I've been researching/shopping around for a good 3 way crossover for a DIY speaker build for nearly a week now. Every time I think I find what I need, it's off by a single factor, so I'm continually running in circles and falling down rabbit holes to no end.
What is needed:
-3 way passive crossover, must be 4 ohms for each driver
-At least 100w
-LP at around 300 to 400hz
-Bandpass at around 300 to 400hz / 3,000 to 4,000hz
-HP at around 3,000 to 4,000hz
-All frequencies line up so there aren't noticeable dips/peaks across the frequency output
-High quality parts wouldn't hurt
I've tried searching for complete boards with all components included and also have tried searching for separate 1 way HP, BP, and LP boards without any solid finds.
Also don't want to break the bank, keeping a budget on crossovers at around $80 each (building two speakers, so around $160 total for either complete boards or separate). Active crossovers are pretty much irrelevant unless there are ones that don't require an amp for every input, or stage-style of plug-ins (home audio compatible), or are expensive.
HELP
What is needed:
-3 way passive crossover, must be 4 ohms for each driver
-At least 100w
-LP at around 300 to 400hz
-Bandpass at around 300 to 400hz / 3,000 to 4,000hz
-HP at around 3,000 to 4,000hz
-All frequencies line up so there aren't noticeable dips/peaks across the frequency output
-High quality parts wouldn't hurt
I've tried searching for complete boards with all components included and also have tried searching for separate 1 way HP, BP, and LP boards without any solid finds.
Also don't want to break the bank, keeping a budget on crossovers at around $80 each (building two speakers, so around $160 total for either complete boards or separate). Active crossovers are pretty much irrelevant unless there are ones that don't require an amp for every input, or stage-style of plug-ins (home audio compatible), or are expensive.
HELP
Sorry to bear bad news... But passive crossovers need to be calculated for, at bare minimum, the actual impedance of the specific drivers you are using.
To purchase a pre-made crossover is a major compromise. Yes, it *will* work, and don't let anyone tell you it won't. But a great pair of speakers won't be made with a pre-made filter.
With that in mind, there is really no need to consider such things as dips/peaks, or details.
The driver impedance, response, spacing, baffle dimensions, etc will have far greater effects on the outcome and prevent any stock crossover from yielding predictable results. Likewise, value that would normally be gained with high quality parts is unlikely to show itself.
If you want to take that route to get started, I totally understand. You can always make modifications in the future. Just buy the closest thing to what you want, and don't spend extra for fancy stuff. You can also calculate your own component values based on data sheet information and put together your own filters with a pre-made PCB. That would be a step above pre-made.
Check out this thread on passive crossovers for a good start and further explanation.
Further questions will generate more response if asked in the Loudspeaker / Multi-Way section.
To purchase a pre-made crossover is a major compromise. Yes, it *will* work, and don't let anyone tell you it won't. But a great pair of speakers won't be made with a pre-made filter.
With that in mind, there is really no need to consider such things as dips/peaks, or details.
The driver impedance, response, spacing, baffle dimensions, etc will have far greater effects on the outcome and prevent any stock crossover from yielding predictable results. Likewise, value that would normally be gained with high quality parts is unlikely to show itself.
If you want to take that route to get started, I totally understand. You can always make modifications in the future. Just buy the closest thing to what you want, and don't spend extra for fancy stuff. You can also calculate your own component values based on data sheet information and put together your own filters with a pre-made PCB. That would be a step above pre-made.
Check out this thread on passive crossovers for a good start and further explanation.
Further questions will generate more response if asked in the Loudspeaker / Multi-Way section.
Not sure if this helps at all. I just need a crossover with the given stats.
Getting fairly close with this but not sure about the quality: 1pcs Speaker Frequency Divider Crossover Filter 3-way frequency 4ohm 120W-150W | eBay
Getting fairly close with this but not sure about the quality: 1pcs Speaker Frequency Divider Crossover Filter 3-way frequency 4ohm 120W-150W | eBay
Often what seem to be completely normal things must be further quantified to be described in a useful fashion when physics is involved.
One significant problem is the 4-ohm parameter you defined.
Your drivers are not really 4 ohms. In truth, each driver most likely varies between about 3.5 ohms and 25 ohms, depending on frequency.
So, what do you do about that? The simplified version: Either calculate the crossover component values based on the actual impedance around the desired crossover frequency, or add impedance compensation to your circuit to counteract the driver's changing impedance.
The stock crossover isn't doing either of these things, it's reactances are calculated to work for a straight 4 ohm impedance and your drivers will not be presenting that load to the crossover. As a result, performance will not be what you're expecting, and no amount of component quality will help that.
Post your full driver complement in the multi-way forum, and ask for help in figuring the crossover. Have some patience, and you will probably get much more help than you thought you need.
If you want to stick with a pre-made crossover, don't sweat over details trying to find what you want. It really doesn't exist in a pre-made form. I would buy based on cost and largest inductor wire gauge.
One significant problem is the 4-ohm parameter you defined.
Your drivers are not really 4 ohms. In truth, each driver most likely varies between about 3.5 ohms and 25 ohms, depending on frequency.
So, what do you do about that? The simplified version: Either calculate the crossover component values based on the actual impedance around the desired crossover frequency, or add impedance compensation to your circuit to counteract the driver's changing impedance.
The stock crossover isn't doing either of these things, it's reactances are calculated to work for a straight 4 ohm impedance and your drivers will not be presenting that load to the crossover. As a result, performance will not be what you're expecting, and no amount of component quality will help that.
Post your full driver complement in the multi-way forum, and ask for help in figuring the crossover. Have some patience, and you will probably get much more help than you thought you need.
If you want to stick with a pre-made crossover, don't sweat over details trying to find what you want. It really doesn't exist in a pre-made form. I would buy based on cost and largest inductor wire gauge.
Thanks, I checked out the impedance graph on the woofers and they range from 3.2hz to 17khz depending. I just bought individual crossovers (LP, BP, HP) for each driver. Hoping it will sound ok. Not going for anything crazy.
Thanks for posting your decision.
I think separate sections are a better choice than a single board. And if you ever want to fiddle with them, it will be easier to do.
I hope it works well for you!
I think separate sections are a better choice than a single board. And if you ever want to fiddle with them, it will be easier to do.
I hope it works well for you!
All frequencies line up so there aren't noticeable dips/peaks across the frequency output
that can take a lot of time to develop , you need to know what kind of response is possible, phase shifts, know your XO boundaries, know the problems of the drivers, the time alignment, the impedance curve and response of each driver in the room, then have a clever way of blending both drivers off-axis responses...
You need a calibration microphone, around a hundred available capacitor and inductors + switch boards to build up to 3 order with Z filter etc... then maybe you will find a way to build an elegant XO that you like the sound of.
Instead of this you can go active with dsp dac, etc,
that can take a lot of time to develop , you need to know what kind of response is possible, phase shifts, know your XO boundaries, know the problems of the drivers, the time alignment, the impedance curve and response of each driver in the room, then have a clever way of blending both drivers off-axis responses...
You need a calibration microphone, around a hundred available capacitor and inductors + switch boards to build up to 3 order with Z filter etc... then maybe you will find a way to build an elegant XO that you like the sound of.
Instead of this you can go active with dsp dac, etc,
Jordan,
Pick up a copy of "The Loudspeaker Design Cookbook". Otherwise you're just flailing around and likely to get frustrated. We want you to get excited about building loudspeakers.
Pick up a copy of "The Loudspeaker Design Cookbook". Otherwise you're just flailing around and likely to get frustrated. We want you to get excited about building loudspeakers.
Some OT posts have been removed from the thread.I've been researching/shopping around for a good 3 way crossover for a DIY speaker build for nearly a week now. Every time I think I find what I need, it's off by a single factor, so I'm continually running in circles and falling down rabbit holes to no end.
What is needed:
-3 way passive crossover, must be 4 ohms for each driver
-At least 100w
-LP at around 300 to 400hz
-Bandpass at around 300 to 400hz / 3,000 to 4,000hz
-HP at around 3,000 to 4,000hz
-All frequencies line up so there aren't noticeable dips/peaks across the frequency output
-High quality parts wouldn't hurt
I've tried searching for complete boards with all components included and also have tried searching for separate 1 way HP, BP, and LP boards without any solid finds.
Also don't want to break the bank, keeping a budget on crossovers at around $80 each (building two speakers, so around $160 total for either complete boards or separate). Active crossovers are pretty much irrelevant unless there are ones that don't require an amp for every input, or stage-style of plug-ins (home audio compatible), or are expensive.
HELP
Hi there,
You can use on-line calculators to come up with component values and make your own, however there are 2 big problems and one lesser problem
1. A 4 ohm driver is only 4 ohms for a small part of the frequency spectrum. This is the "nominal" impedance - one used for mainly advertising purposes.
I recommend you go to any retail site and look at the impedance curve for a woofer to see what I mean.
Depending on where you want to crossover and the specific driver in question, the 4 ohms could be as low as 3 or high as 5 or near resonance (30+ ohms), meaning your pre-made / pre-calculated "4 ohm crossover" won't actually cross over at the intended frequency.
2. Drivers do not have a smooth frequency response. Due to the drivers own non-linearities, combined with the enclosure and baffle shape / placement - this makes the driver behave in a unique way. This must be accounted for in the crossover design.
3. The combination of a pre-made crossover and arbitrary drivers could result in a very (dangerously) low impedance. This may not show in a static multi-meter DCR test. You could end up shorting / blowing your amps output transistors.
Its rare (especially with a 3 way) someone lucks out and gets a great sounding speaker with a pre-made crossover that isn't designed for the drivers and enclosure specifically.
Take a look at the "Designing speakers without measurements sticky" - it will save you some potential frustration when you put all the effort into a pre-made crossover only to find it sounds bad.
- Status
- Not open for further replies.