Given the amount of work that needs to go into making a crossover, the number of prototypes and potential variations, and the need to quickly compare alternate solutions, I was considering building myself a simple board to allow rapid testing of alternative solutions.
Basically a large sheet of aluminium with switches to allow components to be switched in /out.
This would allow rapid changing from 2nd-3rd-4th order configs, L-pad adjustments, impedance matching etc at the flic of a switch.
My intention would be to build it in a binary like manner, with multiples of componenet values, i.e. the tweeter could have two banks of capacitors at 6,4.2,2, etc uF, allowing many other combinations to be arranged.
I've read a lot of threads where people have tried countless combinations, and given that inevitably this hobby is adictive and ends with more and more speakers being built, surely something like this should be ideal?
I've googled it but can't find anything, just wondered if anybody else had thought of this, seen this, or would like to use this?
Maybe we could work together to put a design together?
Basically a large sheet of aluminium with switches to allow components to be switched in /out.
This would allow rapid changing from 2nd-3rd-4th order configs, L-pad adjustments, impedance matching etc at the flic of a switch.
My intention would be to build it in a binary like manner, with multiples of componenet values, i.e. the tweeter could have two banks of capacitors at 6,4.2,2, etc uF, allowing many other combinations to be arranged.
I've read a lot of threads where people have tried countless combinations, and given that inevitably this hobby is adictive and ends with more and more speakers being built, surely something like this should be ideal?
I've googled it but can't find anything, just wondered if anybody else had thought of this, seen this, or would like to use this?
Maybe we could work together to put a design together?
Sounds like a great, but complicated, idea! If you look at the layout in Jeff Bagby's Passive Crossover Designer - you'll see that it would be a big board, even for a 2 way. Add to that a wide range of values and it could become very complex very quickly.
I'll ping Jeff and try to get him in here. He should have some good thoughts and ideas.
I'll ping Jeff and try to get him in here. He should have some good thoughts and ideas.
I've thought about this too, but I really think some of it would have to be modular, Like zobels. I'd probably have a second board for those so it could be in or out.
The only thing that's stopped me from trying this is the ridiculous price of copper right now, I could see the inductor costs soaring into the thousands rather quickly, unless there's a way to use some sort of variable inductor.
It would probably be a lot like the old resistor box I have downstairs. It's got 9 rotary switches on it and around each switch, there's a chain of resistors, where each knob gives 1 step and chains more resistors the higher you turn it. First knob is 0.1, second is 1, third is 10, fourth is 100, and so on.
So, I thought about making a couple of inductance ones and capacitance ones, and depending on how they're arranged you could simulate any slope you wanted, but if you were going to build something like this for multi-way, you'd quickly need lots of boxes, especially if you wanted to try higher order filters.
I can picture an entire 4x8 sheet of plywood that's covered in parts and switches.
The only thing that's stopped me from trying this is the ridiculous price of copper right now, I could see the inductor costs soaring into the thousands rather quickly, unless there's a way to use some sort of variable inductor.
It would probably be a lot like the old resistor box I have downstairs. It's got 9 rotary switches on it and around each switch, there's a chain of resistors, where each knob gives 1 step and chains more resistors the higher you turn it. First knob is 0.1, second is 1, third is 10, fourth is 100, and so on.
So, I thought about making a couple of inductance ones and capacitance ones, and depending on how they're arranged you could simulate any slope you wanted, but if you were going to build something like this for multi-way, you'd quickly need lots of boxes, especially if you wanted to try higher order filters.
I can picture an entire 4x8 sheet of plywood that's covered in parts and switches.
I'm presuming series mounted inductors would have a massively negative effect on phase?
How about if we asked an inductor supplier to wind multiple coils on a single rod, similar to the capacitor bank idea? I'm presuming the wire would need to be the same thickness throughout?
Closely banked these could be easy to switch in/out, I'm not sure how the un-used inductors would interact with the ones we'd be using though, any ideas anybody?
How about if we asked an inductor supplier to wind multiple coils on a single rod, similar to the capacitor bank idea? I'm presuming the wire would need to be the same thickness throughout?
Closely banked these could be easy to switch in/out, I'm not sure how the un-used inductors would interact with the ones we'd be using though, any ideas anybody?
Yeah that's basically what I was thinking, just wasn't sure about phase shift problems.
Trying to work out if it would be better to have a bank of inductors in series or have them wound on the same rod, I might give a few inductor companies a call and see what they have to say - sure they have much more knowledge than I do.
Trying to work out if it would be better to have a bank of inductors in series or have them wound on the same rod, I might give a few inductor companies a call and see what they have to say - sure they have much more knowledge than I do.
I've got a couple of 10 mH coils somewhere, I could try sanding the enamel off one edge and dragging a wire across it. The only problem is, I could really only verify by ear that it was doing what I think it's going to do, I don't have a meter that can measure henries.
I have seen coils wound only partially across the iron core, so I'm guessing that it's going to work..as far as phase, we can always ARTA a driver as the coil is varied and see what's happening.
It certainly won't damage anything to try it, I'll have to put that on the project list for this coming week.
I have seen coils wound only partially across the iron core, so I'm guessing that it's going to work..as far as phase, we can always ARTA a driver as the coil is varied and see what's happening.
It certainly won't damage anything to try it, I'll have to put that on the project list for this coming week.
Anybody ever heard of the Vidsonix Virtual Crossover box? Lots of info here.
Vidsonix Crossover Box
They are sold at discount on ebay by Vidsonix. I got lucky once and got one for a $100. I think they are worth it if you do a lot of speaker designs. The box is good for 2 or 3 way 1st and 2nd order slopes. Dial in the values that make your speaker measure and sound good and then disconnect the box and measure the exact values of each component with an LCR or similar meter.
It's a good empirical tool.
Vidsonix Crossover Box
They are sold at discount on ebay by Vidsonix. I got lucky once and got one for a $100. I think they are worth it if you do a lot of speaker designs. The box is good for 2 or 3 way 1st and 2nd order slopes. Dial in the values that make your speaker measure and sound good and then disconnect the box and measure the exact values of each component with an LCR or similar meter.
It's a good empirical tool.
That's pretty much exactly what I was thinking, though it's a little limited...
It looks like you can use a cheap internal capacitor, or bypass to a an external, is that right?
Certainly it's on the right track, it would be good to know what others think of this - does it need to have the facility for 2nd/3rd order slopes? Pads?
I like the phase switch, that's pretty awesome for A/B comparison.
Any chance you've done a mini review?
It looks like you can use a cheap internal capacitor, or bypass to a an external, is that right?
Certainly it's on the right track, it would be good to know what others think of this - does it need to have the facility for 2nd/3rd order slopes? Pads?
I like the phase switch, that's pretty awesome for A/B comparison.
Any chance you've done a mini review?
A full review was done and published in the Nov. 2002 issue of audioXpress. If you can get a hold of that issue, then you'll have it. I have a hard copy. It was once available for .pdf download via the link at the Vidsonix link I posted, but seems to have been dropped when audioXpress re-designed their website. The reivew was written by Kent English and Nelson Pass.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.