Modula MKII TMM Build Thread

[ligonk]

Hello, before I get started int he build thread, I need to ask a very important question:

I have received my tweeters and I proceeded to remove the faceplate to prep it with the horns I have already modified (that can be seen in another thread) but immediatly the freaking VC pops out of the gap and the whole diaphragm lifted out with the faceplate. I put it back into the gap correctly oriented but Jon Marsh had this to say:

ScanSpeak D2608/9130 and
Mounting plate details
Probably the next thing to do is to
prepare the tweeter for interfacing
with the waveguide and fabricate the
tweeter mounting plate. The
tweeter can be disassembled easily
with a 5/64” hex tool, even though
it’s probably a metric size. Take care
around the VC assembly, and try to
avoid popping it out of the gap and
disturbing the ferrofluid.

So my question obvious: how badly did i just mess up my $80 tweeter?!?!




P.S. i have plans and drawings and pictures and all sorts of fun things to post, but I just did this and decided to start the thread early, before I have any of it organised enough to post:E.

[benchtester]

...

So my question obvious: how badly did i just mess up my $80 tweeter?!?!...
.

I don't think it is a big deal since the ferrofluid is attracted to the magnetic gap. After all, Madisound sells replacement voice coils; so a swap is anticipated.

[sdl2112]

+1, I removed/reinstalled the VC on my D2608 without issue.

I'm looking forward to your build thread!

[ligonk]

Okay this makes me feel better. The ease in which that bugger popped out was stunning, to say the least.


Onto the Meat!

For starts, anyone who isn't familiar with the Modula MKII build should acquire the PDF that is floating around here as it is a seriously well-made piece of DIY Knowledge. Of course it is brought to us by none-other than Jon Marsh and I would like to thank him now for taking the time to prepare such a manual.

Secondly I'll mention that I am going for the TMM build, turnig the project into a 2.5-way system. The details for this are hinted at in the PDF, but most of the XO schematic he has gone ahead and prepared for us. The details that remain include the port tuning of the second woofer's cavity as well as box construction. I have opted to go for the XO as laid out in his design documentation, minus the L7, C8, R9 LCR as this is the least efficient cost/performance element, and it is indeed costly (without using electrolytic caps and laminate-core inductors, and that I may do later).

So to start I made some sketches of the dimensions. I do not have a large workshop, in fact all i have is a fold-down workbench in my garage. Because of this i have no installed tools like a table saw nor a drill press. My father has these, but he lives far-ish away, so in case of emergency i can access them. So my plan is to do as rigorous of a design job so that I can create a panel-BOM for each board i need cutting and be able to get all the cuts done at the local hardwood supply store. This is risky and doesn't allow for much improvisation, so I want to think it out ahead of time. Most of my cuts i do in the garage are done with either a skil-saw or a hand saw, and then the rough edge is laid down on a straight piece of wood and i use the flush trim router bit to true it up. So you can see the value in getting it right the first time.

Here is what I have so far for drawings:



Missing are the holes in the MDF backing of the front baffle, the actual thickness of that MDF needs more consideration (as it will dictate the mating of the tweeter faceplate tot he horn), depending on the final thickness of the front hardwood/LBL. The front panel will be either hardwood or ideally LBL depending on availability. The brace that Jon uses on his front panel, the one that does not stretch the whole way back I may omit, or make it a full brace. I also need to show the cross sections for the internal bracing, as only the edge shots are shown there. I was thinking of a solid piece of MDF with a single large-diameter hole cut in the center, leaving 1" material on the sides. That or i could go for 4 smaller ones, that would leave more center material. My finishing choice as of now is gloss black on the mdf after extensive sealing and to finish the wood/bamboo front plate in natural. Now this front plate only extends the length of the front baffle where the drivers are, and so would resemble Zaph's wave-guide TMM, except i'll go black where he went veneer and finished wood where he went black. I don't particularly want to mess with veneer and i have to keep the cost down.


The first place that i began construction on was the waveguides, as sourcing them seemed the highest risk to the build. I used the "6 inch" horns available from diyaudio.com thinking they may end up to small, but after measuring them and then milling them, I am convinced they are the same as the original, plus $5.50 to boot.




Next I went ahead and worked on creating the silicone gasket, experimentally on one. I had parchment paper, but no wax paper, but it ended up working just as well.





The horn has a sharp lip and overhangs the hole just a little bit, but it appears this gasket is acceptable?



This is the faceplate mated up to the horn after the gasket had been cut
Is there a good way of cleaning that up a bit? It is pure silicone, so i'm not sure how much harder it will get.

I have also purchased two little bamboo cutting boards to cut up and use as the tweeter mounting board. The only issue is, they are just 1/4" thick, and the ones in the PDF appear to be closer to 1/2". Now they are still VERY stiff and are made up of alternately oriented layers, but with extended pressure, I'm not 100% confident. Creating these will be the next step, maybe sometime today, depending on if the wife wants to hang out or not.

I have purchased two tweeters and a woofer as well as three horns just to get me started, but i have to trickle-buy most of this so expect the pace of this project to be slow with intermittent bursts of activity.

[ligonk]

Oh, and here is a bonus post, just showing some of the other DIY audio endeavors i have pursued:

My first set of overnight sensations, that started it all.



Finished in natural tung-oil.
With the XO board:



Then I told my father i could build him a little portable system with blue-tooth that would sound much better than those shitty iHome speakers he wanted to get:



And my in-laws her them when they were over and requested a pair as well. So here they are together:

[JonMarsh]

This is what will happen, you know, if you "advertise"- you'll never run out of things to do! :B

[ligonk]

I have just completed the little work i have materials to do. I just cut the bamboo tweeter plate.




I do not think I will need the corners off since i am making my own cabinet, but that's easy to do later. I guess i can drill the holes to attach the tweeter faceplate to the bamboo, since i do not have a recess with this thinner bamboo, aligning the tweeter plate to center will be more difficult.

still wish there was a good way of cleaning the silicone gasket up a bit.

[dar47]

Making good progress. If your not apposed to removing the silicone you could try the black RTV gasket. I put it on a day before mounting and it hardens enough that it still seals and them can be easily cut away with an exacto knife.

[ligonk]

I'm not, the silicone rips off cleanly enough. does the RTV get a bit harder than the silicone? like sandable hard? cause I guess, even with a brand new exacto, the silicone was too soft to cut cleanly.

[dar47]

Not sandable hard, but harder then the silicone and cuts a lot easier. I didn't try it but a small gouging chisel would even shape the gap well.

[ligonk]

Update after some more work. I stopped by our excellent hardwood lumber yard, Austin hardwoods here in Denver. I was looking for something hard that would yield a 10" by 25" front baffle per the drawing. I found this excellent curly maple board for a reasonable $36.



Can't see any of the figuring in this pic.

I held onto it for a week and then finally made it to my dad's to use his woodworking equipment.

First issue of the build reared its head when i found out that I apparently CAN'T use a ruler and instead of a width of 10 7/8, it was 9 7/8". So i settled for a finished board of 9 1/2", wand i don't expect much BSC loss with just a half inch. I also got to use his planer and drill press for the pin-holes to use the Jasper circle jig.

Here is a shot right afterwards with the figuring much more apparent. The recess for the horns ended up like <1/32" too shallow, so i'm sanding down the horn flange.



And here is tonight, after getting the rest of the holes in place. The first recess i cut i didn't tighten the chuck enough, and the bit slipped too deep. so i packed it with glue and sawdust, and will run over it again with the router set to the same recess when it dries. You can see that to the left of this image.



I didn't want to change the bit in the chuck so i could get a better second cut on the rerun of the recess, so i waited on the chamfer and round. The woofer recess has plenty of room for some sort of gasket material, i have yet to decide what.

I spent a few hours working out the dimensions of the MDF to cut and i realize i have a few important questions:

Currently, my plan is to separate the top woofer's air chamber from the bottom woofer. It then donned on me that this may not be the best course of action. It was my original thought to separate the bottom woofer's bass from the mid range of the top, so that it inherits a more clean character of a three way. But then, do i make the bottom chamber the same volume at it is now (in my current design, it is, but extra the volume of the tweeter assembly)? Then the third question i need resolving is, should i choose to separate the bottom chamber, what do i tune the bass reflex port to? My assumption is that it has the same tuning, but I just have not ran the WinISD simulation yet. I prefer to hear your experienced commentary on the situation first. Other than that, i just have to wait until, maybe this weekend to get the MDF and go get it cut at my dad's. These decisions do not change any of my cuts, as the inner-baffles that would separate the air chamber, i would just punch the large home in them.

[ligonk]

Another question i have been mulling about is the low DCR inductors for the two woofers. The given approach to these are the C-Core toroid solid core inductors from Jantzen with an unprecedented .068 ohms DCR.

Jantzen 3.3mh 14 awg c-coil toroidal inductor

These certainly minimize the series resistance! But what of the effects of saturation we all hear about with solid cores? Furthermore, I have read some research that indicates having vanishingly low DCR for your woofers can cause a loss of bass, as the amps usage of back EMF controls the woofer too greatly (can this even be greater than critically damped? Is there such a thing as too-controlled?). Another issue with the above is that i can't get the required 3 mH naturally, so would have to wind back this coil, something im not 100% sure i can do with the design and construction of the c-coil. With all that in mind i was looking at these sold by madisound:

Solen 3.0 mh perfect-lay inductors 14awg

With a respectfully low DCR of .42 ohms. Plus who can pass up a perfect lay?? Would this DCR introduce flabby bass that reduces the positive attributes of an air core inductor? Or are the positive effects of air-core just not that important when pressed with the deleterious effects of a high series resistance? What physical characteristics of an inductor lead to early saturation? Does the c-coil's massive 700-volt rating mean that it is basically free from saturation at any levels that this speaker would see? Does saturation in solid core inductors even have anything to do with voltage rating?

The ideal response would be one derived empirically from years of self-experimentation (that i plain-old cannot afford at the moment) with a link so some literature that explains this phenomenon for the physics-abled. In lieu of that I'll take any thing you wonderful folk have as far as advice. The air core, in this case, is only $10 less expensive; so cost, for once, can have no bearing in this trade study.

Thanks again,


Kyle

[ligonk]

Update time! Thread's a bit quiet, but I'll diligently continue my sporadic updates just the same.

First pic shows the MDF cut after i got it back from my Pa's place:


It just wouldn't be a build thread without the pile-of-MDF shot.

Next is a rather uninteresting shot of the early glue up progress, before I spent a week in San Jose Cali.





Here is the progress I made this weekend.



With a closer up shot:



The wire I'm using is the stuff BJC sells for regular speakers use. It the 10-gauge 5T00UP from Belden. It seems like solid stuff and at 1$/ft seemed a better choice all around than some of the other options i was considering, Madisound's offering being the chief competition. The only worry i have now is if my 25w soldering iron has enough oomf to solder on a bus wire. The are seen passing through the larger holes in the not-a-baffle but what you can't see is the air-tight pass through for the solid air-chamber baffles

At first i couldn't settle on a pattern of holes for the through-hole support panels. But im pretty happy with how they turned out (Part "D" in the last pic).

Glue is Titebond II. I seem to have made enough calculations and labor ahead of time that so far, I have not come across any mishap of any size, which is a unique first for me, I assume you! I have decided to go with the two separate air chambers, a la Krutke's WG TMM build, and I'm going to stick with the low-dcr coils, though I would love to do some research on what characteristics cause core saturation, and the effects thereof.

*Shrug* - not much, i don't have the stagemanship of some of the better build threads, but I'd be happy to answer any questions or take pictures of any parts not shown. This is a pretty basic build when it comes down to the cabinet, so I am just going to push through it a but. The interesting part will come with the tweeter mating later on.

[ligonk]

Wow it's been a while. I don't have much time, so I'm just going to post some build progress pics.


Here is the 3/4" 3-part foam from PE starting to go in. I realized down the line that I would have loved to have purchased another sheet :-/



Side on! It was easier to do the sides before the foam, I discovered. Sort of a big flaw here: I didn't seal the MDF with anything. Later on, the other dampening material i used used a full and thinly spread layer of Titebond II to adhere it to the wall, so on those sections, im not worried, but the sections with the PE foam only have that self adhesive tape for sealant, which may well be nothing at all. As long as i live in Colorado, I don't foresee any issues.


Skipped a bit with no pictures, but here is the bottom design I am using: 3/8" rabbet, 3/4" deep with wings and T-Nuts. 1/4-20x1.5" for the bolts (not pictured). This will have 1" rubber feet (1/4" tall) on it just outside the holes, overlapping the join. Finishing is after 1 coat, sanding and coat 2 (not sanded yet) of Deft's sanding sealer. I have a 12" steel dry-wall knife coming i the main with some original (312?) formula bondo. I have quite a few issues to clean up with bondo and i hear it works amazingly for MDF. Question for anyone out there: What surface preparation on Deft's sanding sealers should i use before bondo? I was just going to sand it with 100-grit beforehand.


And here is kinda where they are now. Both have the bottoms installed with a temporary installation of felt feet. Next work will be bondo and if its looking flat, primer, Maybe I will finally get around to starting on the front baffle one of these days. I start school Aug 25, so i was hoping to get to functional by then, but i still have to BUY the inductors and resistors, much less solder the beasts up! Primer coat will be okay for now. I just hope I don't fall into that trap of listening to unfinished speakers for 4 months, cause I can't bear to take them away for the weeks and weeks of finishing. At least my current pair are pretty decent sounding (until i hear the new ones, lol).

Jon Marsh: I have a pretty big question for you regarding Horn-tweeter integration, but I don't have the time to spell it out here, so watch this space

[ligonk]

So here is my question. One thing I noticed after looking at JK's Waveguide TMM and looking at the modula MKII TMM is the integration of the horn with the faceplate. The Seas tweeter has a flat faceplate, making it pretty easy to mate the flat backside of the waveguide to it. The scanspeak we are using has that annoying mini-waveguide that requires the use of RTV or some other creative solution to mate up. The thing I am worrying about is the shelf this creates to the tweeter. See below where the blue RTV makes the vertical ledge. I am worried that this ledge will cause bad diffraction and all around not work with the waveguide (despite many people having made these speakers and not indicating any issues).


So I was wondering if it was a good bit of extra credit to file the backside of the horn down so that it inserts ito the waveguide(doesn't have to fit flat, so long as the edge of the horn mates to the faceplate), like below:


Likw I said, no one seems to have any issues with the current implementation, but it just strikes as a potential issue. "Fixing" it could of course mess up the voicing of the box so there is that to think about.




And here is a quick update with where the cabs are at.


Some Post bondo shots.


Port integration


Looking smooth with the first layers of primer going on.

[ligonk]

Also, how do i prevent this?:





Its been ~20 hours since spraying, I'm going to wait three days before i try to sand again, i guess.

[kevinm]

Wet sanding will greatly reduce that since the water keeps it from sticking. But it's not good for a typical orbital sander to suck up the water.

[ligonk]

I have blocks i can hand sand with. I tried that actually and it worked well, except where i broke through the primer, i started raising some "grain" in the MDF and it started looking ugly. I guess i can skip 220 and just jump into 400 wet sanded, as its pretty f-ing flat now.

[ligonk]

Image after popping the grain with Amber shellac, sanding and then BLO, nice and thick!