Anyone using an FIR digital crossover?

Havoc

Member
2002-02-06 9:16 pm
yep. i agree. The hardware to do this can be expensive. I am in fact working on such a project right now. Acutally, it will be a general purpose audio DSP box, but digital crossovers and room acoustic corrections are the two most obvious and important applications.

Like Havoc says, the quality will only be as good as the programming, and let's not forget that real FIR filters are VERY computationally intensive, requiring powerful processors to do the math. Manufacturers might be tempted to take shortcuts by using other algorithms (eg IIR filters or DCT/IDCT transforms), potentially degrading the ideal quality that can potentially acheived.

With digital you are making very few trade-offs from analog. If you're starting from digital source material, then the only real sacrifice is perhaps cost, and maybe a short processing latency for very large FIR filters. What you're gaining is a peice of equipment which can potentially deliver carefully controlled and custom-tailored frequency and phase responses. You can do things which are simply not possible in the analog world. Furthermore, digital crossovers are not subject to component tolerances and ageing, nor any unwanted component interactions or non-idealities. Oh, and perhaps one of the nicest features - reprogrammability! Just change the code, and presto! ...it does something completely different! In the digital world, it's all just math, and nothing more. Your only real limitations are the bit-width of your processing, the amount of computational power you have available, and the quality of your algorithms.
 

dc

Member
2001-12-26 9:04 pm
NYC
Chad-

I've been lurking around a variety of sites looking for a digital crossover with room correction. I'd like to use it with phoenix speakers (www.linkwitzlab.com) that i'd like to build for HT. So, how is your project coming along? Would you be willing to share what you've learned? Is it something that a novice might be able to replicate? Can you give me an idea of the total cost? Will it be able to handle 9 channels across the front plus an LFE and 2 rear channels? Will there be a delay in the audio path that is long enough to require video delay as well?

Thanks.

Brad
 
Brad:

I've been eyeballing the Phoenix loudspeakers myself... my only speaker projects to date have been very limited experiments with ESL panels (not counting all the subwoofer enclosures I built for friends in high school... but that was years ago). I find the Linkwitz designs really appealing, and the analogue crossover could be very easily implemented in a DSP... So, i think the Phoenix is a perfect project for me to start with. Now I just have to figure out where I'm going to build these things... hmm

But, I digress... my DSP project is still moving slowly ahead, though the actual hardware has not yet materialized. At the moment, i have a bunch of other stuff to get out of the way so that i will have the time to sit down and finish off the PCB layout and have it fabbed. The cost is going to be pretty steep, at least for the prototype units, and there are lots of small SMD components (including the 208-pin QFP processor... in a 0.5mm pitch package!). I anticipate about $1500-$2000 per finished unit (for my first run of 2). With all the tiny parts, only a proficient hobbyist is likely to assemble one successfully. As far as I/O, the ADSP-21065L processor I'm using is limited to 8 channels in and 8 channels out... a bit limited for home-theatre and multi-channel applications if you intend to use it for digital crossovers as well...

Rest assured, I'll be posting more as my design nears completion over the coming months.
 
FIR crossover for TI TAS5015 digital amp

Having just completed a digital amps basded upon the TAS5015 chip, I started looking forward to a digital crossover using FIR filter technology and hoped reviving this old tread would generate some new interest.


Has anyone built a DIY DSP with enough MIPs for a 2 way FIR filter crossover with complex boost and room compensation?
 
BruteFIR

This is probably what you want. I guess it's not a standalone box but you could build a small dedicated box that resembles a hifi component.

Give the development time spent on this you are not likely to ever catch up if you choose to do it all yourself.

I have had similar thoughts myself and I think this might be the best solution both from a cost and a time-to-market (well, market of one ;) ) POV.

BruteFIR

Oh, and how does your digital amp sound? Do you have any pictures or project descriptions? Great accomplishment.

Best,
Magnus
 
Re: BruteFIR

LineSource said:
Having just completed a digital amps basded upon the TAS5015 chip, I started looking forward to a digital crossover using FIR filter technology and hoped reviving this old tread would generate some new interest.


magnus said:
Oh, and how does your digital amp sound? Do you have any pictures or project descriptions? Great accomplishment.

Agreed.
How does it sound?
Any chance of seeing some schematics for it?
I would also like to build one (well maybe quite a few), but am a little scared of them sounding bad after putting all the design/building effort in.