they are as good as Wavecor/Revelator woofers. I think I paid $50 apiece, rather than the 150-200 for woofers of similar quality. So, the question is what to do with them?

A 17cm woofer is getting to be a bit big to integrate with a direct radiating tweeter in my opinion. Everyone does it, but polar responses are not going to look great unless the woofer has great HF extension (not the case here - typical paper cone breakup), your tweeter is a beast (beastly tweeters are common nowadays) and you use a second order crossover. For this approach to work, ideally the tweeter's acoustic center would be located over the woofer's, necessitating a sloped or stepped baffle.

A more elegant solution is the use of a waveguide loaded tweeter, but how to approach this?
- Off the shelf waveguide tweeters (Wavecor 30mm, Seas DXT, Morel?) don't have particularly large waveguides, although they are better than nothing
- There is a SEOS waveguide for dome tweeters, but the tweeter it was designed for is NLA
- Visaton makes a waveguide/dome combo, but it is hard to get in the USA and the tweeter itself didn't seem that great

In the end I decided to use the Visaton waveguide (148R) and adapt it to a series of tweeters to see what the performance looked like. I 3D printed adapters with the thickness of the existing faceplate, enabling a very seamless transition from the dome to the waveguide opening:

20190128_194213.jpg20190128_194143.jpg

I tested 3 tweeters - the Dayton RST28f, the SB acoustics SB26STAC and at SEAS 27mm unit. Of the three, the Dayton had the smoothest response and the best high frequency extension. As a bonus, the RST28f is an extremely robust tweeter with a low resonance, allowing a low crossover point even without a waveguide.

Cabinets
For me, the hardest part of a DIY build is making the cabinets; I like woodworking but I don't have a great space for it now. As a result I enlisted https://www.speakerhardware.com/ to build some cabinets to my specifications.

I provided a sketchup model and a detailed drawing of the front and back panels, and for a reasonable price I was provided with two nice cabinets made from baltic birch plywood. Dimensions are 8x11x16", and the cabinet is tuned to around 43hz. Maximum power/xmax limited SPL is 105db/1m,
wg cabinet.png

The cabinet is designed to be DIY friendly - the rear panel can be completely removed to make crossover changes easily. It is attached to cleats which are lined with thin neoprene for an airtight seal, probably not essential for a ported cabinet.

The cabinet is lined with mass loaded butyl rubber:
https://www.amazon.com/gp/product/B01KZ5X7KO/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1
I've never used this but it certainly makes the cabinet 'seem' better; it doesn't ring at all and it's quite a bit heavier.
After that, I used a combination of 2" open cell foam and 1" cotton acoustic dampening to line the cabinet.

I'll put some pictures of the cabinet up later, but basically it looks like a typical european DIY speaker - think Troels Gravesen's designs.

Cost:
People often ask if DIY is worth it - I can't answer that question for you, but I can provide some idea of cost. For me, this speaker's cost breakdown looks like this:
Rival Woofers - $120 (big discount)
Tweeters - $60 (on sale)
Waveguides - $26
3D printed adapters - $40
Cabinets - $160
Hardware, speaker lining and stuff - $70
So, total outlay at this point is under 500 dollars. The crossovers will probably require another $60, just as a rough estimate (I expect crossover design to be simple.) Will this equal the performance of an 700 dollar speaker? I think so... but mostly it's just fun to make your own speakers, and it's great to be able to tweak the tonal balance to your room.

Design Tools:
- ARTA measurement software for distortion measurement and polar measurements
- Vituixcad for 3d visualization and crossover design
- Calibrated Dayton EMM microphone
- Behringer audio interface

It is probably possible to design a speaker using a USB mic but I'd advise against it. A usb mic provides no way of measuring the delay between the outgoing measurement stimulus and the measured signal coming back from the speaker. This makes it very cumbersome to extract the time-of-flight of the sound coming from the speaker - you don't know if the signal you have is delayed due to the physical position of the drivers tested, or because of some kind of latency in the capture process. ARTA allows you to 'loop back' the outgoing signal into one of your mic inputs, which provides higher accuracy. This makes crossover design much more accurate.

Hopefully measurements of the drivers in the completed cabinets soon.
 

dc655321

Major Contributor
Joined
Mar 4, 2018
Messages
1,597
Likes
2,235
Good stuff, @617.
I've been planning on doing a similar project myself, so your info is very helpful.
I was going to go 8" woofer/mids, with a waveguide+tweeter (have the wg-300, also looking at augerpro's 3-d prints).

It is probably possible to design a speaker using a USB mic but I'd advise against it. A usb mic provides no way of measuring the delay between the outgoing measurement stimulus and the measured signal coming back from the speaker. This makes it very cumbersome to extract the time-of-flight of the sound coming from the speaker - you don't know if the signal you have is delayed due to the physical position of the drivers tested, or because of some kind of latency in the capture process.

A USB mic is no problem with REW - a pilot chirp (or pair) can be included to precede or bracket the probing sweep.
Delay is found relative to the pilot chirp. Using a pair of pilot chirps permits sample rate corrections too.
 
OP
617

617

Major Contributor
Forum Donor
Joined
Mar 18, 2019
Messages
2,403
Likes
5,296
Location
Somerville, MA
Good stuff, @617.
I've been planning on doing a similar project myself, so your info is very helpful.
I was going to go 8" woofer/mids, with a waveguide+tweeter (have the wg-300, also looking at augerpro's 3-d prints).

A USB mic is no problem with REW - a pilot chirp (or pair) can be included to precede or bracket the probing sweep.
Delay is found relative to the pilot chirp. Using a pair of pilot chirps permits sample rate corrections too.

I'm not 100% certain that this would work, but if it does that is great news. The proof would be if the summation from your crossover simulation matches measurement - let me know if it does.

What woofer are you looking at? I always thought the Dayton high efficiency 8" (PM220) would make a really interesting choice.

If I was starting now I'd probably do the augerpro WG with the SB26 ceramic coated dome.
 

TimW

Major Contributor
Joined
Aug 15, 2018
Messages
1,065
Likes
1,406
Location
Seattle, WA
What is the affect of loading a standard dome tweeter in a waveguide? From what I gather it affects directivity and frequency response but I've never looked at before and after measurements. I just picked up some DC28F-8 tweeters for $16.50 each that I was thinking of pairing with these DSA215-8 woofers I've had for a while. I've also got a 3D printer and those Visaton waveguides are reasonably priced so this could be a could be a good combo. Thanks for making this post @617
 
  • Like
Reactions: 617
OP
617

617

Major Contributor
Forum Donor
Joined
Mar 18, 2019
Messages
2,403
Likes
5,296
Location
Somerville, MA
What is the affect of loading a standard dome tweeter in a waveguide? From what I gather it affects directivity and frequency response but I've never looked at before and after measurements. I just picked up some DC28F-8 tweeters for $16.50 each that I was thinking of pairing with these DSA215-8 woofers I've had for a while. I've also got a 3D printer and those Visaton waveguides are reasonably priced so this could be a could be a good combo. Thanks for making this post @617

A waveguide will:
- Reduce radiation to the sides of the tweeter, lessening or eliminating baffle edge diffraction
- Narrow radiation of the tweeter's low end, making reflections from near walls quieter
- Move the tweeter back a bit so it is better aligned with the acoustic center of the woofer
- Boost the tweeter's output, making it play louder with less excursion, resulting in less distortion
- Narrow the dispersion of the tweeter's lower range so that it matches the dispersion of a woofer, making off axis sound much more linear, improving imaging
 

dc655321

Major Contributor
Joined
Mar 4, 2018
Messages
1,597
Likes
2,235
What is the affect of loading a standard dome tweeter in a waveguide? From what I gather it affects directivity and frequency response but I've never looked at before and after measurements.

Without waveguide and with, for an inexpensive 1" dome (~$20 tweeter + $15 waveguide + $10 3-d printing for adapter).
These are the initial measurements that sold me on the possibilities afforded (basically @617's points, above).
More interesting developments at diyaudio.
 
OP
617

617

Major Contributor
Forum Donor
Joined
Mar 18, 2019
Messages
2,403
Likes
5,296
Location
Somerville, MA

dc655321

Major Contributor
Joined
Mar 4, 2018
Messages
1,597
Likes
2,235

TimW

Major Contributor
Joined
Aug 15, 2018
Messages
1,065
Likes
1,406
Location
Seattle, WA
So looking at the DX25TG09-04 graphs with and without waveguide, I can see how off axis response has been modified in a way that would allow for better integration (or smoother off axis response when combined) with a woofer. But I also notice that the waveguide acts like somewhat of a band-pass filter, causing a big hump in the response centered around 2 KHz or so. How would you accommodate for this response in a crossover, be it passive or active? I can imagine using some sort of shelving filter in an active crossover.
 
OP
617

617

Major Contributor
Forum Donor
Joined
Mar 18, 2019
Messages
2,403
Likes
5,296
Location
Somerville, MA
So looking at the DX25TG09-04 graphs with and without waveguide, I can see how off axis response has been modified in a way that would allow for better integration (or smoother off axis response when combined) with a woofer. But I also notice that the waveguide acts like somewhat of a band-pass filter, causing a big hump in the response centered around 2 KHz or so. How would you accommodate for this response in a crossover, be it passive or active? I can imagine using some sort of shelving filter in an active crossover.

At least with a passive crossover, you can actually linearize that response simply by tweaking the high pass filter. If you're using a digital active crossover, you can do it in any number of ways.
 

dc655321

Major Contributor
Joined
Mar 4, 2018
Messages
1,597
Likes
2,235
But I also notice that the waveguide acts like somewhat of a band-pass filter, causing a big hump in the response centered around 2 KHz or so.

Yes, the acoustic gain available from the waveguide is only offered over a band of frequencies (1st & 2nd figures in link).
Notice that out around the top end of the tweeter's range (> 10kHz), the with-waveguide output is equal to the without-waveguide configuration (on-axis).
By trimming the output over the gain-band of the waveguided tweeter back to intrinsic levels, one is reducing the input power required.
In this particular case for this relatively "deep" waveguide, that's about a 10dB (max) drop from 1.2-10 kHz.
 

TimW

Major Contributor
Joined
Aug 15, 2018
Messages
1,065
Likes
1,406
Location
Seattle, WA
Is matching a tweeter to a waveguide very complicated? For example if I bought those Visaton waveguides and designed my own adapters for the tweeters I have, what issues could I run into? If frequency response and distortion measurements look good should I assume everything is good?
 

HammerSandwich

Major Contributor
Forum Donor
Joined
Nov 22, 2018
Messages
1,137
Likes
1,497
The attenuated band's distortion improvement was shown by Zaph back in the day, too. It's impressive!

A good waveguide has another benefit I missed above: while it narrows "LF" dispersion, it can widen the HF pattern. IOW, you don't get only an improved mid/tweet handoff, rather the speaker's DI improves across the whole band. See Stereophile's Dutch 8C horizontals (I can't think of a better speaker in that area.) or go back to Zaph (DXT @ 30º versus all else).
 

HammerSandwich

Major Contributor
Forum Donor
Joined
Nov 22, 2018
Messages
1,137
Likes
1,497
Is matching a tweeter to a waveguide very complicated?
Dimensions are critical, because HF wavelengths are so short. 3D printing's popular here for good reason.

For example if I bought those Visaton waveguides and designed my own adapters for the tweeters I have, what issues could I run into? If frequency response and distortion measurements look good should I assume everything is good?
That's most of it, but you should look for reflections in the time domain as well.
 

dc655321

Major Contributor
Joined
Mar 4, 2018
Messages
1,597
Likes
2,235
Is matching a tweeter to a waveguide very complicated?

Yes, careful matching is required.
Have a look through the first few pages of the diyaudio thread I linked above.
Throat-to-dome aspect is very important (i.e. aperture/acceptance matching).

If frequency response and distortion measurements look good should I assume everything is good?

Off-axis spectra will indicate yea or nea (0, 30, 60, 90 deg, etc), in both planes.
 

TimW

Major Contributor
Joined
Aug 15, 2018
Messages
1,065
Likes
1,406
Location
Seattle, WA
Sounds like a lot of trial and error. I may just try printing one of the designs from that diyaudio thread. @dc655321 do you know which page of that thread has stl files for download?
 

Biblob

Addicted to Fun and Learning
Joined
Sep 13, 2018
Messages
633
Likes
600
Thanks for sharing this. I am busy designing one with a 6" WG from Augerpro.
My buddy a woodworker will start on the cabinet this weekend :)

Here is the concept in sketchup. It's going to be a 2.2 system. Woofers are SB17NBAC-4's (14L cabinet) and the subs are CSS SDX10's (18L cabinet).

screenshot design.PNG
 
OP
617

617

Major Contributor
Forum Donor
Joined
Mar 18, 2019
Messages
2,403
Likes
5,296
Location
Somerville, MA
Thanks for sharing this. I am busy designing one with a 6" WG from Augerpro.
My buddy a woodworker will start on the cabinet this weekend :)

Here is the concept in sketchup. It's going to be a 2.2 system. Woofers are SB17NBAC-4's (14L cabinet) and the subs are CSS SDX10's (18L cabinet).

View attachment 39750

Those css ldx10s are beasts! I have two.

These all look like excellent designs.
 

Biblob

Addicted to Fun and Learning
Joined
Sep 13, 2018
Messages
633
Likes
600
Those css ldx10s are beasts! I have two.

These all look like excellent designs.
Thank you. It's my first ever DIY too, so I feel flattered. Have much to learn still.
Still don't know how I'm gonna measure accuately enough to design the XO. I want to cross at about 1,6 khz though.

Does someone have a guide how to do this with REW and UMIK-1?
 
Last edited:
Top Bottom