My DIY speaker project for 2 ch.

gatemplin · December 2004

Thank you Xavier,

The front and rear baffles were made from 1 1/4" Beech butcher block. I wouldn't use it again because it doesn't take screws well. All of the other sides and all of the internal bracing was 3/4" plywood, the best stuff I could find at HD. For the finished version I think I will go back to MDF. It tastes bad, it makes a mess and it's heavy, but the final result is better. I think I will do oak veneer on the finished versions but I am still deciding.

The Extremis drivers have been built but probably wont ship to me until late January. Hopefully Jon will get his sooner so that he can start on the crossover design.

gatemplin · December 2004

Originally posted by Vr3MxStyler2k3
Is that grey fridgerator in the back... the DIY subwoofer?

If so... :eek: :eek: :eek: :eek: :eek: :eek: :eek: :eek: :eek: :eek: :eek: :eek: :eek: :eek: :eek: :eek: :eek: :eek: :eek: :eek:

Looks REALLY good man...

Thanks Sid,

Yes it's the DIY EBS Tempest. It's 49.5" tall. Here is a pic of the side. You can also see the port and Biwire posts on the speakers.

gatemplin · December 2004

Here is a pic of the business end. I got kinda sloppy down here with the paint. The driver is the 15" Adire Audio Tempest and there are two 4" flared ports tuned to just under 16 Hz.

audiobliss · December 2004

Man, and I didn't even see that thing.....:eek:

Amazing looking sub, though. Bet it sounds awesome!!

VR3 · December 2004

Thats not a subwoofer...

Thats like... a bomb...or an earthquake... something dangerous...

You got some SKILL dude... VERY nice...

gatemplin · January 2005

Thank you lads.

Just another update. The Extremis drivers have shipped and I expect to get them next week. Jon might also have the crossover done for next week. Jon is going to do a fourth order Linkwitz riley in the 1.5k-ish range and then modify the standard 24 dB/octave attenuation with the Cauer Elliptic filters to achieve much greater attenuation. I think he is going for about 48 dB/octave or maybe greater. This is really frickin hard to do. He is then only person I know of who can do it. I'm lucky that he is helping me with this project.

Tomorrow I will be modifying the LPG tweeters. Jon has put up a tutorial here
It made a huge difference in the impedance plots. I cant wait to try it.

The results from MarkK's Extremis tests should be in soon. He is a very well known amateur tester with a very good Praxis setup. This will help with the design.

I realized I never posted a link to the official thread. http://htguide.com/forum/showthread.php4?t=9891
It's long and confusing but I have learned a lot about design from Jon.

Here is a pic of the stuffed pole piece. I think that maybe the magnets are so big because these dont use ferro fluid to cool the coil and increase the flux density:

gatemplin · January 2005

And here is a pic of the tweeter after the grill and diffuser have been removed, the grill will be removeable on my tweeters:

Sorry I screwed up, next post.

gatemplin · January 2005

LPG

steveinaz · January 2005

Fantastic job buddy. Those are gonna sound awesome!

gatemplin · January 2005

Thanks Steve.

I finished modding the tweeters today. It takes a steady hand and it is easy to break them. I had a couple of close calls.

Jon has been working on the crossover. Here is a FR graph. This may look ragged but it is early and this is high resolution and not smoothed like most graphs to make it look good. Ignore the low end response.

Here is a early version of the crossover. I think it will probably cost over $300. More than the drivers.:) It is 8th order LR and crossed at 1.4kHz which is really low but that is OK since attenuation is so high.

I'm so exited that this project is almost over. I should be done by the end of the month!

gatemplin · January 2005

Yeah baby! I just got my Adire Audio Extremis drivers! Time to do some excusion tests!

They look great, huge frame and these things are heavy. Can't wait to hear them.

audiobliss · January 2005

awesome....keep us posted!!

gatemplin · January 2005

I tried some excursion tests and boy does this thing move. It has 3 cm (over one inch ) of total movement for a 7" driver. After I finish my chemistry project today I'll throw these in the test cabs with a test crossover (not the one above) and give them a workout. These are more like 7" subwoofers than midranges.

Hopefully Jon will have the crossover finalized by next week or sooner and I can order the parts. I'm using 14 gauge air core inductors and polpropylene capacitors from www.solen.ca .

gatemplin · January 2005

A couple of things have changed with this project.

I sanded the white boxes and painted them with spray on truck bed liner. It turned out pretty well considering how cold it was when I made them. The truck bed liner is very durable and forgiving, but it takes a long time to dry and they stink for about a week. I would suggest to anyone thinking of taking this route to try the roll on liner.

I think I will keep these boxes for a much longer time than I first thought. They aren't my best work but they are really solid and I wont have the time or money to build new ones (cost would be around $200US). This is my big project for the year.

The crossover has changed and will change again probably. This second version is Linkwitz Riley 8th order crossed at about 1.4 kHz. The graph is raw unsmoothed and the low end is irrelevant. Jon is planning on working on the crossover on the weekend. You might notice that these things are about 80 dB efficient. That is because the Extremis is a midwoofer (85dB efficient) and then Jon is using about 4 dB of baffle step compensation, which is pretty aggressive, but the bass will be incredible for the size. These will need a ton of power.

Thanks for following the project and if anyone has questions, or wants to see distortion or waterfall tests etc just ask.

gatemplin · February 2005

Another weekly update. I finished the cabinets and I'll have some more pics up this week. There is still no crossover but just for fun I ran some test tones on the cabinet. There is real deep extension thanks to the low resonant frequency and big box. There are noises coming from the fireplace and the dry wall shakes, with just one 6 and 50w. I'm very happy so far.

Obviously there is a null in the middle, I will try and get a better chart up. This is the speaker sitting on the floor in the middle of the room, the mic 1m away.

gatemplin · February 2005

I thought measuring nearfield in the middle of the room would maybe avoid some of the room affects, but apparently not.

I tried another sweep with the speaker sitting in the same place as I put my sub and measured from the listening position.
This turned out much better. I am thinking about EQing these with the BFD and using them as subs until I get the crossovers done.

jdhdiggs · February 2005

Not bad lower end for a large bookshelf.

One issue I see is how sensitive is your crossover to the values you listed. A lor of those values don't readily exist (to my knowledge) and you're going to have to buy a lot of components to get the "exact" rating. You may want to go back and see if you can simplify some of the components to more traditional values. Have you found a way to get around this?

gatemplin · February 2005

Originally posted by jdhdiggs
Not bad lower end for a large bookshelf.

One issue I see is how sensitive is your crossover to the values you listed. A lor of those values don't readily exist (to my knowledge) and you're going to have to buy a lot of components to get the "exact" rating. You may want to go back and see if you can simplify some of the components to more traditional values. Have you found a way to get around this?

Good eye. Yes a lot of the values dont exist. And yes this crossover is sensitive to the component values because of the steep attenuation. Some components need to be more accurate than others. This is also not likely going to be the final crossover.

My guess is that the crossover point will be lower, maybe 1.2 kHz. Jon is doing some more distortion testing on the Extremis and the LPG to see how low he can go. There are energy storage issues in the cone or suspension at about 1.6 kHz in the Extremis as you can see in the waterfall below.

And he will try and get the component values closer to off the shelf components. I will still have to parallel some capacitors and trim some inductors I suspect.

Here is an example of Jon's work. He is very thorough as you can see.
http://www.mclink.it/com/audiomatica/img/gpphoto/Hancock/slamm.htm
As I have said before. I'm really lucky that he is helping me out.

jdhdiggs · February 2005

I figured you were going to be sitting around with a fluke meter and Pspice tring to make a circuit of those values. I would think that those XBL drivers wouldn'e be THAT sensitive to the crossover range so I would be very tempted to simulate the easiest to make X-over and see how it goes.

Oh, and now you have me thinking DIY again... Audio ADD is horrible.

gatemplin · February 2005

Originally posted by jdhdiggs
I figured you were going to be sitting around with a fluke meter and Pspice tring to make a circuit of those values. I would think that those XBL drivers wouldn'e be THAT sensitive to the crossover range so I would be very tempted to simulate the easiest to make X-over and see how it goes.

Yeah they aren't too bad. The motor is very good, but the soft poly cone is the source of most of the problems. We are definitely blowing the problems out of proportion, but nothing is ever "good enough". It's all part of the fun. The philosophy behind this little design is to keep driver spacing well below 1 wavelength at the crossover, very low THD and IMD, excellent off axis and on axis reponse and as flat a response as possible. The trade offs are the huge size and terrible efficiency (80dB) and the cost of the crossover.

If I were doing the crossover all by myself I would have gone with something much more simple. Like 2nd order Linkwitz Riley. I am not even close to being able to design my own cauer elliptics like this. Maybe in 4 years though.

Oh, and now you have me thinking DIY again... Audio ADD is horrible.

Excellent. That is one of my goals in starting this thread. Spreading the disease

.

Willow · February 2005

are thsoe the speakers you will bring with you ? If it's still on for sure we can try them out at my place, see/hear what a dif sound they wil have. NICE !!

gatemplin · February 2005

Originally posted by anonymouse
Have you considered an active crossover and bi-amping? The crossover would undoubtedly be cheaper, more accurate and tunable.

I will be bi-amping because of the poor efficiency. Jon could have more easily done an active design, but he prefers analog passive.

He explains here and you can see one of his active designs in the link a couple of posts up:

http://htguide.com/forum/showpost.php4?p=148339&postcount=8

I would really like to try active design someday, but I am not there yet. I still have plenty to learn about passive. Working with an expert like Jon is a great learning experience.

gatemplin · February 2005

Originally posted by Willow
are thsoe the speakers you will bring with you ? If it's still on for sure we can try them out at my place, see/hear what a dif sound they wil have. NICE !!

I dont know if they will be ready JF. I am hoping they will, but maybe next time. PM sent.

Current pics of the speakers should be up very soon.

jdhdiggs · February 2005

Originally posted by gatemplin
and terrible efficiency (80dB) and the cost of the crossover.

80dB? I hope not.. You're going to need bridged crown K2's to get that anywhere. Are you sure it isn't higher? If it's that low, have you looked at how loud you can realistically play music at? (Power handling and such) I would think that with the box being so large, you would be much more efficient, no less unless the crossover is eating the power.

I mean, for 102dB peaks, your needing 128W.. Yikes, that would turn me off on the whole thing since most amps "last" watts are much "crappier sounding" than the first watts.

gatemplin · February 2005

Originally posted by jdhdiggs
80dB? I hope not.. You're going to need bridged crown K2's to get that anywhere. Are you sure it isn't higher?

Yep it is really that low. The woofers are about 85 dB and we are using a healthy 4-5 dB of baffle step compensation.

I listen at about 80 dB average with 85-90 dB peaks. I will have about 50 continuous watts per driver, those are vintage SS or HK watts. I am also only about 6 feet away.
Down low these are very efficient because of the EBS alignment.

It's funny you mention the K2s. Jon also was one of the main designers on those Crown K amps. It would be pretty cool to have speakers and amps designed by the same guy.

jdhdiggs · February 2005

That's cool, I just couldn't picture speaks being useful at that low of efficiency. I guess not everyone needs to go deaf...

How much power can those XBL's take in your alignment?

gatemplin · February 2005

The XBL2 drivers are rated at 100w continuous and 300w peak, with this crossover it will be higher. How much higher I am not sure. I would guess about 400w per side.

An MTM design would be more practical for most people at 86 dB per, but the box would jump to like 60L and at $100 per driver you are looking at $400 just for the woofers.

gatemplin · February 2005

Here is the first pic I took. This is with the grill on the tweeter. They are obviously mirror imaged. The driver placement was designed using Baffle Diffraction Simulator. Nothing about these speakers is done for esthetic reasons. You can see the truck bed liner. I think they turned out quite nice.

gatemplin · February 2005

I forgot to load the pic.

Here is the front with the grill.

gatemplin · February 2005

Here is a pic with the tweeter naked.

My DIY speaker project for 2 ch.

Comments