A Modular 2/3/4/n ways Digitally Controlled State Variable Active Crossover

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Dondo
this is realy nice project. Linkwitz also use OPAMP's for his famous active loudspeakers.
My opinion is that OPAMP
crossover is better than any DSP where double convesion A/D-D/A is made. At the moment I use Behringer DCX2496 and I wanted to replace it with P125. But if you finish your project until end of this year maybe I can wait.
 
Hi Tesla Audio,
the 4 stereo volumes, used in my 4-way system, are relay-based R-2R attenuators.
every board (VBoard) has 2 stereo passive attenuators or opamp based active volumes that you can control with 4-bit (16 steps), usually set to 0.5db steps 0db to -7.5db.
The same board can be used as just 1 stereo volume with 8-bit (256 steps) control.


The technology used is a Logarithmic resistor ladder like this:


Circuit_diagram_of_logarithmic_ladder_DA_converter%2C_Mar_2012.png
 
It's possible to add till 8 boards with few modifications to the actual software.
About clicks and pops with volume boards. Actually when you change the volume of just 0.5db for step it's pretty hard to hear a click.
But probably if you measure the signal it's still there. Statistically, listening music, larger the steps louder the click. I have also my preamplifier main volume with the same technology and it happened to me in past (because a software bug) to get a very loud pop jumping from a high volume to a low volume with a wider step like 20db.
Anyway, when you don't have a bugged software the clicks are just so small that is not a big deal.
Physical clicks of the relays in my aluminum chassis are louder then that. It's the same effect you get out of a shortening rotary switch, but controlled from "a robot" that takes exactly 30ms to switch a step instead of your less precise hand on a knob.
Famous and expensive audio rotary passive attenuators like DACT (LINK HERE) offers solutions with the same issues but people love them anyway!
Once I got used to the physical clicks of microrelays, when I start using a regular volume again I kind of miss them. :D
 
Hi Dondo
I like your project. It is very complex. I'm definitely interested if price will not be very high. You've mentioned that everything could be controled by tablet. That will be perfect and best option. Usually once when you set crossover frequencies you don't touch it again but other things like volume control frequently are used.
 
I agree. Volumes are very useful also during a normal listening session.
Also frequency cut can be slightly adjusted depending from the listening position, listening volume, even music genre.
For example some tweeter sounds great at their lower frequencies just when the volume is low. So the sound system could sound better at higher volumes with a higher cut.
You listen the same speakers at 1 meter or you listen them at 3 meters.
Having a tweeter at 1 meter shooting almost in axis into your ears for several hour could not be your best listening experience. You may want to keep the highs lower. Also during the listening of a particular recording, it can supply the lack of a good analog equalizer (that anyway is a complex, expensive machine usually impacting on your sound quality more than a crossover just for its complexity.)

I think flexibility of the crossover in a multi-way system is a nice feature to have.
 
Even better crossover would be in digital domain. For example today most source music files are stored in personal computer. Before otput on PC USB music content can be filtered for LOW,MID and HIGH frequency. Separate I2S signal should go out from USB. But in that case we need 3 DAC's (3way) or 4 (4 way). Such crossover will avoid all problems with noise and distortion produced in crossover OPAMP's.
Another option is to use 3 or 4 DAC's in parallel with crossover filter loaded in DAC. DAC like Soekris DAM 1021. But so far nobody realised such crossover.
This is just my thinking about ideal crossover.
 
Hi everybody,
today I published the new FBoard revision A Version 1.0 (where A is the layout revision, and 1.0 is the schematics version). I will use this standard of nomenclature from now on, I still have to change the site and the files. but I think it will be useful to better communicate without mistakes the possible future PCBs versions.
Also the software was updated and it's now available in a .zip file.
The software is not still at its final/stable version but few things are now missing.
The audio tests done till now, are till better then what I thought on the beginning. But, as I said, I got a bad batch of op-amps and I'm still waiting for the spare-parts to be delivered, a bit in late already. A full audio test was done in 2-way mode using different speakers and amplifiers recently. The sound was just the way I liked to be! A high quality, low noise, deep bass and bright highs Sound. And what is really impressive in the sound system with the addiction of an Xover44 for who comes from the passive crossovers is the sensitivity. Listening to a well recorded instrument or voice will bring you big emotions, and to obtain the same with a passive crossover you need a very powerful amplifier. No frequency oscillation in cutoff-frequency was audible in any condition with my equipment. In past frequency oscillation was an issue for who designed Hi-Q SVF applied to audio synthesizers, especially using old op-amps. This is also why SVFs are not commonly used is frequency sensitive applications.
Also when you don't listen to me guys, i'm always working in the background or studying the next step. I don't have full time to dedicate to this but slowly but surely the project will go on. Just be patient and save money! :D It's not a cheap project unfortunately but I'll work on a cheaper not 4-way component selection too, to beat the costs using cheaper suppliers when possible.
 
cheaper 2 and 3 way configurations

Hi guys,
I'm working on a new part list (BOM), aimed to make the project cheaper.
Talking with Tesla audio, I figured out that a project like this, can become a lot cheaper. This job to find the best supplier for components, check all the datasheets in detail and in a "methodic and scientific way" it is a pretty hard and will take time.
Being for some years a "purchasing expert" / "the main buyer" in my own company, I think I'll be able to do a good job to bring the price lower for everybody's pocket.
Unfortunately this job will take some time. Some companies are slow to answer my questions and cooperation with some third parties is still to be confirmed.
I used in my 4 way project "premium" components, and in a 4-way system is essential.
By the way in a 2 or 3 way system those low tolerances are not needed and some components (like caps) can be replaced with till a different technology.
On top of that I'm working to build some amplifiers for a customer, building quality wooden boxes my self (as my customer asked for).
Anyway one of my goal is to develop the software for esp32, that give us the possibility to save some money, not spending them on a VGA monitor which is now kind of obsolete and shouldn't be used for new projects.
Unfortunately cheap TFT monitors didn't give me enough affordability for a long-life project, so I think your tablet or cellphone would be a better choice (also considering the latency you have using it).
So I'm asking to anybody interested to help me with the software development specialized in java and esp32 c++ (I built an example software just to test the possibilities of this technology and it's working). I don't ask this for free neither.
Everybody will be paid for what he will do (once the project is commercial), as any person with an ethic should do in this world.
So if you are a good programmer contact me! I will love to work with you and I need it because my time is limited right now.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.