DIY crossover components "starter kit"

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I have been considering buying a loot with crossover-components for diy speaker construction, and thought; hey, I´m probably not the only one who would benefit from a list of practical capasitor, inductor and resistor values.

Like resistors, both caps and coils can be coupled in series and paralell to get different values and the formulas too calculate this is pretty simple. This could reduce the lists length considerably.

A good starting point would mabe be a list for each xo-range, for example:
4ohm: 150-450hz
8ohm: 150-450hz
4ohm: 1500-3500hz
8ohm: 1500-3500hz

I´m not shure which values are most practical and most used.
Mabe some of you can help me make a list?

I think the diy community can benefit from a list like this.
 
Administrator
Joined 2004
Paid Member
I don't know if you can assign values like that as the drivers impedance at the XO point will determine the actual cut off and the natural roll off has to be taken into consideration also so it's not that simple but there are lots of charts out there that give the approximations for value.
I've collected and over-bought over the years so my stash includes:

microfarad caps. Assorted types
1.0
1.5
2.2
3.3
4.0
4.7
6.8
10
22
33
47
100

Coils are assorted mH 22 ga. up to 14 ga.
starting at 0.1 (tiny little things)
right up to 6.7 both air and iron core and some salvaged transformers.

Resistors, ohms
1
2
3
4
5
8
12
16

If you were to buy new today probably around $4-500 worth. There are others with tons more but this works for me when I need to scratch the urge. It's nice not to have to order or go and get them.
 
Ahh, thank you. this was the sort of answer i was looking for :)

The values you list are almost exactly what i anticipated, especially the caps.

I know its not actually easy too assume a range of values like that, but if you start too simulate some xo´s you will see that even with the impedance(curves) taken into consideration there will be certain values that can be used. This is also the reason i chose too list such wide range for the xo points. Natural roll-off characteristics I think can be better solved by choosing "matching" drivers, if not; higher order filtering. This still boils down too a certain range of values.
I think it is possible too make this list. I will purchase a bunch of values quite soon, ill most probably do some experimenting and mabe use this thread too post the results.
More people sharing their experience here will also help too see trends.
Then mabe i can get an end result in the form of a list:)
 
I like Cals list. I find that for 2-ways, coils are usually 2 mH or less. Coils in the .10 mH to .4 mH range are often needed for notch filters. Buy the cheap ones. Big caps can be handy and you don't want to use up 6 or 7 caps to make a 200 uf, so having a few large electrolytic caps helps. They are dirt cheap, and save a lot of frustration. Get even more resistor sizes, and quantity. Especially 6 ohms and under. Add some 20s, and a couple 40 ohm to the list.

Tip... Get a marking pen and mark everything so that it can be read without picking it out of the box.
 
sreten: thank you for the post, but i dont think you should dismiss this idea just yet. You do have a point saying the filter is "acoustic", but this is not really true, although you can take a practical approach and reason this way.
The electrical filter only interact with the driver impedance, not the driver acoustics; the only thing that makes the driver sound different after adding a filter component are reduced signal strength at certain frequencies. Mabe that is what you meant, if so, your approach is just easier? :)
I would like some constructive criticism if I´m wrong.

kbgl: Thank you for the post. This is good, wee are now moving in the right direction here. You mention notch filters, I must admit i did not remember those when i started the thread. I am just wondering, how often do you really use those 200uF caps?:p i dont see any use for those, except for filters in the 200hz+- region. Mabe a few really big ones can be an "advanced" package" :D
 
There are no "standard" values for "standard" situations.

Yes. I suppose it would be better to talk about Most Commonly Used Values.

Crossover slopes are acoustic not electrical.

Very true. A point often missed by newbies. The filters described for crossovers are the final acoustic response, which is the combined response of the electrical circuit and the driver's acoustic response as mounted in the enclosure.

Mabe there is a good digital way too do this?

Yes, there is. SoundEasy (and another software; I don't recall which) has an emulator mode called Digital Filter that enables you to listen to your filter design without building it physically. Needs a multi-channel sound card and multi-channel amplifier (though a stereo setup would be fine for a single 2-way).
 
Tip... Get a marking pen and mark everything so that it can be read without picking it out of the box.

I like to measure every component and mark them.

I am just wondering, how often do you really use those 200uF caps?

If you build something like a bi-amped 4 way (check Some of Cal's stuff links), you will see. The small mid bass out front uses them. That's where electrolytics come in handy. I just can't spend as much on a cap as I do on a driver. That's just wrong. :)
 
kbgl: Thank you for the post. This is good, wee are now moving in the right direction here. You mention notch filters, I must admit i did not remember those when i started the thread. I am just wondering, how often do you really use those 200uF caps?:p i dont see any use for those, except for filters in the 200hz+- region. Mabe a few really big ones can be an "advanced" package" :D

Maybe I got carried away using 200 uF as an example. I don't recall where I've used anything that big, but what I was getting at was having some cheap 33 uF, 50 uF and 100 uF on hand can be handy. A 100 uF is $2, and is much cheaper than using 4 or 5 higher quality caps to make large values. Maybe a half dozen each of 50uF and 33 uF would be a good start. I think I've seen some big caps used on woofer circuits for baffle step compensation. Maybe they were around 200 uF.
 
Last edited:
Just bought myself a whole bunch of components and here comes the list:

Resistors:
2x 1ohm
1x 1,5ohm
1x 1,8ohm
4x 2,2ohm
1x 3,3ohm
2x 4,7ohm
1x 5,6ohm
1x 6,8ohm
4x 10ohm
1x 22ohm
2x 27ohm
2x 47ohm

Capacitors:
1x 0,47uF
1x 1uF
1x 2,2uF
1x 3,3uF
1x 3,9uF
1x 6,8uF
6x 10uF
2x 22uF
1x 33uF
2x 47uF
2x 72uF

Coils:
2x 0,1mH
4x 0,2mH
2x 0,25mH
1x 0,39mH
1x 0,56mH
1x 0,8mH
2x 0,9mH
1x 1,2mH
1x 1,5mH
2x 2mH
2x 2,2mH
2x 2,7mH
1x 4,7mH
2x 12mH

I got this for about 2000NoK(about 300UsD)

With these components you should be able too make just about anything :)

Here is a useful link for calculating values when series and/or parallel combinations are used for more specific values:

Series and parallel circuits - Wikipedia, the free encyclopedia
 
Another program is "Pure Music". It has crossover and filter programing. I am not sure how easy it would be to transfer to components though.

I would like to make an alternative argument. The bottom line is: what does the sound do in your room. This is a complex interaction of the room acoustics, driver and box interaction, driver properties, and equipment. Probably in that order.
I left out the XO and other filtering, because that is what you use/make to correct for issues related to the above.
So I need to tune a XO/filter network to not only accommodate the speaker itself, but also the room it is in. In fact the room has the final say, so if my speaker system has a peak at some point, like 2000hz for example, I can add a notch filter and so on. But then what if there were a node or resonance at that frequency in my room? D'oh! That speaker, with it's carefully crafted XO is nice and flat in the design room, but not my listening room. I think that is why mass produced or even high end bought speakers lack, they are made for the average, not my, listening room.
Anyway, being new at this, I would advocate using passive or active networks to get close, but there is little point in trying to get it all just right, then use room analysis and digital processing (it is so good now) to fix it for your room. This can be done for a few hundred dollars worth of equipment (or lots more if you prefer).
This works for me because I use a digital front end. Also because I am not a "purist", I figure that by the time the analog music sound waves hit the recording mike to the time it is finally again an analog signal entering my amp, it has gone through so much processing already, there is no point in worrying about it.

Thus may cause a stir...;-)
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.