'Perceive' Contruction Diary

[ShinOBIWAN]

I'm planning on documenting my latest DIY effort here and hopefully use this thread to pick the brains of others as the project progresses

I chose to call this latest project 'Perceive' because I hope it will offer a clarity and insight into music that is amongst the best available from commercially available loudspeakers.

Going with a 3-way, tri-amplified design with onboard triple mono amplifiers and all active Linkwitz 4th order network.

Just running through a quick synopsis of the components used for each speaker:

Drivers:
1 x Scan-Speak R2904-7000 Ring Radiator
1 x ATC SM75.150 3" Dome Midrange
2 x Seas Excel W22 8.5" Bass

Electronics and amplification:
1 x 3-way Linkwitz 4th order XO with transform,EQ and delay circuits
1 x Active XO power supply
1 x 16v Toroidal transformer for XO power supply

1 x Input/Output board with Balanced XLR and unbalanced RCA inputs
1 x 2 x 120w(8ohm) solid state amp for treble and mid drivers.
1 x Dual PSU for the treble and mid amps
1 x 120w(8ohm) solid state amp for bass drivers
1 x PSU for bass amp
3 x 18v Toroidal transformers for amp power supplies

All of this will be going in to a time-aligned, decoupled sat/sub type arrangement. With the Scan and ATC in the upper cabinet which sits on the bass cabinet containing the two Excel drivers.

Current candidates for crossovers frequencies are 400hz, 2.5Khz. These figures are subject to change after testing but are good starting points.

Most of the above components such as the Scan rings, ATC mid and all of the electronics and amplification I have or am awaiting delivery. The cabinets are in the late planning stages so are virtually ready to built. This will hopefully start in the next week or so. The only thing I haven't ordered or have is the bass drivers which leads on to a question.

Question:
I'm undecided on the bass drivers to use with the design, initial thoughts are the Sea's Excel W22. I'm limited to 8.5" drivers at the most because I'd like to keep the baffle slim.
What I'd appreciate from others are potential candidates to be used in the transmission line enclosure sub enclosure.

Will post details and other questions as the project progresses.

Thanks for looking.

[ShinOBIWAN]

Right pictures of the drivers now:

The R2904 treble units come as a matched pair and just to make sure you realise that you've spent £550 on a pair of tweeters they come in a presentation box - wow, good to know my money is going into sound quality and not useless trinkets :roll:


Whats this! Someones stole one

These things look downright dangerous with that extremely sharp phase plug in the middle. Wouldn't like to have these around with young kids otherwise there could be a nasty accident :? Build quality is faultless on these drivers. Very weighty too for HF units.

[ShinOBIWAN]

Here's a few early designs I've knocked up using CAD.



In the end I chose to go with non of these and decided on the decoupled sub/sat arrangement.
I thought I post these here though just to give an idea of the direction I'm heading in.

Once the final design is complete I'll post a schematic of it.

[ShinOBIWAN]

Active 3-way LR 4th order XO


Power Supply for the active crossover:


Amplifier input and output stage with balanced and unbalanced connections

[ShinOBIWAN]

Treble & Midrange Amplification and PSU

Dual mono amp with 2x120w into 8ohm. Used for treble and mid drivers


Treble and mid amplifier PSU 2 x 56V/2 x 14A continuous current capacity (2 x 70A peak) for the amplifier and 2 x 30V/2 x 0.2A for the input/output stage circuit.

[ShinOBIWAN]

Bass Amplification and PSU

Mono amp with 1x120w into 8ohm. Used for bass drivers


Bass amplifier PSU 56VA 14A

[wigginjs]

Heh, world class project you've got going here =). Did you design these boards yourself or can they be purchased some where? What made you decide to go with an analog active crossover as opposed to a digital one? One solution to your bass driver problem would be mounting the driver on the side of the enclosure instead of the front. If you're only using the driver below about 120Hz or so then it isn't important for the driver to face the listening position (anyone feel free to step in an correct me on this if I'm wrong). It's my understanding that these frequencies are below the point where sound becomes "unlocalizable".

I'm very excited about seeing this project progress. Your pictures are really excellent. Thanks so much for taking the time to document your effort.

[tiroth]

How did you decide on 4th order network? Was it because of your time-alignment requirement?

[ShinOBIWAN]

Originally posted by m0tion
Heh, world class project you've got going here =).

Thanks

Did you design these boards yourself or can they be purchased some where?

I wish I could design PCB's like that!

They were bought from here:

http://www.bmm-electronics.com/Produ...ductgroep_ID=7

They are Nederlands based but will ship worldwide and supply either bare PCB's and schematics, kits or fully assembled and tested PCB's. So depending on your confidence/skill with electronics you can tailor the whole effort to match.

Personally I've gone for the fully assembled and tested to avoid any headache's. I really want to concentrate all my efforts into what I'm good at and thats active XO's and cabinet making. So not having to worry and expend time reaseaching amplifier design and perfecting my soldering skills, not to mention a considerable outlay in time to build all 14 PCB's for a speaker means I can dedicate myself to the parts I'm intimate with.

Some of the amp builders and designers will be shaking their heads but I wanted a tailored solution without the hassle.

What made you decide to go with an analog active crossover as opposed to a digital one?

Again I'm very familar with analogue active XO's. Because of this I can quickly add extra PCB's such as transform, delay, EQ, notch. shelving etc. and know instantly how to use them to correct problems in the measured response during crossover testing/tweaking.

One solution to your bass driver problem would be mounting the driver on the side of the enclosure instead of the front. If you're only using the driver below about 120Hz or so then it isn't important for the driver to face the listening position (anyone feel free to step in an correct me on this if I'm wrong). It's my understanding that these frequencies are below the point where sound becomes "unlocalizable".

Unfortunately the XO point between the bass and mid will be set no lower than 300hz with 400hz more probable. The reason for this is my choice of mid driver, the ATC 3" dome. Its simply not comfortable playing down to something like 120hz.
The side firing bass driver would of course ruin the lower mid's/upperbass and the response curve would suffer as a result.

So the bass drivers have to be used on the front baffle because of the 400hz XO point. Also note that I've chosen to use 2 bass drivers because of the high sensitivity of the mid and treble units - 94dB.

Also I wish to get more accurate, yet extended bass by using two smaller bass drivers such as the 8"ers in a TL loaded sub cabinet.

I'm very excited about seeing this project progress. Your pictures are really excellent. Thanks so much for taking the time to document your effort. [/B]

Its a pleasure and I thank all who share their idea's on here and give inspiration to others.

[ShinOBIWAN]

Originally posted by tiroth
How did you decide on 4th order network? Was it because of your time-alignment requirement?

Choice of order was entirely decided upon because of the time and phase coherency requirement.

The LR 4th order affords both of these in the electrical domain as well as flat amplitude through the pass band and the acoustic sum of the drivers is unity at the XO point.
If I'd gone with butterworth etc I'd would have had an extra notch filter to correct the 6dB summed gain at the XO point.

Of course there will be phase differences since the natural acoustic rolloff of the drivers will cause this. An all pass filter will correct these phase discrepencies.
The Scan Ring will definitely require an all pass but the mid and bass will have very close to what is a 4th order acoustic rolloff at their XO point because the drivers have a flat and large overlapped response.

I also have a preference for 4th order when it comes to active XO's. But I wouldn't sacrifice the sound of a design just to keep within this preference. So of course this choice will be evaluated when testing and tweaking the crossovers and may change based upon the results. For now though these are going to be the starting points.