Weird crossover settings?
Hi Alan!
Great wright-up.
I've been trying to discern why (given some of your criteria for crossovers) these bookshelves would use a crossover frequency as high as 6500hz.
If I take for granted that their 5" driver (as you say is possible) is extremely directional which allows for a higher crossover - it doesn't make sense that they'd then use ostensibly the same exact drivers in their center and (midrange in their) tower but cross them over at 2300.
Any guesses as to why that might occur?
Seen under 'specifications' here:
Fluance-Signature Series Hi-Fi 5.0 Home Theater Speaker System
Thanks!
Sean.
Hi Alan!
Great wright-up.
I've been trying to discern why (given some of your criteria for crossovers) these bookshelves would use a crossover frequency as high as 6500hz.
If I take for granted that their 5" driver (as you say is possible) is extremely directional which allows for a higher crossover - it doesn't make sense that they'd then use ostensibly the same exact drivers in their center and (midrange in their) tower but cross them over at 2300.
Any guesses as to why that might occur?
Seen under 'specifications' here:
Fluance-Signature Series Hi-Fi 5.0 Home Theater Speaker System
Thanks!
Sean.
Anything I say would only be a guess.
On first thoughts these small waveguides would probably behave best above 6,500 and the 5" drivers might match them somewhere around the lower frequency.
I notice the mids have a lip around them but I wouldn't expect much from those. The tweeter waveguides probably wouldn't be doing much around 2k3 either.
The mains should probably have a fairly flat full range response but the surrounds aren't going to be as critical. The centre channel should probably be working best in the midrange.
It is advisable to have a consistent response between the channels which using the same drivers can help to achieve, but it's hard to say why they made the decisions they did here.
On first thoughts these small waveguides would probably behave best above 6,500 and the 5" drivers might match them somewhere around the lower frequency.
I notice the mids have a lip around them but I wouldn't expect much from those. The tweeter waveguides probably wouldn't be doing much around 2k3 either.
The mains should probably have a fairly flat full range response but the surrounds aren't going to be as critical. The centre channel should probably be working best in the midrange.
It is advisable to have a consistent response between the channels which using the same drivers can help to achieve, but it's hard to say why they made the decisions they did here.
Hello Allen.
Excellent tutorial!
I'm designing a 2-way between a tweeter and midrange, and midbass and bass will be XO actively. The tweeter is a Beyma TPL-150H and the midrange will be a horn-loaded cone (Faital M5N12-80).
The TPL exhibits a constant impedance at 4.9 ohm (see attachment). Hence I don't need to use a resistor in paralel, right?
I have the values to run the numbers on the midrange cone. But will the values change when I horn-load it? Meaning I should build the horn first, measure, and then design the crossover?
Excellent tutorial!
I'm designing a 2-way between a tweeter and midrange, and midbass and bass will be XO actively. The tweeter is a Beyma TPL-150H and the midrange will be a horn-loaded cone (Faital M5N12-80).
The TPL exhibits a constant impedance at 4.9 ohm (see attachment). Hence I don't need to use a resistor in paralel, right?
I have the values to run the numbers on the midrange cone. But will the values change when I horn-load it? Meaning I should build the horn first, measure, and then design the crossover?
(Attachment?..I've seen it before so going from memory..) What you say is basically correct. I think there is a very slight peak which can probably be ignored. You'll find as well that a second order filter is relatively insensitive to such peaks when used in the normal way.
Yes. The directivity will change affecting your choice of crossover point and prompting you to decide on the listening angle, the response will change affecting the needed equalisation, and the impedance will change.
Perhaps this has been suggested and it already does exist, but it would be great if the tute was captured in some kind of an online flowchart automated where all someone has to do is say choose the number of drivers, 2 or 3 way, xo order, baffle width, plug in or import driver values and info, with options for impedance, baffle correction etc.
Is there such a flowchart anywhere perhaps created by others.
Is there such a flowchart anywhere perhaps created by others.
Some notes on upgrading or changing components
There are two angles to upgrading an existing crossover assumed to already be close to optimum.. Redesigning the crossover, or simply replacing components. A redesign will involve a number of measurements, and would be worthy of a look at the cabinet design as well, maybe rounding some edges or even something more involved than that, or at least having a look at its construction before proceeding.
Replacing damaged components
Over time, electrolytic capacitors can dry out and their capacitance value can change. Well constructed modern types won't suffer as badly as older designs or poorly made modern ones.
Inductors and resistors can be damaged by heat. Usually this will be obvious but sometimes not. Checking their value with a meter will help as they can take on values that are out of spec. An inductor that has suffered from a shorted turn shouldn't be used. Components can be uprated by using inductors with lower gauge wire and resistors of a higher power dissipation rating.
Component differences
There are ways that a component can behave in a non-ideal way, and components of the same value may have differences. They are not always significant in a crossover. Eg resistors can be responsible for noise, non-linear distortion, and have some inductance in practice, in small amounts. While these effects exist, good electrical design is about ensuring they are out of range in their application either in frequency, or by being at a low enough level. Resistors are often chosen for their robustness in a crossover.
Non-linear distortion at moderate levels, and of a simple characteristic nature has been demonstrated not to be an issue in speakers. It may impart a small tonal variation in a fairly benign way.
Plastic capacitors, in particular polypropylene are usually used due to their positive characteristics. Electrolytics are typically used when larger values are needed, or as an inexpensive alternative. Capacitors can have small amounts of inductance due to their construction and size, resulting in self-resonance at a high frequency. This is not usually an issue with crossover use (at audio frequencies).
Capacitors show some resistance which appears in series with them. This is small and sometimes shows when changing a capacitor in series with a tweeter as the treble level can change. When there is a resistor placed in series with a capacitor they behave as one combined resistance so it can be compensated for by adjusting the resistor value if needed. Capacitor resistance can also affect deep notch filters.
Inductors don't ordinarily need changing but they can have problems so when doing so, they have a series resistance due to the wire. If the inductor is in series with a resistor this is not really a problem and the combined resistance can be adjusted if necessary. The resistance of an inductor in a tweeter circuit is not usually critical. Sometimes the inductor resistance is part of a deep notch filter where it might be critical.
Used in series with woofers, inductors are usually chosen for low resistance. Higher resistance here could cause enough heating to cause damage to the inductor. It will also potentially affect the woofer response and alter the treble balance. Two ways to reduce the resistance with these larger inductances is either to use a core such as iron which doesn't need the same quantity of coil for the same inductance, or to use heavy gauge wire making for a large air cored component. Iron cores can lead to minor non-linear distortion that isn't necessarily a problem. An undersized core for the application can result in saturation at higher levels that reduces the effectiveness of the core and might be a problem.
Self-resonance in inductors can happen in the audio band for values larger than are usually used in a crossover, due to capacitance between the windings appearing in parallel. Inductors made for crossover use are normally fine.
Capacitors and inductors can be subject to microphonics. Their construction can allow mechanical resonance that affects frequency response. This may be caused by current, magnetism and/or sound. It can be reduced by soft mounting, potting the component in bitumen or varnish, shielding or replacing the component, depending on the cause.
It is worth noting that more attention is often paid to the devices in tweeter circuits partly because hearing is more critical at middle/higher frequencies. It might appear that series components have a greater potential for improvement than shunting ones but more important is the amount of affect each has on the circuit as a whole, considering the other components in the circuit. Sometimes drastically different versions of a component can show no apparent difference in a crossover.
Combining components
Using 1.4 times the wire on an inductor doubles the inductance and increases the resistance by 1.4 times. Using two in series instead will give twice the inductance but twice the resistance. Although for a given wire gauge there is a minor resistance benefit in placing inductors in parallel, it is more practical to use the correct single unit.
Using two half values of capacitance in parallel can reduce the parasitic resistance and inductance, but this is mostly used simply to create specific values of capacitance from available components in a crossover in preference to the series option.
Two double value resistors in parallel, or two half values in series, of the same size is one way to double the power handling when only that size is available to use.
There are two angles to upgrading an existing crossover assumed to already be close to optimum.. Redesigning the crossover, or simply replacing components. A redesign will involve a number of measurements, and would be worthy of a look at the cabinet design as well, maybe rounding some edges or even something more involved than that, or at least having a look at its construction before proceeding.
Replacing damaged components
Over time, electrolytic capacitors can dry out and their capacitance value can change. Well constructed modern types won't suffer as badly as older designs or poorly made modern ones.
Inductors and resistors can be damaged by heat. Usually this will be obvious but sometimes not. Checking their value with a meter will help as they can take on values that are out of spec. An inductor that has suffered from a shorted turn shouldn't be used. Components can be uprated by using inductors with lower gauge wire and resistors of a higher power dissipation rating.
Component differences
There are ways that a component can behave in a non-ideal way, and components of the same value may have differences. They are not always significant in a crossover. Eg resistors can be responsible for noise, non-linear distortion, and have some inductance in practice, in small amounts. While these effects exist, good electrical design is about ensuring they are out of range in their application either in frequency, or by being at a low enough level. Resistors are often chosen for their robustness in a crossover.
Non-linear distortion at moderate levels, and of a simple characteristic nature has been demonstrated not to be an issue in speakers. It may impart a small tonal variation in a fairly benign way.
Plastic capacitors, in particular polypropylene are usually used due to their positive characteristics. Electrolytics are typically used when larger values are needed, or as an inexpensive alternative. Capacitors can have small amounts of inductance due to their construction and size, resulting in self-resonance at a high frequency. This is not usually an issue with crossover use (at audio frequencies).
Capacitors show some resistance which appears in series with them. This is small and sometimes shows when changing a capacitor in series with a tweeter as the treble level can change. When there is a resistor placed in series with a capacitor they behave as one combined resistance so it can be compensated for by adjusting the resistor value if needed. Capacitor resistance can also affect deep notch filters.
Inductors don't ordinarily need changing but they can have problems so when doing so, they have a series resistance due to the wire. If the inductor is in series with a resistor this is not really a problem and the combined resistance can be adjusted if necessary. The resistance of an inductor in a tweeter circuit is not usually critical. Sometimes the inductor resistance is part of a deep notch filter where it might be critical.
Used in series with woofers, inductors are usually chosen for low resistance. Higher resistance here could cause enough heating to cause damage to the inductor. It will also potentially affect the woofer response and alter the treble balance. Two ways to reduce the resistance with these larger inductances is either to use a core such as iron which doesn't need the same quantity of coil for the same inductance, or to use heavy gauge wire making for a large air cored component. Iron cores can lead to minor non-linear distortion that isn't necessarily a problem. An undersized core for the application can result in saturation at higher levels that reduces the effectiveness of the core and might be a problem.
Self-resonance in inductors can happen in the audio band for values larger than are usually used in a crossover, due to capacitance between the windings appearing in parallel. Inductors made for crossover use are normally fine.
Capacitors and inductors can be subject to microphonics. Their construction can allow mechanical resonance that affects frequency response. This may be caused by current, magnetism and/or sound. It can be reduced by soft mounting, potting the component in bitumen or varnish, shielding or replacing the component, depending on the cause.
It is worth noting that more attention is often paid to the devices in tweeter circuits partly because hearing is more critical at middle/higher frequencies. It might appear that series components have a greater potential for improvement than shunting ones but more important is the amount of affect each has on the circuit as a whole, considering the other components in the circuit. Sometimes drastically different versions of a component can show no apparent difference in a crossover.
Combining components
Using 1.4 times the wire on an inductor doubles the inductance and increases the resistance by 1.4 times. Using two in series instead will give twice the inductance but twice the resistance. Although for a given wire gauge there is a minor resistance benefit in placing inductors in parallel, it is more practical to use the correct single unit.
Using two half values of capacitance in parallel can reduce the parasitic resistance and inductance, but this is mostly used simply to create specific values of capacitance from available components in a crossover in preference to the series option.
Two double value resistors in parallel, or two half values in series, of the same size is one way to double the power handling when only that size is available to use.
Last edited:
I just stumbled upon this thread and like many I have found it to be of great value, however, working on my first crossover I would appreciate if someone could tell me if I'm on the right track by answering a few questions I have.
I'm building a crossover for a 2-way speaker using the Dayton RS180S-8 woofer and the Dayton DC28F-8 tweeter. From my understanding looking at the frequency response charts the sweet spot for my crossover should be somewhere in the region of 1700-2000Hz. I have selected 1700Hz when working out parts values.
Using AllenB's guid I was able to work out that I need to use an 8ohm Resistor and a 12uF capacitor in order to "Flatten Woofer Impedance". I selected a 20ohm resistor to also "Flatten Tweeter Impedance".
Q1) Now do I solder these capacitors and resistors between the positive terminals on the tweeter, woofer and the main crossover board?
Q2) The woofer crossover requires just one component which is the "Inductor" (I worked out I need a 0.75mh inductor)?
Q3) The tweeter crossover on the other hand requires 3 components, Capacitor(8.2uF 1%), Inductor(1.1mH) and a Resistor(2.5ohm). So therefore there should be 2 inductors and 1ea of Capacitor and Resistor on my crossover?
Q4) Finally, do my values seem about right to achieve a good crossover or am I way of the mark?
Cheers
I'm building a crossover for a 2-way speaker using the Dayton RS180S-8 woofer and the Dayton DC28F-8 tweeter. From my understanding looking at the frequency response charts the sweet spot for my crossover should be somewhere in the region of 1700-2000Hz. I have selected 1700Hz when working out parts values.
Using AllenB's guid I was able to work out that I need to use an 8ohm Resistor and a 12uF capacitor in order to "Flatten Woofer Impedance". I selected a 20ohm resistor to also "Flatten Tweeter Impedance".
Q1) Now do I solder these capacitors and resistors between the positive terminals on the tweeter, woofer and the main crossover board?
Q2) The woofer crossover requires just one component which is the "Inductor" (I worked out I need a 0.75mh inductor)?
Q3) The tweeter crossover on the other hand requires 3 components, Capacitor(8.2uF 1%), Inductor(1.1mH) and a Resistor(2.5ohm). So therefore there should be 2 inductors and 1ea of Capacitor and Resistor on my crossover?
Q4) Finally, do my values seem about right to achieve a good crossover or am I way of the mark?
Cheers
The Q factor of the tweeter filter looks a little low. In other words the response right at the crossover will dip somewhat. Maybe the resistance you used in the calculation was higher than 2.5 ohms. As it is it should be just lower than 2.5
That is also a small value to use. It can work. This method does work better with smaller values and the tweeter you are using has a peak right where you want to work. An alternative is to use an impedance peak filter. It is a little more involved but more accurate.
The components working on the impedance should be close across the driver terminals, but it isn't very critical.
You'll discover once you tweak the levels, tame the peaks and balance the bass with the baffle, whether this is the best frequency to use. Other than that, it should be a good place to start.
That is also a small value to use. It can work. This method does work better with smaller values and the tweeter you are using has a peak right where you want to work. An alternative is to use an impedance peak filter. It is a little more involved but more accurate.
The components working on the impedance should be close across the driver terminals, but it isn't very critical.
You'll discover once you tweak the levels, tame the peaks and balance the bass with the baffle, whether this is the best frequency to use. Other than that, it should be a good place to start.
Hi Allen,
Just to clarify that I understand what you are saying; The 2.5ohm resistor on the tweeter crossover is a little high? or do you mean the parallel resistor used to flatten the tweeter impedance?
Using your formula I worked out the following values; 8ohm tweeter impedance x 10ohm resistor divide by 8+10 which gave me 4.44ohm and 5.71ohm using the 20ohm resistor which as I understand gives me new impedance value of the tweeter?
Just to clarify that I understand what you are saying; The 2.5ohm resistor on the tweeter crossover is a little high? or do you mean the parallel resistor used to flatten the tweeter impedance?
Using your formula I worked out the following values; 8ohm tweeter impedance x 10ohm resistor divide by 8+10 which gave me 4.44ohm and 5.71ohm using the 20ohm resistor which as I understand gives me new impedance value of the tweeter?
I really need to say a massive thanks to Allen for giving an understanding of how crossovers work and a very special thanks to bwaslo' s amazing Xsim software. I have just finished satorique1 (https://www.loudspeakerbuilding.com...Formula-of-small-speakers_8636,en,901195,9871) without their crossover because I was intending to mate them with a pair of 830452 peerless subs. All now done and my old IMF TLS80mkII are redundant. these new speakers are at least an order of magnitude better than the IMF''s,
Totally stunned by them. And thanks to all on this forum I'm sure most have contributed something to my thoughts and plans somehow :0 :0
Totally stunned by them. And thanks to all on this forum I'm sure most have contributed something to my thoughts and plans somehow :0 :0
Hello all. I'm new to this board. I'm hoping I can seel advice and not offend?
My studio monitors are active Adam A7X and fine. Recently bought so pretty empty wallet.
However, for context, my listening pleasure...
I have refoamed a pair of AR-14 1976 era 10" 2 way speakers this summer. They are stunningly good.
But I am respectfully asking for help and advice in choosing new capacitors as the ones in them are precisely 40 years old and electrolytic... BUT not seeming to be leaking.
The AR-14 x-over is 1300hz. In each speaker, the caps are 16uf and 20uf both at 400v I want to upgrade to polypropylene. I gather I may have to add resistors as well?
Electronics, I normally get as far as changing electric guitar 'pups' and valves in my guitar amps, solder speaker cables etc and that's about it. Audio engineer / production yes I use gear creatively, not really electronics build and maintain.
I buy tube amp valves here and this is the most I pay for a type 4x £34 for el34 -
EL34-SVETLANA WINGED C YEAR 1999-2000
As I have to also buy 7x eec83 at £16 to £20 each... but they can be good for 10-20 years if the guitar amp isn't touring. Yeah, I'm over 50...
I get that basic Solens are cheap here in the UK at £6 odd each. People seem to be sniffy about them?
Mundorf MKP 15uf but no 16uf grrrr, is £5.90 and no 20uf only 22uf for £8.26.
Then I see Claritycap’s Premium Polypropylene Audio Capacitors CMR Range 400v... "Taking inspiration from the highly regarded MR range..." (MR 16uf is £79.06 ouch).
CMR range has a 16uf 400v but no 20uf and prices on some retail sites are nuts / mugging buyers.
Frankly, I'm thinking it's easier to ugrade the capsule in a Rode NT1a condensor microphone from a K67 type to a K47 and the capsule is gold and only costs £125 odd. Make the Rode at £130 very close in sound quality of capture to a Neumann U87 at £2,500.
Sorry, I digress.
I read this site review and understand there are many different capacitor sound flavours... but one of many variables...
Humble Homemade Hifi - Cap Test
Has anyone ever updated capacitors in AR-14's? Even as they are these speakers sound more realistic than many £1K B&W and ilk home speaker systems I have heard in my rather well heeled friends' homes.
Any input would be greatly appreciated.
Thank you in anticipation.
Yours faithfully,
Patrick Robins
My studio monitors are active Adam A7X and fine. Recently bought so pretty empty wallet.
However, for context, my listening pleasure...
I have refoamed a pair of AR-14 1976 era 10" 2 way speakers this summer. They are stunningly good.
But I am respectfully asking for help and advice in choosing new capacitors as the ones in them are precisely 40 years old and electrolytic... BUT not seeming to be leaking.
The AR-14 x-over is 1300hz. In each speaker, the caps are 16uf and 20uf both at 400v I want to upgrade to polypropylene. I gather I may have to add resistors as well?
Electronics, I normally get as far as changing electric guitar 'pups' and valves in my guitar amps, solder speaker cables etc and that's about it. Audio engineer / production yes I use gear creatively, not really electronics build and maintain.
I buy tube amp valves here and this is the most I pay for a type 4x £34 for el34 -
EL34-SVETLANA WINGED C YEAR 1999-2000
As I have to also buy 7x eec83 at £16 to £20 each... but they can be good for 10-20 years if the guitar amp isn't touring. Yeah, I'm over 50...
I get that basic Solens are cheap here in the UK at £6 odd each. People seem to be sniffy about them?
Mundorf MKP 15uf but no 16uf grrrr, is £5.90 and no 20uf only 22uf for £8.26.
Then I see Claritycap’s Premium Polypropylene Audio Capacitors CMR Range 400v... "Taking inspiration from the highly regarded MR range..." (MR 16uf is £79.06 ouch).
CMR range has a 16uf 400v but no 20uf and prices on some retail sites are nuts / mugging buyers.
Frankly, I'm thinking it's easier to ugrade the capsule in a Rode NT1a condensor microphone from a K67 type to a K47 and the capsule is gold and only costs £125 odd. Make the Rode at £130 very close in sound quality of capture to a Neumann U87 at £2,500.
Sorry, I digress.
I read this site review and understand there are many different capacitor sound flavours... but one of many variables...
Humble Homemade Hifi - Cap Test
Has anyone ever updated capacitors in AR-14's? Even as they are these speakers sound more realistic than many £1K B&W and ilk home speaker systems I have heard in my rather well heeled friends' homes.
Any input would be greatly appreciated.
Thank you in anticipation.
Yours faithfully,
Patrick Robins
Capacitors can be constructed badly enough to cause bad sound, but most hardly have a significant sound of their own if at all. Still, it's easy to try and that makes it popular. Claritycaps are good, but I've had just as much luck with some electrolytics and even random caps I've found in odd places.
Normally your mention of a polypropylene makes the right sense for a crossover but your budget will decide. If anything you might favour the ones in series with the tweeters.
You should probably find that the capacitance and ESR of new vs old and dry caps to be the main difference.
Normally your mention of a polypropylene makes the right sense for a crossover but your budget will decide. If anything you might favour the ones in series with the tweeters.
You should probably find that the capacitance and ESR of new vs old and dry caps to be the main difference.
I found this newly drawn schematic... just looking for original pdf I have found online recently...
http://www.classicspeakerpages.net/...atic.jpg.3dce13e342393536364b31145544e349.jpg
http://www.classicspeakerpages.net/...atic.jpg.3dce13e342393536364b31145544e349.jpg