2-way floorstander beginner build log

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hey everyone,

Let me first introduce myself since I haven’t posted anything around here yet. My name is Joost and I have little experience in speaker building (yet, hopefully), but I do have a lot of reading under my belt, both on this and other forums and some of the more recommended books on speaker design. My brother studies electrical engineering and has done a project on designing a speaker so he does have access to all necessary measuring equipment and a fair bit of know how. Lastly, I do have a background with woodworking as a hobbyist so on the cabinet making side I am quite confident.

A few years back I build a pair of small full range speakers that were great fun to make and to listen to. However, with 4-inch drivers they simply don’t go very low or loud. So the new plan is to build a pair of modest 2-way floorstanders.

The inspiration for this build will be from Troels Gravesen, specifically the Illumina 66 speaker (Illumina-66). I really love the look of this speaker, and maybe more importantly the girlfriend also sort of likes them (but why do they have to be so big??). And I feel based on many forum posts that a 2-way is a better place for me to pick up speaker building again than a 3-way. The problem I have is that this speaker is way outside of my budget (approx. 2000 euros in parts alone). Aside from this I would love to design my own crossover.

The goal of the speakers is 3-fold. First of all they will be for listening to recordings at a specific listening position, secondly they will be used for home theater purposes, and finally they should be able to hold their own when I throw a small party in my apartment (nothing to serious). The room is an absolute mess sonically speaking as far as I can tell, a lot of large flat surfaces, proper placement of the speakers will be difficult without having interference of a couch, and a lot of corners. However in this regard I have very little say in what happens; you know how it is….

Sources will be the TV, a raspberry pi DAC (Spotify playback over wifi), 3.5mm jack in from mobile phone. The speakers will be driven by the myref Fremen edition chipamp.

So the plan:
  • Two 2-way floorstanders with a 5”-7” + tweeter setup
  • Cabinet: approx. 30 liters (1.1 cuft) vented, relatively flexible in this regard but WAF is important.
  • Relatively low sensitivity. For at least the foreseeable future these speakers will provide all the base so I am happy to sacrifice some sensitivity for more low end output.
  • Custom cabinet design, including facetted front baffle
  • Custom cross over design, preferably not too complicated since it will be the first one I design myself. However measurement equipment will be available in the room.
  • Budget: approx. 200 euros per speaker for drivers, another 300 euros for cabinet work/cross over components etc. (so a total budget of 700 euros). Lower is never a problem of course :D but also in this regards I am relatively flexible.
Now for the driver choices. I must admit I am overwhelmed by the amount of choices out there. The original design by Troels uses the Illuminator 18WU/4741T00 and D3004/660000. He states the following about the woofer “And not least important, the 18WU does not loose articulation at low levels! This is a quite unique feature of the 18WU drivers. Too many 6" drivers loose grip at low levels due to high-loss suspensions. High mechanical Q is one of the things we have to look for.”. This in combination with the fact that I am looking for something which is relatively simple to cross over was the starting point of my search. So I am looking for: a relatively flat response up to at least 2/3 kHz, deep base extension, high mechanical Q, large piston surface area, and a large x-max value.

So below some options, I modelled all of them in Unibox by setting the volume of the box at 30L and tuning for a flat response:

  • SB Acoustics 6” SB17MFC35-8 (57 euros): A driver which I have seen recommended a lot here on the forum for starting speaker builders, and it seems to tick all the boxes. However I do have some head room in terms of budget. Unibox f3 = 34Hz
  • SEAS prestige 6.5” CA18RNX-H1215-08 (91 euros): slightly larger piston area (135cm2 vs 118cm2) however the FR seems to be less flat than the SB17, there seems to be distinct increase from 500Hz upward. When modelled in Unibox for a flat response f3 = 37Hz
  • 6.5” Satori MW16p-8: (140 euros): more in the upper part of my budget , slightly lower piston area again 119cm2. Response is more similar to the SB16 than the seas prestige however this one also has an f3 = 37Hz when modelled.

These are obviously just a few of the possibilities out there and I would love to hear any suggestions for a good driver in my price range. For the tweeter to me it seems logical that I will make that choice after choosing a woofer but here too I’m not really sure so if anyone disagrees do let me know!

The idea of this thread will be to show the progress I make on building a set of speakers. However since I am a beginner I hope to get some advice/answers to some questions as well! Hopefully by the end this will be another good thread for someone to use as a resource. I do have a full time job and my weekends are also quite full a lot of the time so progress might be slow at time…

For now I would love to hear suggestions on driver selection and of course any general comments/remarks. Please feel free to be as harsh as you want to be if I completely overlooked anything or the plan in general is **** ;)

Cheers,

Joost
 
I have experience with the drivers you mentioned. Seas CA18RNX were disappointing. Comparatively even the SB17MFC will blow the seas away with the Satori MW16P being really your only choice.

Worry not about the 37hz modelling as my SB kit speakers Ara containing the MW16P are also tuned around there but measured almost flat to 20hz response in a 700sqft living room.
 
Hi Joost,
I was looking at a couple of the woofers you are interesting in. It appears they will run into Xmax issues in a vented enclosure at moderate listening levels (i.e. 95-97db @1meter). The Satori MW16P performs the best of the two that I modeled. Note: I only tried the Satori and SB. The SB 17MFC just does not have enough power handling capabilities for home theater use or as a party speaker. No matter what, I would be looking to add a sub down the road. Just my 2 cents FWIW.:)

Have Fun,
Rich
 
Thanks for the responses!

One thing to keep in mind that for music playback what I have right now are a pair of sub 3 liter 4 inch full range speakers. And that the tv itself is serving as the HT speaker. So I'm quite convinced that this will improve the overall sound quality.

After looking around some more on the SEAS it seems they are not very well regarded so I won't be using them at least. And @lozjek I hadn't noticed they weren't measured in an standard way thanks for pointing that out! I'll have to be more critical when looking at these figures.

In terms of volumes and power levels, I live in an appartment complex. It is quite new (build last year) so the isolation is good, but it is still an appartment complex so volume won't be through the roof anyway. However an alternative would be to use the 7.5" Satori's, in the Unibox sim it seems these can be driven to 103dB which seems more than adequate :D. A sub would be very nice however I think that I will be relatively slow to build this pair and it will take even more time after that to start again on a new project.

Lastly I found these drivers after looking through quite a lot of them, but as I said I am mainly overwhelmed by the amount of choices out there. If anyone has other suggestions for drivers I am more than open to them!

Cheers
 
Hi Joost

I am surprised that no one has said this yet, but you would be far better off using a proven design. Quicker, simpler, cheaper.

As you know, Troels uses very expensive components, which is a big risk when you have not heard the end result. However excellent they might be, they might not be to your taste.

I have built Scanspeak revelator designs and have enjoyed them.

Since then, I have come to realise that you can get similarly excellent results with cheaper drivers if designed correctly.

Check out Jeff Bagby designs. Full kits are available on Meniscus Audio website. In you budget range, I would look at Docere and Mandolin.

N.B.The Docere has a tower design as well as a bookshelf/standmount design, but you can always build a bookshelf design into a tower cabinet, by only utilising a portion of the volume, and leave the rest separated and unused.

It should be simple to mimic the elements of the Illumina 66 that you like in any construction. Baffle layout and width need to be maintained from the design, along with cabinet volume.

Docere Full Kit (Pair) - Meniscus Audio

Mandolin Loudspeaker Kit - Meniscus Audio
 
Last edited:
Hi Joost,
I do agree with @dublin78 about building a proven design. This would especially hold true if it was going to be your final build. But, if you intend to continue on the speaker building journey; its not a bad idea to formulate your own design from scratch. You mentioned in the first paragraph of your introduction that your brother has access to measurement gear which becomes necessary to achieve outstanding results (unless you have golden ears).:D My thought is you will learn much more about the in's and out's of speaker design and building by rolling your own. Just beware it can become quite addictive once you start down this path! All my wife can do is look at me and shake her head. At least it is the only addiction she has to worry about. Although anything else would be far less expensive.:)

Best,
Rich
 
Thanks for the suggestions! I sort of expected that to be the first response too ;)

I do understand that using a kit or proven design will (almost certainly) give me a better result. But... and I think this has been said quite a lot of times on these forums before the largest part of the fun for me is the process of learning/expanding all these skills in different fields. woodworking, acoustics, electronics, design etc. etc. So that is why I would like to go through the entire proces myself (with ofcourse the help of the omniscient internet), even knowing that I could do it quicker, simpler, cheaper, and very likely better using a proven design.

That being said i am certainly not opposed to stealing good ideas from proven designs and the 2 you send do inspire confidence that with relatively cheap components I can get a good result!

EDIT: Is there any such thing as a final build? ;)
 
Last edited:
Joost. I admire your quest and wish to enjoy the journey.

My reluctance to do the same is based upon the knowledge that the more that I learn about loudspeaker design, the more variables I discover. It is not as simple as crossover slope and amplitude. The best designers take into account far more complex issues that I can't get my head round. Yet.

Good luck.
 
Pff those write ups were quite something and send me in another reading frenzy... especially off-axis response and especially interactions between the drivers is something i can't get my head around... oh well

In the meanwhile I have been tinkering in all kinds of different simulation tools. Vituixcad is quite amazing I must say. I started out using Unibox to simulate the Satori MW19P-8, and I think I found a nice sweet spot with a volume of about 30 liters, however this was as far as that could take me.

So I turned to VituixCad... After some less than fruitfull experimentation I had a look at the guide :rolleyes: and everything started to make some more sense. The pictures below show the results!

What I did was the following. First I imported the FRD and ZMA data from the manufacturer (drivers used were: Satori MW19P-8 and the SB26ADC-C00-4) using the SPL trace tool. Then I recreated the enclosure from Unibox which did give slightly different results but it happens, the final volume was 30L tuned to about 36Hz.

Then I made a quick mock-up of how I envision the enclosure and let the program calculate the baffle step response. I set the listening distance at 30 cm because then the response was approx. -3dB which I thought I had read on Troels Graven his site was a more appropriate figure for actual response in a room than the theoretical -6dB.

Then I used the merge tool to combine the low end which was modelled using just the T/S parameters with the high end which was just the original frequency response graph from the manufacturer. This frequency response and impedance file I used in the crossover design. For the tweeter I simply used the impedance, and frequency response curve from the manufacturer.

To design the crossover I started with a theoretical 4th order Linkwitz-Riley crossed over at 2kHz. Then I added an L-pad to the tweeter and a Zobel network to the woofer. And then I started tinkering around (both manually and using the optimatization tool from the program). Anf in the picture you can see the final response.

I know I need to do measurements on the actual speakers to do all this properly but I must say that the results do look promising. Maybe even promising enough to actually start buying the components.

I'm quite sure I have missed quite a lot of stuff, so please do feel free to enlighten me if you feel like it. One thing I am already aware I have not addressed is the acoustical off set between tweeter and woofer, but I'm thinking about physically off setting them to solve that, but I'm not sure yet.

Let me know what you think!

Cheers
 

Attachments

  • 2-way MW19P8 SB26ADC 0004 overview.png
    2-way MW19P8 SB26ADC 0004 overview.png
    195.4 KB · Views: 292
  • bafflecorrection.png
    bafflecorrection.png
    79.5 KB · Views: 257
  • box.png
    box.png
    241.8 KB · Views: 251
  • merge.png
    merge.png
    119.8 KB · Views: 260
The Satori MW19P-4 with a 1650Hz LR4 circuit works well with the 1.1" SB29RDC-4 dimple-dome tweeter with either BW3 or LR4 circuits at 1650Hz. The extra area and durability of the 1.1" dimple dome tweeter is necessary for 1650Hz. There are several low cost Bi-amplifiers with active crossovers capable of 1650/1650Hz which will also perform well with these two drivers.

A tall'ish cabinet designed with a MLTL Mass Loaded Transmission Line alignment(shape) will both extend the deep bass, and smooth any internal resonances.
 
Last edited:
Good point about the BSC, I added a correction for this in the next iteration. You can see this in the picture below that the aside from a small peak at 3.7kHz, and a slight dimple at 1.2kHz the response is pretty flat. Also phase coherency seems to be excellent.

This time around I did include offsets, I based these on a sketchup model I made (156mm vertical offset, and 12mm in the "z" direction), and made the assumption of a 20mm acoustical offset if they were mounted on a vertical baffle. The 20mm I took because of a article I read at some point on measuring acoustical offset, where they recommended 20mm as a first guess for the input.

As far as the 4-ohm version and the other tweeter that does sound good, because it would solve the issue of the peak at 3.7kHz. I will give that a try!

As always, let me know what you think!

Cheers,
Joost
 

Attachments

  • crossover 01-02-19.png
    crossover 01-02-19.png
    195.5 KB · Views: 214
  • cabinet.png
    cabinet.png
    120.1 KB · Views: 142
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.