Classix II Build with miniDSP

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hello,

I'm cutting my teeth by building a set of Paul Carmody's Classix II. For the price of analogue crossover's I can buy a miniDSP which can be repurposed when I upgrade down the road. Solen here in Canada is also out of stock of some crossover components so I'd either be waiting or shipping from multiple sites.

Does anyone know how I implement the Classix II crossover digitally?

Thanks
Brad
 
Last edited:
Do you have a measurement microphone? If so, you can design a crossover with a better amplitude response than the passive implementation. Adding a notch filter or two is cheap in a DSP, while it often is omitted in passive crossovers due to the component costs. A fourth order Linkwitz-Riley crossover also is no problem in a DSP, which again is expensive to implement as a passive crossover.
 
Thanks - Let me add a few things.

I bought a receiver this weekend from a gentleman who has a UMIK-1. He's the one who pointed me towards miniDSP. He said I'm free to use it, since he hasn't and actually sent me home with it. Said he'd call if he needs it back. A really nice guy.

Allen, I am unfortunately not well learned on electronics.. at all. I'm a comp sci student but there's no electronics discussed so far. Trying to learn on YouTube but will need help on this build.

I really appreciate any help to implement this. I'll do my part and learn as I go.

Thanks
Brad
 
Classix II - undefinition

writ up of the crossover says a 4th order for the tweeter and 2nd order for the woofer crossed over at around 2100-2200 hz.

This is where I would start with the Minidsp, LR4 + LR2 at 2200 and see what you get.

The minidsp can do a number of things that would be impossible/expensive in hardware. So, basically you can set the crossover point and slopes (above) plus get the sensitivities the same (usually reducing the volume of the tweeter, like and lpad in the writeup), then you can delay the tweeter to align the phase (a few ms, you can calculate using geometry and the speed of sound or use an online calculator for delay), then you can start to iron out the various lumps and bumps in the frequency trace using notch filters (+ and - gain) initially because of the drivers (say 500hz upwards) and then because of the room (500hz downwards) and finally if you can mess around with the general profile to get the sound you are after e.g. 6db slope, BBC 1-3khz dip, bass boost, treble taming etc.

There are loads of articles and tutorials out there so happy googling but worth starting with the minidsp site as they have a number of good papers on there.
 
But if the passive crossover was desirably done to begin with?


I would want a UMIK-1 with the minidsp so that I can take full advantage of the room EQ options in the dsp which are obviously not implemented in the analogue crossover. Also reverse engineering the crossover is OK but looking at the final frequency plot on the write up page there looks like there could be quite a bit of improvement available as it is no where near smooth or flat (if that is the desired outcome).
 
Allen, a couple things turning me off from the passive atm. Solen doesn't have everything I need so I'll either have to wait or order from more than one vendor. I'm on a short break between courses so want to get this done before they start again. Also I've never built a passive so will have to learn the ins and outs of that, I'd rather put the time into learning digital design right now. And I don't plan on sticking with these speakers long term. These will get me through school and once I'm working I'll probably want something new. The miniDSP can likely be repurposed.

What do you think, does this sound reasonable?
 
Last edited:
diyAudio Moderator
Joined 2008
Paid Member
Brad, yes it sounds reasonable. I have looked at Pauls original page. Considering the data that is available, I can see a couple of ways to get the information you need.

The first option would be to use a simulator. Use the driver impedance data but use a flat response plot. Apply the crossover halves and the responses will show the transfer functions.

Second, build the speakers and tweak the responses until they look like Pauls plots.
 
Last edited:
diyAudio Moderator
Joined 2008
Paid Member
I would want a UMIK-1 with the minidsp so that I can take full advantage of the room EQ options in the dsp which are obviously not implemented in the analogue crossover.
Correcting room modes electrically is a little beyond the scope of this thread, I would imagine.

there looks like there could be quite a bit of improvement available as it is no where near smooth or flat (if that is the desired outcome).
I share you curiosity, but agree that it is right to be unsure whether it is right to do this. Perhaps the design isn't acoustically smooth and flat, and both these needed to be the way they are shown?
 
I won't be looking to correct room modes. I am curious about whether smoothing will actually make these speakers sound better or worse. Don't get me wrong, I subscribe to flat response and controlled directivity. It's just that these speakers are just about the only ones in my budget, even though they achieve neither.

It would be nice if smoothing made them better. I'd like to think I can get a leg up using the miniDSP.
 
I assume that you bought the basic 2x4 miniDSP (not the HD version) and that your receiver has preamp outputs that are separate from the power amp inputs, so that you can insert the mindDSP in between. In that case, you don't need to worry about gain structure, because your only option is the preamp volume control.

With the HD version you could alternatively control the gain digitally in the miniDSP using their optional remote.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.