I'd like to get a rough understanding of how to choose coils for my home and possibly car audio. Was looking at Parts Express, which seems to have a good enough selection to choose from, but the coil descriptions are lacking for someone with minimal knowledge of them.
1. Suppose the big question is how do you derive what freq. they cut?
2. How can you decide what effect, either good or bad they'll have on the woofer? Does coil diameter have a huge impact? Any coil designs more accepted in audiophile-ranges?
3. Are the pre-made cross-overs any good? I'm more picky about highs and mids, so I could always swap in a better film cap, but that's a $30 x-over and bump up another $10-30 for higher-end film from say parts conexxion, etc., and woofer must remain distortion-free either with lows to mid-lows.
4. I've got too many speakers to list all of them, but home cabs are 4-way with paper 15" and ribbon surrounds, high 92 SPL 8ohm. Might swap them out eventually for the Realistic Mach series "clones" from PE as my buddy had the Mach 2 or 3 in HS and those were excellent woofers and better than his dad's Mach 1's, and the replacements are getting universally glowing reviews as being better than any of the originals, so possibly looking at coils that would match well with those, or possibly a 3-way cross-over and I'll retrofit a paper mid.
**Car audio, I have numerous amplifiers; about six 6.5" poly woofers, and debating running some inexpensive paper GRS 12" to improve vocals and ambiance, so coils that could match those sizes w/o introducing distortion (rock music).
5. I seem to remember running coils directly to woofers in my early teens to home audio 12" woofers, after harvesting them from older boxes that were gutted, and my 77' MCS 3233 rocked 4, high SPL 12" woofers loudly and cleanly. Is there any actual issue with doing this? Or does there need to be a resistor or other componentry in series or parallel to the coil?
Thanks.
1. Suppose the big question is how do you derive what freq. they cut?
2. How can you decide what effect, either good or bad they'll have on the woofer? Does coil diameter have a huge impact? Any coil designs more accepted in audiophile-ranges?
3. Are the pre-made cross-overs any good? I'm more picky about highs and mids, so I could always swap in a better film cap, but that's a $30 x-over and bump up another $10-30 for higher-end film from say parts conexxion, etc., and woofer must remain distortion-free either with lows to mid-lows.
4. I've got too many speakers to list all of them, but home cabs are 4-way with paper 15" and ribbon surrounds, high 92 SPL 8ohm. Might swap them out eventually for the Realistic Mach series "clones" from PE as my buddy had the Mach 2 or 3 in HS and those were excellent woofers and better than his dad's Mach 1's, and the replacements are getting universally glowing reviews as being better than any of the originals, so possibly looking at coils that would match well with those, or possibly a 3-way cross-over and I'll retrofit a paper mid.
**Car audio, I have numerous amplifiers; about six 6.5" poly woofers, and debating running some inexpensive paper GRS 12" to improve vocals and ambiance, so coils that could match those sizes w/o introducing distortion (rock music).
5. I seem to remember running coils directly to woofers in my early teens to home audio 12" woofers, after harvesting them from older boxes that were gutted, and my 77' MCS 3233 rocked 4, high SPL 12" woofers loudly and cleanly. Is there any actual issue with doing this? Or does there need to be a resistor or other componentry in series or parallel to the coil?
Thanks.
Maybe I'm misunderstanding their significance? I thought they acted as upper range voice chokes when avoiding full-range signals?
Only time I've ever had large coils in a speaker cross-over, was from late 70's boxes, when they didn't skimp on quality and probably had a modern equivalent of a $70 passive cross-over complete with L pads.
Several of my smaller speakers, and even my 15" woofers have really tiny chokes, that look like what you'd see in an old AM handheld radio (ferrite speaker choke?) running to the woofer. I'd like to upgrade those. Are they only to suppress unwanted noise in the signal path? Or are they effectively cutting high midrange vocals? When running the previous coils with my old Realistic 12" polypropylene woofers (excellent SVC bass speaker back in the day), it muted the higher frequencies, and to my ears sounded better. Was hoping to recreate. Tried reading up on coils, and in all honesty, not looking to do a thesis on them, hence I'm here asking. Thanks.
Only time I've ever had large coils in a speaker cross-over, was from late 70's boxes, when they didn't skimp on quality and probably had a modern equivalent of a $70 passive cross-over complete with L pads.
Several of my smaller speakers, and even my 15" woofers have really tiny chokes, that look like what you'd see in an old AM handheld radio (ferrite speaker choke?) running to the woofer. I'd like to upgrade those. Are they only to suppress unwanted noise in the signal path? Or are they effectively cutting high midrange vocals? When running the previous coils with my old Realistic 12" polypropylene woofers (excellent SVC bass speaker back in the day), it muted the higher frequencies, and to my ears sounded better. Was hoping to recreate. Tried reading up on coils, and in all honesty, not looking to do a thesis on them, hence I'm here asking. Thanks.
A place to start would be the driver nominal impedance divided by (6.3 x frequency).
The usual trade off is the resistance of the inductor. Sometimes this resistance is wanted. When in series with woofers, often less is better up to a point. Air cores are good though they often have larger resistances. The diameter helps in getting the most inductance for the copper used.
You'll no doubt miss the target with these but they're the easiest option.
Finding the best values is far more important than the quality of the components. It can make a difference but keep it in perspective.
Common distortion in woofers is also a lesser priority. Their response needs to be made useable.
There is always series resistance in any real circuit, it comes down to how much. There is not generally a problem with running a coil directly in series with a woofer.
Last edited:
15" woofers are often more efficient, and using less drive signal can allow smaller inductors. These cores allow more inductance for the size compared to full sized air cores, but the cores will saturate and effectively cease to exist when over driven.
You answered a bunch of a questions Allen and I thank you. I'm not over-driving or getting any distortion (higher-end amp), but was looking to improve the highs in the cabs and add a non piezo-style mid-high, and have been thinking about better coils, but wasn't sure where to start with them. Are you more of a fan of air coils over copper foil?
Do air cores get any unwanted "effects" from air actually moving across them? If so, would covering them in something like rubber or a small wooden box help?
Does their placement in relation to woofer magnet matter? I've seen where 2 coils near each other have a detrimental effect if placed incorrectly when audio signal is viewed with software.
Do you know of any good 3 or 4 way cross-over schematics out there for DIY, that incorporate a coil? I have a bunch of freedom with speaker impedances and amp can do 400wpc @16 ohm, if need be.
Thanks again!
Do air cores get any unwanted "effects" from air actually moving across them? If so, would covering them in something like rubber or a small wooden box help?
Does their placement in relation to woofer magnet matter? I've seen where 2 coils near each other have a detrimental effect if placed incorrectly when audio signal is viewed with software.
Do you know of any good 3 or 4 way cross-over schematics out there for DIY, that incorporate a coil? I have a bunch of freedom with speaker impedances and amp can do 400wpc @16 ohm, if need be.
Thanks again!
Hi Johnny,
Wow, where to start?
Listen, I think your best bet is to grab XSim (free) and experiment for a while. To properly design a crossver takes at least a cabinet, and the impedance and FR of the drivers.
XSim let's you learn with ideal drivers (the default) or if you get some sample data files from Parts Express you can do some experimentation. They have them for all the Dayton drivers. It would really help you get a feel for what's going on.
And I wouldn't do any part swapping until you know your crossover has the right design to begin with.
The data files XSim uses can be created by OmniMic + DATS V2. for any drivers you have, but dont' have data for.
In addition to XSim, I like WinISD for the box design. WinISD's data can be created by DATS in case you don't have the specs. Actually the specs are usually wrong so the best idea is to go this route anyway.
Use WinISD to create the ideal cabinet/port for each driver that needs one (typically woofers and mids). Use XSim to simulate the ideal crossovers. You can use pre-generated CircuitBlocks to start with. Learn about 1st and second order high and low pass filters this way. The coil in series wiht a woofer is a 1st order low pass filter. Try one out.
Have fun!
Erik
Last edited:
OK. These sometimes give a lumpy response (not even at each frequency in their band). They can sound harsh. Response can (should) be smoothed although the process can be fiddly. Replacing them with something better behaved is one way out of the problem, but get a correctly specified unit.
Coils generally don't wear out, they can be damaged by heat but this is sometimes obvious by their colour.
The first thing to find out is what resistance you want. With woofers/series, and in general this can mean how much can you tolerate (you can always add if necessary). Huge coils might have higher frequency deficiencies but in crossovers this is not a large concern.
I don't mind wire, foil or whatever. As long as it has the right resistance and inductance. Smaller gauge wire where allowable makes practical sense. Physical stability, ie tight and fixed windings is not a bad thing. Iron cores are a good way to get low resistance and I'm ok with these if they can handle the power, especially good with efficient speakers.
I wouldn't say so, no. Rubber might reduce physical vibration which if it deforms the coil might modulate its value, but a minor concern.
If the two coils are dropping a different part of the signal as they ordinarily do then interaction can give unpredictable results. In the case of an iron cored inductor where the coil is highly dependent on the core and the core is susceptible to interference, a magnet might cause saturation.
The field in and around an inductor is shaped like a doughnut. The angle of interaction greatly affects the amount of interaction.
If you are sure you are happy with the existing layout/design of these speakers, you should choose some drivers that fit the bill then work on the crossover. It really depends on what you are trying to cross.
Thanks Erik and Allen. You have both been very helpful, and I'll look into the freeware for the cross-over design.
What exactly is "mH" with regards to coils? What's the difference between 0.80 mH vs. 2.0 mH? They are already listing wire thickness, so is this an inductance value?
Since wattage is apparently concern with coil wire thickness, will that be a huge sound changer, one way or the other? Would it behoove me to get something with thinner wire gauge? Do they get hot?
Sorry for the ton of questions. I'm literally the only person I know that "likes" speakers and no one is into audio.
What exactly is "mH" with regards to coils? What's the difference between 0.80 mH vs. 2.0 mH? They are already listing wire thickness, so is this an inductance value?
Since wattage is apparently concern with coil wire thickness, will that be a huge sound changer, one way or the other? Would it behoove me to get something with thinner wire gauge? Do they get hot?
Sorry for the ton of questions. I'm literally the only person I know that "likes" speakers and no one is into audio.
Last edited:
mH = milliHenry's. It's the microfarads of coils.
That is, in capacitors you talk microfarads, in coils millihenry's are the measure of the inductance. With an ideal capacitor, or ideal coil, the Farads and Henry's are all you need to determine the impedance of the device at any given frequency, and impedance is complicated. Do you know how Resistance, Voltage and Current are related? If not, you should learn that first. Wire thickness helps you pick the right coil and affects DCR (DC Resistance). That is, you could buy an 0.8mH coil with 16, 14 or 12 guage. As you go from 16 to 12 guage the DCR goes down but cost and power ratings go up. In your crossover simulations, the DCR acts as an additional resistor in series with the inductance. Try it. XSim let's you set the inductance as well as the DCR for each coil. Using a very small gauge coil can add unreasonable amounts of resistance. Depends on the application. It's perfectly reasonable to use 22 gauge coils in a tweeter section, and 12 in the woofer. Used appropriately the resistive losses should be negligible. For instance, you find the ideal 0.82 mH coil, which uses 16 gauge wiring but the DCR is too high for your application. Go back and look for 14 gauge. If that's still to high, try to find another gauge. That's kind of how I use wire gauges when it comes to that.
DCR is roughly analogous to ESR in capacitors, but ESR is not a static value, and changes based on frequency, but is independent of the impedance. Confused yet?
With a cap it's the impedance which you can derive by the Farads and the frequency + the ESR. Yes, they get hot. Fortunately coils are rated by wattage, so you should never use an inductor for a lower wattage than you expect.
Last edited:
An inductance on its own will not get hot. All energy taken in is released, manifesting in a magnetic flux and is not consumed. It is the resistance of a practical coil, which is not a necessary property and is referred to as a parasitic resistance, which can produce heat.
The resistance is small and is usually only a problem with some woofers used at high power. It is reasonable to play music then feel them or smell them.. but anything can be used at lower power.
mH = milliHenry's. It's the microfarads of coils.
Wire thickness helps you pick the right coil and affects DCR (DC Resistance). That is, you could buy an 0.8mH coil with 16, 14 or 12 guage. As you go from 16 to 12 guage the DCR goes down but cost and power ratings go up. In your crossover simulations, the DCR acts as an additional resistor in series with the inductance. Try it. XSim let's you set the inductance as well as the DCR for each coil. Using a very small gauge coil can add unreasonable amounts of resistance. Depends on the application. It's perfectly reasonable to use 22 gauge coils in a tweeter section, and 12 in the woofer. Used appropriately the resistive losses should be negligible. For instance, you find the ideal 0.82 mH coil, which uses 16 gauge wiring but the DCR is too high for your application. Go back and look for 14 gauge. If that's still to high, try to find another gauge. That's kind of how I use wire gauges when it comes to that.
DCR is roughly analogous to ESR in capacitors, but ESR is not a static value, and changes based on frequency, but is independent of the impedance. Confused yet?
Yes, they get hot. Fortunately coils are rated by wattage, so you should never use an inductor for a lower wattage than you expect.
You and Allen are the types that make or break some forums, and ultimately their true usefulness to convey thorough knowledge in a friendly manner. In this case, guys like you definitely "make" this one.
Thank you for using the comparisons to capacitors, as I'm at least semi-familiar with ESR, ripple current, etc. from re-capping electrolytics and upgrading to better quality/material film caps in home audio (and some car audio) amplifiers, netting minor improvements with better sounding audio, so having rough parallels to work from is very appreciated in the early stages of learning something new, basically from scratch.
I'm am curious though, what practical, real-world benefit does adding a coil to a tweeter section provide if, say a 10uf to 1uf cap is already cutting low frequencies?
Also, does going to a 3-way or even 4-way compared to a 2-way cross-over design, introduce any detrimental negatives to the audio signal, as heard from the speakers?
What do you guys typically build? 3-way cross-over designs for traditional tweeter, mid, woofer? Or 4-way with tweeter, mid, mid-bass, lows?
You and Allen are the types that make or break some forums, and ultimately their true usefulness to convey thorough knowledge in a friendly manner. In this case, guys like you definitely "make" this one.
Thank you for using the comparisons to capacitors, as I'm at least semi-familiar with ESR, ripple current, etc. from re-capping electrolytics and upgrading to better quality/material film caps in home audio (and some car audio) amplifiers, netting minor improvements with better sounding audio, so having rough parallels to work from is very appreciated in the early stages of learning something new, basically from scratch.
That's odd, because I thought I was trying to deter people from doing it themselves.
First, read the section on "Crossover Basics" on my blog, here. Don't worry about the mods part yet.
Usually you are adding "poles" or sections to a filter. Each pole increases the slope of the filter and rate of phase change. In XSim, drop down two ideal speakers, and add a first order high pass to S1 and a 2nd order High pass to S2. You'll find them under Circuit Blocks.
Notice the slope of the high pass curve of each. Also notice the 2nd order has a coil to ground. It's key the component going to ground is farther away from the one in series or you'll get horribly low impedances, low enough to smoke most amps.
Convert the circuit block to separate components. Switch the order around and examine the impedance chart.
Return them to their original location, and add a resistor to the coil. Lower it to around 1-2 ohms. Notice the impedance plot again. You can use the right-click menu to short the resistor while watching the charts.
When you are done with this, do this again only with a low-pass filter.
Notice how the patterns repeat, but the type of device is reversed.
For more examples, check out the V-Cap crossover designer page. For each type of filter, there's a triangle labeled "Additional Information." Click to expand and you can see how the circuit layouts are for each type of filter.
4th is much harder, but reduces off-axis effects. I don't think there's much of an audible difference between a 1st and second, but others strongly disagree.
How easy or hard a particular topology is to implement depends on the drivers and their innate slopes. With ideal speakers, 4th order is easy as drinking beer while getting a manicure.
Usually it's somewhere between wrestling a kid into the shower and wrestling a sleeping bear. You really really have to want it. Others may have an easier time than I do. I'm a slave to the frequency and phase plots. I do what they tell me, not the other way around. I'm working on a crossover for the LM-1 on my blog. The current crossover is assymetrical. 3rd order low, 2nd order high.
Regardless of your choice in filter types (active or passive) your final speaker slope is, as I believe Allen has pointed out elsewhere, the box+room+driver+filter.
That is, you can implement a perfect 2nd order slope, but the electro-acoustical slope will not be that because no speaker is as ideal as it is in a blank XSim speaker.
Again, nothing like experience. Grab some data files from PE and Dayton to play with.
Best,
Erik
Last edited:
Just to drive people mad I pulled out the Crossover Basics page to it's own blog post here.
That way I do not pollute the minds of the innocent with my upgrade heresy.
Erik
That way I do not pollute the minds of the innocent with my upgrade heresy.
Erik
Thanks, JD. Traditional thinking is that a driver should be used within the band that it is most comfortable and this has driven audio types to use greater numbers of drivers to cover the spectrum. Old school ways emphasised efficiency, and efficiency trades against bandwidth forcing the need for more drivers. Conventional direct radiating hifi type drivers are reasonably wide band and some of the reasons used for splitting up the spectrum have included things like: trying to avoid disturbing the midrange by preventing that driver from having to do bass duty etc.
Modern thinking takes into account our sensitivity at different frequencies such as where in the spectrum we can better get away with allowing disturbances such as crossovers, or diffraction, as well as what is in control of the sound in a given band eg: the diaphragm, baffle or room.. at least that's how I see it, and it allows me to design a speaker into a given situation.
Efficiency is still a good thing but not essential, and harmonic distortion is not the monster it was once thought to be.
Modern thinking takes into account our sensitivity at different frequencies such as where in the spectrum we can better get away with allowing disturbances such as crossovers, or diffraction, as well as what is in control of the sound in a given band eg: the diaphragm, baffle or room.. at least that's how I see it, and it allows me to design a speaker into a given situation.
Efficiency is still a good thing but not essential, and harmonic distortion is not the monster it was once thought to be.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.