Active Crossovers

[Kingrex]

Let me see if I can break this down.
The signal from pre to amp is small and delicate. So minimal devices are audible when inserted into it.
The signal from a power amp to speaker is robust, so more can be put in its path. But its still audible.

 

[DasguteOhr]

Use this here
you just have to specify your crossover frequency + slope and how many 2 or three way speakers.

Xkitz Electronics

Supplying High Quality Electronic Devices at Reasonable Prices to Hobbyists, Makers, and DIY Enthusiasts all over the World Since 2010.

www.xkitz.com

P.S You need psu for that(plug-in power supply)

 

[aLLeARS]

This is an interesting topic. Lots to be said about the differences in implementing line level vs speaker level crossovers. As another member has already mentioned, once you remove the speaker level crossover and go the line level route, you enter the territory of speaker designer. Once you do that the number of different permutations and practically unlimited options can be daunting. It can also be really fun and rewarding if you have the patience. If you intend to go the line level route, make sure that you first have a good understanding of the speaker level crossover that you are replacing, exactly how it works and what every component in it does. This will give you the necessary information needed to begin setting up a line level version. As someone else here noted, besides the typical L.P. and H.P. crossover filters, many speakers have additional networks in the crossover to deal with driver and or cabinet resonances or other anomalies that the speaker designer has taken the time and effort to overcome.

For instance, the 15” dual -concentric Tannoy’s that I’m actively bi-amping have a parallel RC filter in series with the H.F. driver to make up for the fact that they were originally designed when amplifiers had a current source output and produced more power into a higher impedance load. This produced more SPL at high frequencies above 5kHz where the driver impedance is highest (12.5 Ohms at 1kHz to around 20 Ohms at 20kHz IIRC). With most modern amplifiers acting as voltage sources that parallel RC filter is necessary to get the proper H.F. extension out of that driver. There is also a series RCL filter with variable level control in parallel with the H.F. driver that flattens out its 3kHz resonant peak. Both can be done with a line level crossover, but it sounds different that way. I was taught to always correct a problem at its source.

It's not unheard of to spend months dialing in a crossover to be exactly right for a given speaker. Simply changing the levels between drivers by as little as 0.5dB can make or break the sound. If you can get it right, it can be really rewarding.

I’ve been using a Pass Labs XVR-1 which is set using jumper pins on the circuit board. It’s a two-way crossover that can be set from 1 pole to 4 pole at any practical frequency. For an analog crossover it’s great for dialing things in. However there are around 50 jumpers that the signal goes through. Each one amounts to another contact in the signal path. It also uses a bunch of electrolytic blocking capacitors in the signal path.

Now that I have the crossover point and slope dialed in, I have on order a Merchand XM26 tube crossover. The owner Phil Marchand is a great guy and will do custom orders. The one I have coming uses HEXFRED diodes for the B+ bridge rectifier as well as a 12at7 instead of a 12ax7 in the output stage to drive a 10k Ohm load. It will be interesting to see how it compares to the XVR-1.

 

[Rob181]

This is an interesting topic. Lots to be said about the differences in implementing line level vs speaker level crossovers. As another member has already mentioned, once you remove the speaker level crossover and go the line level route, you enter the territory of speaker designer. Once you do that the number of different permutations and practically unlimited options can be daunting. It can also be really fun and rewarding if you have the patience. If you intend to go the line level route, make sure that you first have a good understanding of the speaker level crossover that you are replacing, exactly how it works and what every component in it does. This will give you the necessary information needed to begin setting up a line level version. As someone else here noted, besides the typical L.P. and H.P. crossover filters, many speakers have additional networks in the crossover to deal with driver and or cabinet resonances or other anomalies that the speaker designer has taken the time and effort to overcome.

For instance, the 15” dual -concentric Tannoy’s that I’m actively bi-amping have a parallel RC filter in series with the H.F. driver to make up for the fact that they were originally designed when amplifiers had a current source output and produced more power into a higher impedance load. This produced more SPL at high frequencies above 5kHz where the driver impedance is highest (12.5 Ohms at 1kHz to around 20 Ohms at 20kHz IIRC). With most modern amplifiers acting as voltage sources that parallel RC filter is necessary to get the proper H.F. extension out of that driver. There is also a series RCL filter with variable level control in parallel with the H.F. driver that flattens out its 3kHz resonant peak. Both can be done with a line level crossover, but it sounds different that way. I was taught to always correct a problem at its source.

It's not unheard of to spend months dialing in a crossover to be exactly right for a given speaker. Simply changing the levels between drivers by as little as 0.5dB can make or break the sound. If you can get it right, it can be really rewarding.

I’ve been using a Pass Labs XVR-1 which is set using jumper pins on the circuit board. It’s a two-way crossover that can be set from 1 pole to 4 pole at any practical frequency. For an analog crossover it’s great for dialing things in. However there are around 50 jumpers that the signal goes through. Each one amounts to another contact in the signal path. It also uses a bunch of electrolytic blocking capacitors in the signal path.

Now that I have the crossover point and slope dialed in, I have on order a Merchand XM26 tube crossover. The owner Phil Marchand is a great guy and will do custom orders. The one I have coming uses HEXFRED diodes for the B+ bridge rectifier as well as a 12at7 instead of a 12ax7 in the output stage to drive a 10k Ohm load. It will be interesting to see how it compares to the XVR-1.

Look forward to reading that comparison...

 

[morricab]

This is an interesting topic. Lots to be said about the differences in implementing line level vs speaker level crossovers. As another member has already mentioned, once you remove the speaker level crossover and go the line level route, you enter the territory of speaker designer. Once you do that the number of different permutations and practically unlimited options can be daunting. It can also be really fun and rewarding if you have the patience. If you intend to go the line level route, make sure that you first have a good understanding of the speaker level crossover that you are replacing, exactly how it works and what every component in it does. This will give you the necessary information needed to begin setting up a line level version. As someone else here noted, besides the typical L.P. and H.P. crossover filters, many speakers have additional networks in the crossover to deal with driver and or cabinet resonances or other anomalies that the speaker designer has taken the time and effort to overcome.

For instance, the 15” dual -concentric Tannoy’s that I’m actively bi-amping have a parallel RC filter in series with the H.F. driver to make up for the fact that they were originally designed when amplifiers had a current source output and produced more power into a higher impedance load. This produced more SPL at high frequencies above 5kHz where the driver impedance is highest (12.5 Ohms at 1kHz to around 20 Ohms at 20kHz IIRC). With most modern amplifiers acting as voltage sources that parallel RC filter is necessary to get the proper H.F. extension out of that driver. There is also a series RCL filter with variable level control in parallel with the H.F. driver that flattens out its 3kHz resonant peak. Both can be done with a line level crossover, but it sounds different that way. I was taught to always correct a problem at its source.

It's not unheard of to spend months dialing in a crossover to be exactly right for a given speaker. Simply changing the levels between drivers by as little as 0.5dB can make or break the sound. If you can get it right, it can be really rewarding.

I’ve been using a Pass Labs XVR-1 which is set using jumper pins on the circuit board. It’s a two-way crossover that can be set from 1 pole to 4 pole at any practical frequency. For an analog crossover it’s great for dialing things in. However there are around 50 jumpers that the signal goes through. Each one amounts to another contact in the signal path. It also uses a bunch of electrolytic blocking capacitors in the signal path.

Now that I have the crossover point and slope dialed in, I have on order a Merchand XM26 tube crossover. The owner Phil Marchand is a great guy and will do custom orders. The one I have coming uses HEXFRED diodes for the B+ bridge rectifier as well as a 12at7 instead of a 12ax7 in the output stage to drive a 10k Ohm load. It will be interesting to see how it compares to the XVR-1.

Thx for that useful info! I am really curious about the tube xover because I have been thinking about this for quite some time.

 

[Kingrex]

This is an interesting topic. Lots to be said about the differences in implementing line level vs speaker level crossovers. As another member has already mentioned, once you remove the speaker level crossover and go the line level route, you enter the territory of speaker designer. Once you do that the number of different permutations and practically unlimited options can be daunting. It can also be really fun and rewarding if you have the patience. If you intend to go the line level route, make sure that you first have a good understanding of the speaker level crossover that you are replacing, exactly how it works and what every component in it does. This will give you the necessary information needed to begin setting up a line level version. As someone else here noted, besides the typical L.P. and H.P. crossover filters, many speakers have additional networks in the crossover to deal with driver and or cabinet resonances or other anomalies that the speaker designer has taken the time and effort to overcome.

For instance, the 15” dual -concentric Tannoy’s that I’m actively bi-amping have a parallel RC filter in series with the H.F. driver to make up for the fact that they were originally designed when amplifiers had a current source output and produced more power into a higher impedance load. This produced more SPL at high frequencies above 5kHz where the driver impedance is highest (12.5 Ohms at 1kHz to around 20 Ohms at 20kHz IIRC). With most modern amplifiers acting as voltage sources that parallel RC filter is necessary to get the proper H.F. extension out of that driver. There is also a series RCL filter with variable level control in parallel with the H.F. driver that flattens out its 3kHz resonant peak. Both can be done with a line level crossover, but it sounds different that way. I was taught to always correct a problem at its source.

It's not unheard of to spend months dialing in a crossover to be exactly right for a given speaker. Simply changing the levels between drivers by as little as 0.5dB can make or break the sound. If you can get it right, it can be really rewarding.

I’ve been using a Pass Labs XVR-1 which is set using jumper pins on the circuit board. It’s a two-way crossover that can be set from 1 pole to 4 pole at any practical frequency. For an analog crossover it’s great for dialing things in. However there are around 50 jumpers that the signal goes through. Each one amounts to another contact in the signal path. It also uses a bunch of electrolytic blocking capacitors in the signal path.

Now that I have the crossover point and slope dialed in, I have on order a Merchand XM26 tube crossover. The owner Phil Marchand is a great guy and will do custom orders. The one I have coming uses HEXFRED diodes for the B+ bridge rectifier as well as a 12at7 instead of a 12ax7 in the output stage to drive a 10k Ohm load. It will be interesting to see how it compares to the XVR-1.

We assume speaker designers have things all worked out. But do they.
As you were explaining impedance as frequency rises, I was wondering if that is a reason some speakers may seem to get brighter sounding as you turn up the volume. As volume increases, the voltage changes.

 

[Nuforce]

I recommend the sublime audio analogue crossover previously mentioned. That is what I use and it is incredible. The SNR on it is exceptional.

I use it to run my magnepan 20.1 speakers in active mode. The massive external passive crossovers with nasty caps and inductors are in a box in the cellar where they belong! Running big maggies active is almost a religion experience...

Atma-sphere MA1 150w triode monoblocks for the mid/treble and a pair of inexpensive emotiva 600w monos for the bass section.

Rather than using DSP I use the bass gain on the crossover to dial back the bass slightly and use bass traps.

I do not recommend a digital crossover. I've tried all the usual makes and they ALL ruined the sound, adding glare and hardness to the sound.

I run a completely digital system but use a resistor ladder dac and with the sublime audio crossover the sound is totally natural. With the incredible Rose RS130 as the transport.

 

[Nuforce]

This is an interesting topic. Lots to be said about the differences in implementing line level vs speaker level crossovers. As another member has already mentioned, once you remove the speaker level crossover and go the line level route, you enter the territory of speaker designer. Once you do that the number of different permutations and practically unlimited options can be daunting. It can also be really fun and rewarding if you have the patience. If you intend to go the line level route, make sure that you first have a good understanding of the speaker level crossover that you are replacing, exactly how it works and what every component in it does. This will give you the necessary information needed to begin setting up a line level version. As someone else here noted, besides the typical L.P. and H.P. crossover filters, many speakers have additional networks in the crossover to deal with driver and or cabinet resonances or other anomalies that the speaker designer has taken the time and effort to overcome.

For instance, the 15” dual -concentric Tannoy’s that I’m actively bi-amping have a parallel RC filter in series with the H.F. driver to make up for the fact that they were originally designed when amplifiers had a current source output and produced more power into a higher impedance load. This produced more SPL at high frequencies above 5kHz where the driver impedance is highest (12.5 Ohms at 1kHz to around 20 Ohms at 20kHz IIRC). With most modern amplifiers acting as voltage sources that parallel RC filter is necessary to get the proper H.F. extension out of that driver. There is also a series RCL filter with variable level control in parallel with the H.F. driver that flattens out its 3kHz resonant peak. Both can be done with a line level crossover, but it sounds different that way. I was taught to always correct a problem at its source.

It's not unheard of to spend months dialing in a crossover to be exactly right for a given speaker. Simply changing the levels between drivers by as little as 0.5dB can make or break the sound. If you can get it right, it can be really rewarding.

I’ve been using a Pass Labs XVR-1 which is set using jumper pins on the circuit board. It’s a two-way crossover that can be set from 1 pole to 4 pole at any practical frequency. For an analog crossover it’s great for dialing things in. However there are around 50 jumpers that the signal goes through. Each one amounts to another contact in the signal path. It also uses a bunch of electrolytic blocking capacitors in the signal path.

Now that I have the crossover point and slope dialed in, I have on order a Merchand XM26 tube crossover. The owner Phil Marchand is a great guy and will do custom orders. The one I have coming uses HEXFRED diodes for the B+ bridge rectifier as well as a 12at7 instead of a 12ax7 in the output stage to drive a 10k Ohm load. It will be interesting to see how it compares to the XVR-1.

I understand your logic regarding a tube analogy crossover but my gut tells me that a tube approach would add colour / distortion.

At the very least I'd recommend also trying the latest Sublime Audio analogy crossover as a different comparison. It uses cards for the crossover points not pots or relays and has a very very good SNR.

And please do share your results!

 

[Kingrex]

I recommend the sublime audio analogue crossover previously mentioned. That is what I use and it is incredible. The SNR on it is exceptional.

I use it to run my magnepan 20.1 speakers in active mode. The massive external passive crossovers with nasty caps and inductors are in a box in the cellar where they belong! Running big maggies active is almost a religion experience...

Atma-sphere MA1 150w triode monoblocks for the mid/treble and a pair of inexpensive emotiva 600w monos for the bass section.

Rather than using DSP I use the bass gain on the crossover to dial back the bass slightly and use bass traps.

I do not recommend a digital crossover. I've tried all the usual makes and they ALL ruined the sound, adding glare and hardness to the sound.

I run a completely digital system but use a resistor ladder dac and with the sublime audio crossover the sound is totally natural. With the incredible Rose RS130 as the transport.

They look nice but you have to understand everything about your drivers and enclosure to even begin guessing at what cards to insert. I wish I could send my crossover to someone to have it mapped.

They also have Op Amps. I don't know much but I did watch a OCD Hifif guy where Mike was saying the Op Amps add a little glare. He paid the money to have Merchand make him a crossover where all the Op Amps were replaced with something else.

I know little about all this. I am probably going to get a Ashely Protea soon. JR recommended it as something that works for the swarm. It does not have true phase. It only has a delay. But it works for the purpose and a used one is $600. Its fully active so I can dial to my hearts content and determine what parameters I want. Once I figure it out, I will look for a active analog. I will circle back around to Sublime

 

[Nuforce]

If you don't know the exact crossover then as you say, get a crossover with adjustable slopes and frequencies. Use your ears and REW to dial in the best sound. Then you can jump to a proper crossover using cards. Good plan.

Regarding the comment about glare on the sublime audio crossover, I can safely report that is not the case. I have a good ear, and the 20.1 true ribbon tweeter is highly revealing.

The sound is perfect and extremely natural. I suggest the aversion of op amps is unnecessary.

Using a digital crossover with A/D conversion on the other hard, now that is a path to glare. No thanks!

 

[Robh3606]

it almost sounds like your saying the impact to transparency on the front end is worse than the back end. I thought people said all the chokes, caps, resistors in a passive crossover were more damaging than a device on the signal side.

And how many active and passive parts are you adding into the signal chain going active? I don't but that at all. I run analog bi/multi amp systems where the passive crossover parts are used to do any compensation needed to flatten/tailor the frequency response of the individual drivers. The actual driver crossover is done actively. An example would be a CD horn or waveguide where you can just remove a passive crossover, sub in a simple active and get the same results.

Simple would be a typical 24dB L/R where the only adjustment options you have are frequency and driver level and polarity.

If you have drivers that are smooth and need no response tailoring you can go with a simple analog crossover but many times that is simply not the case.

You need to look at the individual driver response measured in the actual enclosures used of the voltage drive that the passive crossover provides for the drivers. Lot's of times in passive they are not using canned and textbook curves to get the desired response.

Rob

 

[Kingrex]

And how many active and passive parts are you adding into the signal chain going active? I don't but that at all. I run analog bi/multi amp systems where the passive crossover parts are used to do any compensation needed to flatten/tailor the frequency response of the individual drivers. The actual driver crossover is done actively. An example would be a CD horn or waveguide where you can just remove a passive crossover, sub in a simple active and get the same results.

Simple would be a typical 24dB L/R where the only adjustment options you have are frequency and driver level and polarity.

If you have drivers that are smooth and need no response tailoring you can go with a simple analog crossover but many times that is simply not the case.

You need to look at the individual driver response measured in the actual enclosures used of the voltage drive that the passive crossover provides for the drivers. Lot's of times in passive they are not using canned and textbook curves to get the desired response.

Rob

You lost me???
More and more I get why so many people with well tuned systems say they tried DSP and walked away. It sounds like a can of worms. I'm lumping active crossovers into this.

 

[Robh3606]

You lost me???
More and more I get why so many people with well tuned systems say they tried DSP and walked away. It sounds like a can of worms. I'm lumping active crossovers into this.

Well not trying to dissuade you from experimenting. If you have a well tuned system you are happy with leave it be. The old "If it's not broke don't fix it" comes to mind. The idea though is you can always go back to passive or do one channel as a time and use the original as reference.

It can be a can of worms if you want an exact duplicate of a passive system converted to active. It takes the required tools and how to use them along with a really good understanding exactly what the passive crossover is doing. To clarify the voltage drives are passive crossover does to tailor the frequency response to each driver to get the target curve.

Here is an example of two speakers showing their on axis summed response and the voltage drives needed to achieve the end result on axis response. One is a horn loaded system 4800 the other an L250 all direct radiator system using 6bd slope. Looking at the voltage drives it's plainly obvious the L250 would be easier to do a "conversion".

Rob

 

[Kingrex]

Well not trying to dissuade you from experimenting. If you have a well tuned system you are happy with leave it be. The old "If it's not broke don't fix it" comes to mind. The idea though is you can always go back to passive or do one channel as a time and use the original as reference.



Rob

I'm more the, if it aint broke, I can break it.

 

[Robh3606]

I'm more the, if it aint broke, I can break it.

LOL! Yeah I know what you mean sometimes it does go that way. It's happened to me often enough. Tinkering has it's hazards!

Rob

 

[Kingrex]

LOL! Yeah I know what you mean sometimes it does go that way. It's happened to me often enough. Tinkering has it's hazards!

Rob

Its that tinkering bug that drug me down the power rabbit hole. I never knew my power was broken. Not until my Rega Osirus was humming. And even then, everyone wanted to sell me a power conditioner. I actually bought a Isotek Syncro power cable. That actually fixed the hum issue. And took away a bit of edge in the sound. At the time, the guy who sold me the Rega said I should try changing the wire from 12 to 10 at my rack. Even with the hum problem solved, I was piqued at the idea of fiddling with power since it was basically free to me. And so it began.

 

[Bones13]

I think Bryston makes an active crossover device to run the elements in passive speakers via separate amps. They also sell the crossover-less speakers.

There is also a DIY Nelson Pass design for full range driver and low frequency driver combination.

 

[Nuforce]

Marchand XM44 Analog Active Crossover Review

This is a review and detailed measurements of the Marchand XM44-3 3-way Analog Active Crossover. It is on kind loan from a member. This is a modular 3-way system. As configured, it costs US $1,900. So not cheap. Question the owner had was how it compares to a digital crossover which he also...

www.audiosciencereview.com

This put me off Marchand.

 

[morricab]

Marchand XM44 Analog Active Crossover Review

This is a review and detailed measurements of the Marchand XM44-3 3-way Analog Active Crossover. It is on kind loan from a member. This is a modular 3-way system. As configured, it costs US $1,900. So not cheap. Question the owner had was how it compares to a digital crossover which he also...

www.audiosciencereview.com

This put me off Marchand.

It shouldn’t have put you off. Those measurements aren’t nearly as bad as they make out. Most of the distortion is 2nd order and at -75dB, it’s inaudible.

 

[Kingrex]

And questionable measurements. Someone in the comments said his test equipment can create the noise. I don't know if it true or not, but Amir does not have a proper tool basket nor the technical expertise to measure much. He have very crude and limited techniques. He is only giving a small data point.
I am more interested in the comments by users that have the unit, as well as the Mini DSP and the Beringer I think it wass.