DIY Nearfield DSP Studio Monitor Project - feedback needed!

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

spintonik

Member
Joined
Apr 21, 2015
Messages
5
Location
Oulu, Finland, EU
Hi!

My name is Mikko, nice to meet you guys  :) I run my own company, a small boutique shop, where I design and hand-craft on-wall, high-end speakers. If you're interested, please take a quick look at http://www.sonolux.fi . I have my background in mobile industry, where I worked as audio designer for mobile devices. Fiddling with music gear is my hobby - I have a modest home studio, geared for making electronic music.

As it happens, I need new near field studio monitors. As I am capable of creating loudspeaker designs from scratch, it felt stupid to buy commercial products. So I've been working on a near field studio monitor project, for quite some time. This is plenty of work, and waste of energy if I just make one pair just for myself and then scrap the design.

So I plan to make a website around it publish all documents free of charge, to those brave ones wanting to build a pair for themselves. Although this project will be outside my Sonolux -brand, I've been considering creating a webshop with a construction kit, if people find it interesting.

However, let me emphasize;

Apart from speaker cabinet, his design is using 100% 3rd party stock items, freely available from Madisound or Europe Audio, for instance. You'll need to CAM/build the speaker cabinet, but you do not have to buy anything from me. I have no intention of making a commercial success from this project. However, this project costs me up to 2000€ to make, depending on how crazy I go with the (anechoic chamber) testing. + 2-6 man months of work. So I do not mind getting some of that back.

Also, CAM/machining costs include both CNC router programming and actual part machining costs. I plan to offer cabinet parts, ready for assembly, which means that CNC router programming costs are divided by several users - cheaper for everyone, including myself.... IF you live in same continent with me (EU).


This is how much I have now:

1. Target of the design (audio performance and added value to the builder) has been set
2. All components selected and in stock, apart from cabinet.
3. Cabinet acoustics calculated, basic design ready, (so it looks good on paper, we'll see about that...)
3. 3D CAD model of speaker cabinet ready

I'll try to finish 2D drawings for my CNC Workshop partner during next week, and order first pair.


"Technical specs":

2 x 5,25" Scan-Speak Midwoofers (15W/4424G00 I presume, or 15W/4434G00), configured to 2.5 way.
1 x Seas 27TFFNC/G tweeter
Hypex AS2.100 DSP Plate amp (2 x 100W Amp with a DSP crossover)
Birch Plywood cabinet with Glued laminated timber board on front,

Cabinet volume is 15 liters, target is to get a flat response between 50Hz and 30kHz.
Phase coherent design, right from acoustics.
Large, Extremely heavy and Expensive ( 1000€ or more. Yes. Hypex AS2.100 is about 350€  each. )
Because of DSP, it will introduce a latency of 2-3ms.
(unbalanced) RCA Analog IN, S/PDIF and USB in. Slave RCA OUT for Subwoofer.

About target setting:

Ok, so I am building a Nearfield Studio Monitor, with a DSP plate amp. It has gotten it's inspiration from Eve Audio SC305, but is not a clone, just looks similar. Below, is the target of the design:

1. Very neutral, flat sound with good distinction of low-mid, mid frequencies. Aims to be great with acoustic music, spoken word, mixing. Off the shelve. Will not blow your mind with loads of bass, but aims to be very precise all the way to 50Hz (hence THAT Scan-Speak and 2.5 -way).
2. DSP Crossover. Which allows - if you know what you're doing - you to alter the sound to your liking. After I've set up the basic configuration (files), it's just a PC Software and a USB connection. And a 20 page manual. I am not fan of correcting room acoustic issues from loudspeaker, but Hypex AS2.100 allows it.

---Steak and BBQ Sauce -part ---
Look, if You're busy and can afford Eve Audio SC305, buy it.

I am pretty sure it will sound better than this, off the shelve. If you cannot, buy something cheaper. This is for a Tinkerer, one that has audio engineering background, is avid at audio measuring and does not find EXACTLY right sounding speaker from the market.

For instance, if you want to delay woofer 2ms from tweeter, (for what ever kinky reasons) this system allows it. Or you just need to have a crossover frequency of 2737Hz, Yes, bring it on! Obviously, while commercial products offer frequency response adjustments with 1-3dB steps, this one allows 0,1dB step changes from a frequency area YOU select.

I provide you a platform, which should be good off the shelve, you take it from there.

Nearfield Studio Monitor - Boutique-Style.
---Steak and BBQ Sauce -part -ends ---


RISKS:

Power response - I can measure it with gear I have (RME, ARTA, NTI Class-1 compliant mic, I can rent Anechoic chamber with AP and Bruel&Kjaer), but I cannot check it well enough to say it will be 100% perfect.

Reference and real life examples - I have professional studio/sound engineers queuing for a test drive, but I do not live near Dave Pensado.


Where I need your help:

Most important is your comments about target setting, if this makes sense to do in a first place.
It's a different story to whack a pair for a myself, than putting it out there, with my name on it.
If no-one cares, I'll save a lot of effort by making just a pair to myself.

To separate "Yes" -people from ones seriously interested, I may introduce my website and webshop before work goes seriously labor intensive (or I rent a anechoic lab for a week). In that case, one pre-order is all it needs for me to go onwards. However, let's leave this to future, with White Market coming in play when needed.

Tweeter waveguide - I'd like to debate with someone about that for a while.

If you make it all the way here, thanks for reading! Here's the pic you're waited! :)

521 by Mikko Sundman, on Flickr
 
So I am interested. 

The concept looks good. A nice 2-way design and glad to see you are not thinking about ribbon tweeter.

I as a mixer am not looking for hyped or loud, in a perfect world I am looking for broad response with as flat as a curve as possible.

The crossover frequency is best if it can be kept out of the vocal range IMO. I also like my drivers in polarity usually although I have heard some boxes where the design has been to run the drivers out of polarity and they sound fine.

Phase coherence of the drivers at the acoustic level is a must.

Lastly I question porting and or waveguide as a design element. It usually results in hyped not real results IMO.

 
If they are to be used as mid or near-field monitors I think the front face should be as small as possible and the depth increased to get the required cabinet volume

And if you have DSP and lots of power available, why not make them sealed boxes?

Nick Froome
 
pvision said:
If they are to be used as mid or near-field monitors I think the front face should be as small as possible and the depth increased to get the required cabinet volume

And if you have DSP and lots of power available, why not make them sealed boxes?

Nick Froome

Nick,

Great comments. Let me have a look on both matters, and see if they improve the design.  I'll shake up the acoustic calculation tools tomorrow and get back to you then.

-Mikko
 
Pip said:
So I am interested. 

The concept looks good. A nice 2-way design and glad to see you are not thinking about ribbon tweeter.

I as a mixer am not looking for hyped or loud, in a perfect world I am looking for broad response with as flat as a curve as possible.

The crossover frequency is best if it can be kept out of the vocal range IMO. I also like my drivers in polarity usually although I have heard some boxes where the design has been to run the drivers out of polarity and they sound fine.

Phase coherence of the drivers at the acoustic level is a must.

Lastly I question porting and or waveguide as a design element. It usually results in hyped not real results IMO.

Drivers selected are there because I know them very well, I use them in Sonolux professional products. Especially the Scan-Speak 5,25" midwoofer is fantastic, response is nice and flat up to 10kHz, allowing me to try crossover frequencies between 2-4kHz, which should allow me to find a good one. However, in a two-way it is very hard to get a setup where crossover frequency is not at most sensitive area of human ear - and around vocals.

This design does not have throat loading waveguide, I have just added a inlay at 21 degrees angle to drop tweeter cone to same depth than woofer - allowing them to play better in phase. One could replace it with ordinary tweeter and delay at DSP, however this design will be a combination of passive crossover and DSP, because of 2.5 -way design.
 
I think one reason waveguides are popular at the moment is directivity control. If you're sitting a speaker on a meter bridge there's a large, flat surface below it to reflect the signal up to the engineer's ears - probably a worse case than most domestic setups

I would think the distance between the woofers is going to dictate your crossover lowest frequency but I'm not up on the maths of d'Appolito designs

The more I read about directivity and lobing the more I realise the benefit of the Geithain coaxial approach. It's intriguing that Altec and Tannoy started work on coaxials seventy years ago

Nick Froome
 
pvision said:
I think one reason waveguides are popular at the moment is directivity control. If you're sitting a speaker on a meter bridge there's a large, flat surface below it to reflect the signal up to the engineer's ears - probably a worse case than most domestic setups

I would think the distance between the woofers is going to dictate your crossover lowest frequency but I'm not up on the maths of d'Appolito designs

The more I read about directivity and lobing the more I realise the benefit of the Geithain coaxial approach. It's intriguing that Altec and Tannoy started work on coaxials seventy years ago

Nick Froome

I agree I think the coaxial approach even though it has its own problems is the best. That said all but one of my monitoring systems employs Tannoys coaxial design. So please lets see what you come up with.

 

Latest posts

Back
Top