Davids new 8"+1" build, Dayton RS225-8, Scanspeak D2604/833000 in WG300, ~42 litre

[DaveFred]

Hello Everyone,

This middle aged woodworker, wannabe speaker maker, has decided to throw his hat into the ring at the up coming Parts Express 2023 Speaker Design Competition.

My entry will be a rather chunky stand mount bass reflex design speaker with the Dayton RS225-8 8" woofer and Scan Speak D2604/833000 tweeter in a Monacor WG300 waveguide. Enclosure size will be 21-1/2" tall by 13-3/8" wide by 15-1/4" deep, baffle thickness is 2-1/2", top/bottom/back are about 1-1/8" thick, sides are around 7/8" thick, yiedling an internal volume of about 42 litres. Front/back/top/bottom will translam style panels made from glued up 3/4" baltic birch strips and the sides will be 3/4" baltic birch with a shop made 1/8" thick bookmatched padauk veneer applied.

Here are some pics of the build,

Chuck of padauk I bought to resaw into veneer,



Four half sheets of 3/4" baltic birch ply, I have to have it cut into 1/2 sheets to get it home in my smallish SUV,



I bought a copy of Cut List Plus FX, and it was money well spent,



Sheets get reduced to what looks like a small pile rather quickly,



Gluing up the strips into the translam panels,



Padauk resawn,




All the glued up translam panels before surfacing,




Gluing up the bookmatched padauk veneers,



Sanding padauk glueup flat,



Quickie shop made cold press veneer press to glue padauk to baltic birch ply side panels,



All the blanks for the speaker boxes glued up, sanded flat and ready to be cut to size,

[JonMarsh]

My entry will be a rather chunky stand mount bass reflex design speaker with the Dayton RS225-8 8" woofer and Scan Speak D2604/833000 tweeter in a Monacor WG300 waveguide. Enclosure size will be 21-1/2" tall by 13-3/8" wide by 15-1/4" deep, baffle thickness is 2-1/2", top/bottom/back are about 1-1/8" thick, sides are around 7/8" thick, yiedling an internal volume of about 42 litres. Front/back/top/bottom will translam style panels made from glued up 3/4" baltic birch strips and the sides will be 3/4" baltic birch with a shop made 1/8" thick bookmatched padauk veneer applied.

Looks to me like you'll be having a lot of fun there! That's something unfortunate about living in the west- participating in the PE events just isn't practical.

How do you like the WG300 and 833000 combo?

[technodanvan]

Translam! Always a good looking choice if you have the elbow grease to spare. Though it helps having that large sander, I'm sure!

[theSven]

So many fun toys to play with! This looks like a really fun build and I'm looking forward to seeing it next month!

[Steve Manning]

Nice looking shop set up Dave. I used that same combination of woods on my first translam project .... I loved the look.

[DaveFred]

How do you like the WG300 and 833000 combo?

I will let you know in a week or so when I start the XO, might end up changing to Peerless XT25TG30-04 or XT25BG60-04 if I don't like it.

[tktran]

Hard domes or ring/dome radiators work great in waveguides. Soft domes: due to their breakups; are a bit more difficult to predict/model.

the XT25BG60 is a fine tweeter; with or without a waveguide.

When the RS225 was $35 a pop; it was ideal for isobaric.
(It needs a big box 40-50L; but when you squeeze two in isobaric you can have a smaller box) - which I had thought of that earlier!

[DaveFred]

Here are a few more pics of the build,

Cutting out driver holes and rebates. Baffle is 2.5" and router bit is not long enough...



Have to finish cut with jig saw.



Use flush trim bit to make jig saw cut flush with router circle jig cut





Drilling holes for driver mounting screws.



Expanding the holes in the back with a step drill bit,



Using a forstner bit the same size as the last step on the step bit to drill down for propell nuts



Opening the back of the baffle so the woofer can breathe, also expands the internal volume of the box just a little bit more,

[theSven]

The Baltic Birch glued up makes me think of Bamboo when I see it in the photos, then I have to scroll to the top to remind myself it wasn't bamboo, it just looks similar from a distance 😂. Did it route smoothly or did you have to remain cautious for splitting?

[DaveFred]

The Baltic Birch glued up makes me think of Bamboo when I see it in the photos, then I have to scroll to the top to remind myself it wasn't bamboo, it just looks similar from a distance 😂. Did it route smoothly or did you have to remain cautious for splitting?

Routes decently, but I am using all sharp carbide and spiral when I can.

[DaveFred]

More pics,

Dry fit of lock rebates, look for the tiny tabs between the vertical parts and the front and back,



Making the back thicker where the terminal cup and port will be,



Routing the terminal recess,



Drilling for rear port,



Spiral flush trimming rough hole to port tube,



What it looks like trimmed on the back,



And now with a 1" roundover on the outside,



Using a nice Milwaukee hole saw to cut out the braces, much nicer than other brands I have used, really clears the sawdust with good set on the teeth so it doesn't clog and burn like cheap ones,



Adding propell nuts, and to be sure, drilling small holes in them and securing them with 1/2" #4 screws,





Gluing up the translam pieces,

[JonMarsh]

Milwaukee hole saws are my favorite, too.

[theSven]

The build is starting to come alive as all the pieces are put together! Nice share on the hole saw, I'll keep that in mind when I'm ready to buy one. Why would you use propel nuts over threaded inserts, is that a preference or easier to install? Do you also use a little epoxy or glue to keep that set screw from never coming out?

[DaveFred]

The build is starting to come alive as all the pieces are put together! Nice share on the hole saw, I'll keep that in mind when I'm ready to buy one. Why would you use propel nuts over threaded inserts, is that a preference or easier to install? Do you also use a little epoxy or glue to keep that set screw from never coming out?

I wanted to install something that would not pull through or spin out, these guys with the added screw I think fit that bill...

[theSven]

Good to know and thanks for the speedy answer Dave.

[DaveFred]

Another update,

Lapping sides of glueup on a 60 grit sanding block, you can see the highs and lows..



Drilling dowel holes in box,




Using dowel pins to mark the sides of the box,




Dowels are drilled more into box than the sides,




Sides with shallower dowel holes,



Gluing sides on,



As someone who has a Porter Cable biscuit jointer and Festool Domino, neither can this simple job so easily in such a low tech, inexpensive way. If you have been fighting with panels sliding around when attaching them, just use a few dowels!

[Mikerodrig27]

Some people will throw salt into the joint. the salt gives it grit and melts at a rate that lets the glue get a grip on the joint. Although, maybe salt and high-end drivers do not mix.

[Steve Manning]

That's a heck of a sanding block! Cool idea for small projects.

[DaveFred]

That's a heck of a sanding block! Cool idea for small projects.

32" x 49" sanding block...

[DaveFred]

More progress,

Boxes all glued up, fresh out of clamps,



Flush trim with spiral upcut bit,



Still wasn't sure if I was going to do the edges the easy way with a 1.5" roundover, or the hard way with a big tapering bevel,



Bevel won out, and I made two end for end mirrored jigs to cut 90% of the bevel. It's too big a bevel for a 10" saw, but I can get most of it this way and clean it up by hand,



One jig for one side,



The other jig for the other side,



What remains that has to be cleaned up by hand. A mixture of hand planes and sanding blocks. I really wish I had of had a thin kerf low set Japanese pull saw, but I didn't have one on hand (I have since ordered one for next time),



Boxes sanded and ready for finishing,

[technodanvan]

You are great at keeping momentum on these builds. Your progress is nothing short of amazing!

[Evil Twin]

The Japanese saws are VERY useful... have a couple. Try to avoid needing to use them, but sometimes they're the best way.

[DaveFred]

Speaker boxes are done, time to mount the drivers and take 108 measurements.

Tweeter and woofer measurements taken on axis 0º to 180º horizontal and -180º to 180º vertical in 10º increments.

[DaveFred]

For better or worse, this is the crossover I have come up with,


I am not 100% sold on the value of the padding resistor, I will need to measure and listen to it more in the next few days before pinning it down.

[DaveFred]

For anyone who wants to play along at home, here is the complete impedance and measurement set,

Daves R2604-8320 RS255-8 2023 files.zip

[theSven]

Looking forward to seeing these live next week/end! The cabinets have a fantastic look with the translam Baltic Birch.

[DaveFred]

Final XO circuit and response is a little different then posted above, will post later, but here are the crossovers being built, with just a couple of days left to wrap this up...

[Evil Twin]

For better or worse, this is the crossover I have come up with,


I am not 100% sold on the value of the padding resistor, I will need to measure and listen to it more in the next few days before pinning it down.

I might recommend checking the depth of reverse null if you haven't already- this is very revealing of phase tracking in the crossover region as well as establishing the optimum forward response on the test and listening axis.

Of course, I realize you are on a tight schedule...

[DaveFred]

I might recommend checking the depth of reverse null if you haven't already- this is very revealing of phase tracking in the crossover region as well as establishing the optimum forward response on the test and listening axis.

Of course, I realize you are on a tight schedule...

How does this look?

(measured last night with the assembled XO's)

[Evil Twin]

That looks exactly like what you should see!

Most impressive!

[PANDINUS]

Well, that's interesting, because according to the data that Dave made available, it looks different in vituixcad.

[PANDINUS]

In the published image from vituixcad, it can be seen that the phase in the place of the dividing frequency are not completely aligned.

[Evil Twin]

I noticed this also- But this is a simulation based on input data- which must include estimates of the drivers Z axis offset- not always made correctly.

The measured response is the critical factor, which is why I brought up reverse phase null check, as it is highly revealing of actual phase issues on the measurement axis.

His result is most satisfactory.

[Evil Twin]

Whether by skill and actions he has not reported, or by luck, he has achieved the necessary result.

[DaveFred]

This was the final XO,



Make sure you have Vcad distance set to the 900mm the result was measured at, not 3000mm.

David.

[DaveFred]

Final XO modeled at 3 metres,


Inverted,



Modeled at 900mm, the distance I measured the big null at,





And inverted,

[DaveFred]

Some other actual measurements of the finished speakers,

Impedance of entire speaker one and two with XO,





Windowed measurement of both speakers 1dB apart,



In box measurement of full ported bass response, solid F3 of 40Hz and F6 of 35Hz



Real measured 0º to 50º response (gated and spliced at 300Hz with combined woofer and port nearfield merge),

[DaveFred]

A couple of more final pictures, installing the XO's, tweeter and then some shots of the finished speakers. My basement lighting does not show these in the best light...



















Thanks for watching and wish me luck at Dayton this weekend!

David.

[theSven]

I'm flying up tomorrow afternoon and looking forward to getting to see these speakers in action!

[PANDINUS]

This is interesting, I think this is where the error occurs. Yes, when entering a distance of 900mm in vituxCad, the phase traces get closer and the reverse null is nice. But those boxes will be 2-3 meters away from the ears when listening. That is, if they are not near-field monitors. So I'm interested in the phasing at this distance. The fact that the measurement is made from a lower distance is necessary to reduce room reflections. But it must be simulated with the input of the listening distance, which is why the setting is available in the simulator. If the measurement were carried out in an anechoic chamber, the microphone would be placed at a real listening distance and not much closer.

[Evil Twin]

Your assiduous effort and attention to details and construction should bode well for your experience at Dayton this weekend... your speaker crafting skills are strong with the Force!

[DaveFred]

Contest results were satisfactory.

[JonMarsh]

Says it all!

[tktran]

Winner winner chicken dinner!

In the end what made you switch tweeters?

[DaveFred]

I was basing my initial choice on the work of Heissmann Acoustics, but my results were not the same as his. Fortunately, I have a number of Scanspeak and Peerless tweeters here in stock that all mount the same way and I tried at least 5 different tweeters. This one had, in my opinion, the best response in the wave guide. I can post them if you would like to see them.

Thank you,

David.

[JonMarsh]

Yes, an overview of the waveguide and tweeter testing combo's you did could be very useful for folks.
Though, I gotta tell you, I think the reason why it worked out well is because it's basically a plane wave generator, unlike the way dome tweeters behavior. JMO. What this series of ring radiators doesn't do well, in it's normal bare configuration, is off axis dispersion, but the waveguide addresses that.

[tktran]

Yea please Dave.

Yes it would be informative to all those interested if you have comparative plots to demonstrate!


Congratulations on the win! With a passive crossover I don’t think one could get any better than what you’ve done with the crossover!

I was a bit surprised you didn’t win with your last entry; you know the one with the Wavecor MW and SEAS DXT IIRC (also documented here). You had a very methodical approach trialling lots of different midwoofers and tweeters.

Any ideas what clinched it this time round?

[DaveFred]

Some more pics that I didn't post previously for some reason,

This is my favorite wood filler, trying to match colours always looks out of place, but to me, black always fits,





So I go in the house for a drink cause I am thirsty, tired and sweaty from trying to lap the sides of these boxes on my big sanding block (it took a long time and was tons of work) only to find my wife left a helper out to help me finish the work....



Made me laugh anyway...

The Monacor WG300 waveguide really is a flimsy piece of plastic, so I filled it with Epoxy to add mass,



Stock pic of what it looks like,



Here is taking apart a stock Scan Speak or Peerless tweeter,











And just a shot of what Padauk look like unmilled with lots of exposure to oxidation and UV vs. freshly milled. The freshly milled orange will change over time to the dark red again.

[DaveFred]

Here are five different tweeter responses I tried.

What I was first going to go with, the Scan Speak D2604/833000 (based on Heissmann Acoustics work),



Peerless DX25TG59-04,



Peerless XT25BG60-04,



Peerless XT25TG30-04,



And finally, what I went with, the Scan Speak R2604-832000,


Which on do you like Best?

David.

[PANDINUS]

xt25bg60-04

[JonMarsh]

Here are five different tweeter responses I tried.

What I was first going to go with, the Scan Speak D2604/833000 (based on Heissmann Acoustics work),



Peerless DX25TG59-04,



Peerless XT25BG60-04,



Peerless XT25TG30-04,



And finally, what I went with, the Scan Speak R2604-832000,


Which on do you like Best?

David.

When I saw you design and build choices, I was wondering if the D2604/833000 was excluded based on total price limitation for your competition class. And that came to mind as one I would want to check out if I got my hands on some Monacor WG300.

Then, too, the SB Acoustics parts came to mind- not that fond of the bigger one, the SB29, but the SB21RDC might be interesting- (considering the linear excursion and very smooth response as a bare driver- enough to probably order a pair soon.

[JonMarsh]

Data items that would also be useful to evaluate would be the Power and DI as Vituix CAD reports- and personally, I could confine that analysis to something like 0-60 degrees.

Also, I like one of the options in Fuzzmeasure in analyzing measurements, the ability to take the RMS average of the combination of selected measurements.

[DaveFred]

Yea please Dave.
Congratulations on the win! With a passive crossover I don’t think one could get any better than what you’ve done with the crossover!

I was a bit surprised you didn’t win with your last entry; you know the one with the Wavecor MW and SEAS DXT IIRC (also documented here). You had a very methodical approach trialling lots of different midwoofers and tweeters.

Any ideas what clinched it this time round?

This is only the third "meet" I have been to, but I learned some at the first two.

First speaker InDIYana 2019, Too dark was the consensus, tuned to my ear, and I don't like Bright.

Second speaker InDIYana 2023, Tuned flat on listening window, speakers were toed in, on the bright side.

Third speaker, MAF 2023, Tuned flat to the on axis knowing I would toe them into the judges. This next bit is important to how I made this XO, and I cannot remember who said this to me or where I read it, could have multiple sources as I need to hear things several times for them to sink in. "People won't notice small holes in the frequency response, but they will sure hear a peak." <- This explains my tweeter response curve, yes there are holes at the high end, but there are no major peaks. That was a very specific design decision on my part. I thought about raising up the upper end and adding a wide notch to squish the peak, but decided against the added complexity to the XO.

Here is a quote from Jeff Bagby who was absolutely instrumental in me not leaving this hobby in frustration within the first couple of months, he was consistently helpful to me back in 2018-2019 when I was getting started, him not getting upset with all my dumb questions (played out on Javads FB DIY Loudspeaker Project Pad). His XO's always had minimal parts, and here he explains why,

"As a result my rule is to avoid overly complex electrical networks with lots of reactive components, keep the crossover as simple as possible for good results, and use high quality components when you do. From my own experience if you keep adding components to a circuit to correct for everything you can think of you will find that somewhere along the way your sound became compressed, flat, and lifeless. There is a cumulative effect to the degradation from all of those components."

The XO is a simple second order electrical on the tweeter and fourth order electrical on the woofer causing an almost perfect textbook LR4 XO at 1550Hz. Jeff always talked about how a perfect reverse null was not the most important thing, but that phase tracking was. If you look back at my posted XO with phase showing, they track (parallel) almost perfectly.

At 3m,



When the tweeter I planed on using response didn't look like what I hoped it would, I was able to try four other tweeters the same morning, having an inventory of drivers on hand was a big plus.

Another point about this speaker, some might have tuned the box lower for seemingly maximum extension, I did not. I deliberately tuned the box a little high to give a bump in the bottom end. Here is me playing with port lengths, (I went with an alignment close to the second shortest port length tested). Notice how F3 is about 37hz for a number of port lengths, where you might get a lower F6, but I didn't like the roll off, I preferred flat with a little bump.




This next bit is in no way meant to be a slight on InDIYana, it is what it is, and I knew that going this year. Ben comes up with a challenge, people bring speakers to meet that challenge, they are played in rapid succession in groups of three with the winner from the groups of three advancing (or real close to this). Everyone in the room votes, but not everyone in the room hears the same thing, guys at the front middle are not hearing the same thing as the guy in the back corner, and everyone has their own reasons/criterion for voting for a given speaker. "I" think it is meant more as a get together and, Hey, lets listen to each others speakers than it is a raw contest. (I hope I am not coloring this in the wrong way, these guys have been meeting for a long time, and I am just some newbie from Canada who showed up out of the blue)

At MAF it is a speaker contest, the three judges sit centered 10' from a pair of speakers 10' apart and a specific set of criterion is used for judging.

Nice looking cabinets could have only helped my entry.

Waveguide, Waveguide, Waveguide. Overall good directivity and power response.

tktran ^Some blend of the above is why I think I won, oh, and, a bunch of newbie dumb luck .

I will know more about why the judges picked my speaker when the score sheets hopefully show up in the mail in the next few days. I was told the judges decision on first place was unanimous.

And, Thank you Reid.

[JonMarsh]

Very good and insightful comments on how perception will play out, relative to design decisions and polar behavior and the way judging is setup. Keep up the comments on design insight and your personal methodology- I think that is very helpful for other potential designers.