Introducing the SSA-WG

[Steve Manning]

Pretty slick Scott, looks like that worked nicely ......

[ergo]

Nice work!
These build pics are so great for getting ideas and hints on tools. I had no idea this type of handle exists for spay cans.

[sdl2112]

Thanks ergo. Built threads are where I have gotten some great ideas...I think this is where I saw the spray handle too. Highly recommended, less fingler fatigue = better results.

[TEK]

Just cool, really cool...
And as we all know - you NEED new tools for every stage in every new project!
That's just the way it is - there is no point fighting it!

[flamethrower1]

Nice job, very nice looking

[ergo]

And as we all know - you NEED new tools for every stage in every new project!
That's just the way it is - there is no point fighting it!

Oh yeah - isn't that the truth. And usually one will conveniently neglect to add that cost to the project cost when telling the wife how much was spent on a project etc. Not very different to many R&D projects in big corporations I'm sure.... but thats part of the fun of this hobby.

Also I often think how much luckier I am compared to my grandfather for example who would have to literally make many of the tools he needed for some tasks as those were not available even if you had bit of money (I grew up on the other side of the iron curtain back then).

[TEK]

I think making tool, the one I want to, can be really fun ;-)
About the cost information - take care, someones wife might be lurking around

[JonMarsh]

Very nice looking work, dig your version of a home made spray booth. Particularly your whole house air reservoir!

Now, that's a funny thing- I'd almost expect there to be more espionage by wives and GF's over what is going on here, but even though I've volunteered the links even, it's as if there's a sign out front that says "Boys Only- nothing interesting here, these aren't the projects you're looking for" with regards to checking out what's going on and what shop talk we might be engaged in.

Yeah, NO ONE ever lists the new tools they buy to finish a project as an "expense"- that's an investment, which is depreciated. Actually, more like "appreciated" given the future use we get...

[sdl2112]

Thanks everyone for the positive comments. I think just about every woodworking tool has been purchased with loudspeaker building in mind.

I still remember starting out saying...ok all I need is a my dads Craftman circular saw, jigsaw, a sheet of plywood and saw horses. That sheet of plywood turned into my work bench top with vise, it then grew legs...then I bought this thing called a router...then....

[sdl2112]

Little update...I added an avatar. I like cats...we'll see if sticks, I may change it.

Now a little speaker update. The speakers are getting quite heavy so I made a transport sled so when I slide it around it won't mar the surface.

They are unfinished but I want to install drivers and take measurement now that I have the real enclosures.

Here I twisted up 14 awg Teflon silver plated wire for internal wiring.


Here's the sled.




Here I did a quick assembly to see how it looks. A lot of little things still to do before measuring.

[TEK]

Looking good, will be interesting to see what your measurements show.
Are you planning on going passive or active with these?

[BobEllis]

The follow on picture to your avatar is the cat knocking the phonograph over.

[sdl2112]

Looking good, will be interesting to see what your measurements show.
Are you planning on going passive or active with these?

It will be passive. The measurements shouldn't be too different from my test box measurements posted earlier in this thread.

[sdl2112]

The follow on picture to your avatar is the cat knocking the phonograph over.

I think you got that right. :lol:

[flamethrower1]

Nice job, those are looking quite nice

[Bear]

Looks great!

[JonMarsh]

Little update...I added an avatar. I like cats...we'll see if sticks, I may change it.

Now a little speaker update. The speakers are getting quite heavy so I made a transport sled so when I slide it around it won't mar the surface.

They are unfinished but I want to install drivers and take measurement now that I have the real enclosures.

Here I twisted up 14 awg Teflon silver plated wire for internal wiring.
[ATTACH=CONFIG]26038[/ATTACH]

Here's the sled.
[ATTACH=CONFIG]26039[/ATTACH]

[ATTACH=CONFIG]26040[/ATTACH]

Here I did a quick assembly to see how it looks. A lot of little things still to do before measuring.
[ATTACH=CONFIG]26041[/ATTACH]

Looks like a good plan for moving them around and working on them. Any idea of the current weight, Scott, without drivers and crossovers?

I'm thinking something similar, but with appliance dollies I have... Plus putting them on furniture dollies, with individual lazy Susan's (like I did the Ardents) when it comes to finishing....

Where did you source your moving blankets from, out of curiosity?

[sdl2112]

Looks like a good plan for moving them around and working on them. Any idea of the current weight, Scott, without drivers and crossovers?

I'm thinking something similar, but with appliance dollies I have... Plus putting them on furniture dollies, with individual lazy Susan's (like I did the Ardents) when it comes to finishing....

Where did you source your moving blankets from, out of curiosity?

Thanks everyone for the kind words.

I made the sled so when I'm moving them around (like on the table for measurements) I don't gouge the surface. Once finished a dolly should suffice. I also have two lazy Susan's from Woodcraft to help with finishing.

The weight of these are:
Empty cabinet = 110lbs
Base = 20lbs
Drivers = 30lbs
x-over + misc = 25lbs est

So the total should be ~185lbs. The empty cabinet and base at 130lbs is manageable but much bigger and I don't think I would be hefting them up on a table. :E

The blankets I got from Home Depot.

[JonMarsh]

Cool- I've got a Home Depot much closer than the nearest U-Haul.

[sdl2112]

I'm finishing up testing and creating driver files. With the test cabinets I only tested the sealed version. Here I tested both sealed and ported to verify low frequency alignment. This involves merging the near field driver and port with the far field data. It's not perfect but gives a visual of an estimated anechoic response.

To test two drivers in the same enclosure space I created a blanking plate to test on axis with one driver for the far field data. Another way is to test the mid point with both drivers but doesn't give as accurate high frequency data as it's slightly off axis for both drivers. For the near field response both drivers are installed measuring only one as that amplitude is only relative.

The sealed data is with the standard 22W8857 drivers. For the ported version I added the bucking magnets. This lowered the driver Qts from 0.438 for the standard driver to 0.354 with the added magnet. This allowed me to try sealed and ported with the same cabinet volume.

Here I show measuring far field with the blanking plate.


Here's a shot of the added magnet.


Next is the response and impedance of the sealed and ported versions with the same input voltage (both drivers parallel connected). The green curves are sealed, red is ported. The tuning frequency is 25Hz.

[sdl2112]

I finished the base skirts. I applied 4-5 coats of matte clear enamel. It should be a pretty durable finish. I installed one to see how it looks...:T

[cochinada]

Beautiful craftsmanship! :T
What is this blue line or ribbon for?

[JonMarsh]

Curious what kind of feet those are..

[Steve Manning]

Clear coat turned out nice Scott .....

[scottvalentin]

Wow those bases look great! Nice job!
What are you using for your measurement set up? Need to add items to the Christmas list! Is a Umik-1 a good choice?

[Face]

Beautiful craftsmanship! :T
What is this blue line or ribbon for?

Appears to be painters tape.

Curious what kind of feet those are..

I believe I've seen those at Rockler, but their site is down ATM.

[sdl2112]

Beautiful craftsmanship! :T
What is this blue line or ribbon for?

Thank you Joaquim...the blue line is just painters tape to protect the edges during assembly.

[sdl2112]

Curious what kind of feet those are..

Those are Slipstick CB505 furniture feet...see post A Day to Remember

https://www.htguide.com/forum/showthread.php?42901-Introducing-the-SSA-WG&p=607973&viewfull=1#post607973

[sdl2112]

Clear coat turned out nice Scott .....

Thanks Steve...about time they're done. The bases and skirts took on a life of their own...:E

[sdl2112]

Wow those bases look great! Nice job!
What are you using for your measurement set up? Need to add items to the Christmas list! Is a Umik-1 a good choice?

Thanks Scott.

I'm using a Behringer ECM8000 mic and 802 mic preamp. For software I'm using Soundeasy. I'm not familiar with Umik-1 but it appears to be the same microphone. For measurement I see a lot of people using Dayton's Omnimic.

[Steve Manning]

Thanks Steve...about time they're done. The bases and skirts took on a life of their own...:E

I can so relate to your pain ..... the Al trim fell into the same category. It has drug things out forever it seems.

[sdl2112]

Yep...I saw you fighting with the Al trim also...though there is a silver lining. Once done it makes the project all the more rewarding. :T

[Steve Manning]

Yep...I saw you fighting with the Al trim also...though there is a silver lining. Once done it makes the project all the more rewarding. :T

Very true ..... if nothing else it was a good lesson learned on what not to do next time around.

[jim1961]

Next is the response and impedance of the sealed and ported versions with the same input voltage (both drivers parallel connected). The green curves are sealed, red is ported. The tuning frequency is 25Hz.
[ATTACH=CONFIG]26138[/ATTACH]

Is this what you were expecting?

[sdl2112]

Yes...The ported design with the extra magnet it a little more efficient. It is also a little more extended. Similar to what is predicted by Unibox. The raw curves may look quite ugly to some but this is reality. No funny business going on. In view of baffle step loss and metal cone break-up it's logical. Next the fun part comes shaping this into a smooth response to match a target transfer function.

[jim1961]

You might consider doing the same test at the listening position. Then calculating a differential curve. Then you will know what kind of room attributes, namely bass compensation, the room will provide.

[sdl2112]

I'll look at room interaction once the proto crossover is up and running. For now I'm keeping them separate.

[sdl2112]

Proto2 crossover development is underway. I changed the Duelund target slightly. The center frequency is now 825Hz with an Aleph value of 3.1. This places the crossover frequencies roughly at 300Hz and 2.3kHz. I also moved the system target SPL down 1dB to approx 87dB.

So here's the new Duelund target graph.


Here's the new proto2 schematic. It looks very similar to my proto1 version but I really had to start from scratch. All the graphs below are for the ported version. It looks like I can get away with just changing the series L to the woofers for the sealed version and a tweak to the woofer conjugate impedance filter.


Next is the system SPL and impedance responses with all drivers at 2M, 15 deg off axis.


Target vs Actual at 2M on-axis


The mid is now in phase showing the reverse null at the crossover frequencies. Again this is at 2M and 15 deg off axis...this took some time to get right. :B We'll see in real life...

[Evil Twin]

I was beginning to wonder if I needed to provide additional motivation for you to complete these updates, but I see you have anticipated my wish, with some credible results to post.

I'm also curious at to the driver alignment and acoustic offset, and how that models in to your crossover; I've often found that enough of an issue with "conventional" drivers in a Duelund that in the end I moved away from the Duelund approach for drivers with typical Z axis time offsets for the woofer and midrange. With the Cell drivers I'm encouraged to explore this option again...

[sdl2112]

The motivation was there...I was caught up in the details...what if this... what if that...:roll:

The way I proceeded was; measure all drivers on their axis at precisely 1M to the baffle. I didn't try to measure the acoustic offset. The relative time of flight is contained in the measurement. Saved each driver file. From there I optimized each driver response to the Duelund target imported from the excel spreadsheet.

Displaying all drivers at this point in the "CAD" screen they are coincident with each other. No additional offsets are included. From there I proceed to the "System" screen where the drivers are placed on the baffle with the correct dimensions to account for the additional offsets. My system measuring axis is midway between the midrange and tweeter. With the waveguide on the tweeter the acoustic center to the midrange is close. This is evident in the original phase plots from the CAD screen. The midrange to woofers was not close as expected.

From there I manipulated the response slightly to bring the phase in-line. Mainly the response tails where their amplitude differences don't influence the overall response as much.

I'll show more when I get back to my home computer. It's a very interesting learning experience.

[sdl2112]

Some more measurement detail...

Here's a screen shot of the system panel layout. The cross-hairs show the mic position. In this case it's on-axis at 2M centered between mid and tweeter.


The next graph shows the reverse nulls at 1, 2, 3, 4M distances, on-axis.


I broke up the phase plots for clarity.

Here are the two woofers. The lower one is in green, upper in red and the sum is blue.


Now the woofer sum in red and mid in green.


The mid in red and tweeter in green.

[JonMarsh]

It looks like you've got very good phase tracking across the board, which is what should occur with a successful implementation of a Duelund.

Do you have a pair working? How do you like the sound with these drivers?

BTW, PCB's and more parts (for 1.8dB version) arrived early this week, but other things have kept me pretty busy..

[sdl2112]

I don't have them up and running. I haven't committed to the crossover values yet. My plan was to use lower cost components for the initial verification but I'm just itching to go for the premium components and adjust (i.e. shell out more money :roll as necessary.

I wasn't sure of a few things.

One was when I merged the near-field (+ baffle diffraction) with the far-field data on the woofers. There was a difference in slope in the 200-500hz merge region between the two. If they cross and are not parallel you don't know where to merge and the relative amplitudes between the two becomes arbitrary. I initially relied on the predicted low frequency response as a guide but I had my doubts. The far-field data was taken with a 6ms window thus there are not many samples (1/6ms = 167Hz). You want to merge where the near and far are both accurate so in this region they should have similar slopes. I re-ran the far-filed data gradually bringing the mic closer to the speaker watching to maintain the baffle step profile. Once closer I was able to extend my window for better low frequency accuracy. Now the merge region became more apparent. This lowered the low end by a little less the 1dB.

The other question was target SPL. You really have a range of target system SPL the can be obtained. It really comes down to the relative bass level to the mid/high frequencies.

Below are crossovers for target sensitivities of 87dB and 89dB and the response of the two. I see other designs with parallel 8 ohm woofers at 90dB or above. The Sony SS-AR1 uses the same woofers as mine and Stereophile measured 90dB. Wilson Audio is also typically over 90dB as well for their Watt/Puppy..etc. If I shoot for full 6dB baffle step, 87dB is about correct; for 4db, 89db is about right. With a general downward tilted response common, a slightly elevated low end seems ok or desired. Even with the Voicing circuit Jon has offered, it seem like a good idea.

Comments are welcome...

Here is the 87dB target...only slightly changed from earlier


Here is the 89dB version.


Comparison of the two.


Next on the list...Finishing shootout.:W Well maybe not a shootout... I'll try a couple and pick what I like best.


BTW Jon...no hurry on the voicing PCBs...as you see I have a ways to go...

[JonMarsh]

Mmm, you going to try those all out on scrap pieces you have lying around? Are you spraying or wiping/brushing? I can vouch for GF HP sprayed, and GF Seal-A-Cell wipe on.

I'm thinking, realistically, with the right technique, you should get good results with any of these... sometimes it just boils down to what application technique you prefer or have available...

[sdl2112]

I've tried all except the Tung oil products. Some of them are plan b, c...etc. My goal is a close to the wood finish. I don't want a film build up. The Arm-R-Seal is quite good on it's own but the application can be a little difficult for a multi-sided horizontal/vertical surfaced project. You really want to brush a uniform wet coat with overlap.

What I want to try is a wiping varnish blend of pure or more likely polymerized Tung oil, Arm-R-Seal and mineral spirits or naphtha. This gives me great control and flexibility. It will be a wipe on, soak...wipe off process. No worries of brush marks or runs. Down side is more coats, longer drying time and not as durable. Durability is not really an issue as this is not a piece of furniture. Also the last coat or two I can add a higher percentage of Arm-R-Seal.