Actif analog crossovers

[Theta]

Seems to me there are lots of benefits to using actif analog crossovers over DSP crossovers, if you are using a Dac to convert to analog then using a dsp, are you not converting back to digital? Can't be good. If you are using an actif analog crossover you don't need a 8 chanel Dac. Also with a DSP the volume control is in the software, is there not a chance of frying your speakers? Of course the DSP unit will offer digital equalization, but Dacs like the RME have that capability and you can connect it to an analog crossover. I welcome corrections and opinions on this.

 

[DVDdoug]

Active, like this?

That's fine if you don't need additional EQ. And of course, you can aways add DSP or an analog equalizer.

is there not a chance of frying your speakers?

IMO - Your tweeters (and possibly midrange) should have a capacitor as an additional-protective high-pass filter. The same goes with an analog crossover in a bi-amped/tri-amped setup.

 

[Theta]

Active, like this?

That's fine if you don't need additional EQ. And of course, you can aways add DSP or an analog equalizer.


IMO - Your tweeters (and possibly midrange) should have a capacitor as an additional-protective high-pass filter. The same goes with an analog crossover in a bi-amped/tri-amped setup.

Yes I am using a DBX active crossover, sorry for the mis-spelling, I am French.

 

[Keith_W]

Bonjour, bienvenue to ASR (sorry for my bad French).

You are right, analog active crossovers have some advantages over digital:

- only requires one DAC, as you pointed out.
- have negligible latency and propagation delay. The XO doesn't have to "think" about your crossover, so signal passes through it in an instant.
- analog volume control. It will not glitch and accidentally send max volume to your speakers, as you said.
- simplicity and low cost.

However, digital crossovers have many advantages over analog, FIR filters in particular:

- steep crossover slopes and unusual configurations are possible
- phase rotations can be adjusted independent of amplitude
- are extremely flexible, it is easy to create new crossover configurations in a few minutes. This is great if you love tinkering.
- it is possible to linearize every driver individually, using measurements taken of that driver
- time alignment via digital delays is possible. I use 48kHz sampling rate, so I can align down to 0.0208ms. If I used higher sampling rates, it could be even finer than this. With analog crossovers, you use all pass filters which are nowhere close in precision.

The volume control question is valid with digital. However, you can protect yourself by limiting the maximum output from your amplifier or DAC. I have never seen a digital volume control on a DAC glitch and accidentally send maximum volume to the amp, and this concern is completely absent with analog volume controls on amps. In my own system here, every channel has its own volume trim, implemented three times. My 8 channel DAC has individual volume trim, and so do the amps. Finally I can implement volume trim via the crossover software. The maximum volume produced can be set with these volume trims, so even if the player software I am using for volume control (JRiver) glitches and sends out maximum volume, it will not damage the speakers. Only annoy the neighbours.

IMO, which solution is superior depends on your application.

 

[ppataki]

Do we have measurements on any analog crossovers?
I guess they would have an impact on the signal in terms of SINAD, frequency response, etc.
If so, then that would be yet another disadvantage

Regarding volume control: the only place I control volume is in Jriver (internal 64-bit volume control) using its Volume Protection feature - that way the volume can never be maxed accidentally (for that matter it is not even easy to max it even deliberately)

 

[Sokel]

Do we have measurements on any analog crossovers?
I guess they would have an impact on the signal in terms of SINAD, frequency response, etc.
If so, then that would be yet another disadvantage

Regarding volume control: the only place I control volume is in Jriver (internal 64-bit volume control) using its Volume Protection feature - that way the volume can never be maxed accidentally (for that matter it is not even easy to max it even deliberately)

Here's some of a dirt cheap one.go some posts down where I took some care of it:

Cheap analog electronic crossover

I've been playing lately with a very cheap analog 2-in/4-out analog crossover. Is somehow DIY but very easy to fix. To my surprise it measures very decently! (note that has been measured with the worst possible conditions,long cables soldered just to see,adapters,noisy PSU and by someone totally...

www.audiosciencereview.com

On the second page there's also some dirty measurements of a behri which is not nice.
Amongst the ones I have measured the best was a Marchand XM9 which simply maxed my ancient interface,I maybe make a thread about it just for the fun of it would be better to wait for my Cosmos.

Edit:Also here's some couple of days ago testing a unbal to bal conversion scheme

Unbalanced to Balanced conversion

Hi, I have an unbalanced preamp and a balanced amp. Why is there a converter box like this https://www.amazon.com/ROLLS-MB15b-Promatch-and-More/dp/B0002IL4B4 if you can just use cables like these? https://www.amazon.com/gp/product/B0957Z55HH/ I am trying to understand the...

www.audiosciencereview.com

 

[radix]

Linkwitz used active crossovers in some speakers and minidsp in some others. I think quite a few active speakers use non-DSP active crossovers and it's only the newer designs that have gone DSP (newer being many years).

I used to be hesitant about the D/A to A/D to D/A chain, but eventually got over it. I think it's pretty negligible. There are other threads on ASR about multiple analog-digital conversions.

dbx 223xs Crossover Review

This is a review and detailed measurements of the dbx 223xs active analog balanced crossover. It is on kind loan from a member and costs US $245. I must say this is a nice and sturdy package for such a bargain price. Controls feel nice as well. Functionality though is confusing given the...

www.audiosciencereview.com

Rolls SX45 Review (Analog Crossover)

This is a review and detailed measurements of the Rolls SX45 analog, 2-way crossover. It is on kind loan from a member and costs US $99. As you see, it is a compact box, powered by a small external power supply: I like the clipping indicator in addition to mono low frequency output to drive a...

www.audiosciencereview.com

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

There's probably more. And, of course, any active speaker without a DSP likely uses an active crossover.

 

[ppataki]

Linkwitz used active crossovers in some speakers and minidsp in some others. I think quite a few active speakers use non-DSP active crossovers and it's only the newer designs that have gone DSP (newer being many years).

I used to be hesitant about the D/A to A/D to D/A chain, but eventually got over it. I think it's pretty negligible. There are other threads on ASR about multiple analog-digital conversions.

dbx 223xs Crossover Review

This is a review and detailed measurements of the dbx 223xs active analog balanced crossover. It is on kind loan from a member and costs US $245. I must say this is a nice and sturdy package for such a bargain price. Controls feel nice as well. Functionality though is confusing given the...

www.audiosciencereview.com

Rolls SX45 Review (Analog Crossover)

This is a review and detailed measurements of the Rolls SX45 analog, 2-way crossover. It is on kind loan from a member and costs US $99. As you see, it is a compact box, powered by a small external power supply: I like the clipping indicator in addition to mono low frequency output to drive a...

www.audiosciencereview.com

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

There's probably more. And, of course, any active speaker without a DSP likely uses an active crossover.

If I take a look at the dbx for example: it has a SINAD of 100
So if I connect a DAC with a SINAD of 123 to it, the dbx will decrease that to 100 overall, right?
I know SINAD is not everything but just for the sake of example - it will have an impact on the signal (-23dB of SINAD in this example)
Please correct me if I am wrong

 

[Sokel]

If I take a look at the dbx for example: it has a SINAD of 100
So if I connect a DAC with a SINAD of 123 to it, the dbx will decrease that to 100 overall, right?
I know SINAD is not everything but just for the sake of example - it will have an impact on the signal (-23dB of SINAD in this example)
Please correct me if I am wrong

You're right but every x-over will have impact with it's filters applied,have a look here for example.

DSP Measurements and Rising Noise Floor

In several threads related to DSP platforms I've seen requests for measurements with filters engaged. I thought it unlikely that implementing filters in a DSP would have measurable effect on noise / distortion other than the obvious change in the fundamental, however @Sokel recently mentioned...

www.audiosciencereview.com

Note the above are pure digital measurements,their analog equivalent are much worst of course.
They did applied some fix after this thread but the penalty is still there,only lower.

 

[radix]

If I take a look at the dbx for example: it has a SINAD of 100
So if I connect a DAC with a SINAD of 123 to it, the dbx will decrease that to 100 overall, right?
I know SINAD is not everything but just for the sake of example - it will have an impact on the signal (-23dB of SINAD in this example)
Please correct me if I am wrong

I agree with @Sokel.

But what is the SINAD of your amp and your speakers and your room? A lot of amps are < 100 and many are only 1% THD at full power. Then consider your speaker distortion. I'm not saying it's good to add noise, but sacrificing some noise for active crossover or DRC seems worth it.

I think the devil is in the details of the overall system.

 

[voodooless]

sorry for the mis-spelling, I am French.

It's the same word You guys both stole it from Latin (as did the Dutch for that matter)

if you are using a Dac to convert to analog then using a dsp, are you not converting back to digital?

Then don't if you have a problem with that It's not a general property of a digital crossover.

If you are using an actif analog crossover you don't need a 8 chanel Dac.

But you do need a whole lot of electronics to accomplish that. DACs are about as cheap as a bunch of opamps, so I don't see a real benefit or drawback here.

Also with a DSP the volume control is in the software, is there not a chance of frying your speakers?

Again, this is not a general property. You can still have an analog volume control if you want.

In the end, an analog active crossover can work very well, as long as the limitations fit your use case, especially if you couple that with DSP EQ and/or time alignment.

If you have more specialized use cases, you're out of luck. For instance, when you need a lot of delay between drivers, or if one of the drivers has a severe resonance that needs to be dealt with. Now the latter can also be done passively, so there are ways around some of the limitations for sure. At the end of the day though, DSP is very much about convenience, speed, and ease.

 

[ppataki]

You're right but every x-over will have impact with it's filters applied,have a look here for example.

Even those that are fully digital? I mean all the VST plugins that one can use on a PC (that is what I do btw)
I would be surprised if those had any impact on SINAD for example (but I might be wrong)

 

[voodooless]

Even those that are fully digital? I mean all the VST plugins that one can use on a PC (that is what I do btw)
I would be surprised if those had any impact on SINAD for example (but I might be wrong)

That will highly depend on the plugin. Some may actually be designed to lower SINAD

 

[ppataki]

That will highly depend on the plugin. Some may actually be designed to lower SINAD

True - but take some well-known ones like Pro Q3, CraveEQ, Jriver's PEQ, etc. if you use those as crossover (among others, just for the sake of the example) I reckon there will be zero loss

 

[voodooless]

True - but take some well-known ones like Pro Q3, CraveEQ, Jriver's PEQ, etc. if you use those as crossover (among others, just for the sake of the example) I reckon there will be zero loss

There is never zero loss. By definition, you do these things to add distortion The problem is when it becomes harmonic, which isn't usually a thing anyway. It's more about the noise bit of the SINAD. Generally, these things use enough bits to not need to worry about this. The Jriver PEQ uses 64-bit float math, so nothing to worry about there. But even then, as soon as you lower gain, you will lose about the same equivalent of SINAD points.

 

[Sokel]

One of the benefits analog el. x-overs may have is that if chosen right they can act as a gain stage to elevate a weaker signal,after all it's some op-amps they use so no reason for not been clean other than the filters.
For example a 2V traditional unbal RCA dac signal can easily make it to 4V or much higher unlike the 0db fixed digital limitation.
So no reason for lowering the overall volume of the system if for example positive gain is applied.

Now,there may be DSP x-over with generous output or pre-amp (in terms of adding gain) ability but their price is usually much higher.

 

[ppataki]

There is never zero loss. By definition, you do these things to add distortion The problem is when it becomes harmonic, which isn't usually a thing anyway. It's more about the noise bit of the SINAD. Generally, these things use enough bits to not need to worry about this. The Jriver PEQ uses 64-bit float math, so nothing to worry about there. But even then, as soon as you lower gain, you will lose about the same equivalent of SINAD points.

I guess this is still the 'cleanest' way to go when it comes to crossovers and DSP in general

 

[Sokel]

On the second page there's also some dirty measurements of a behri which is not nice.
Amongst the ones I have measured the best was a Marchand XM9 which simply maxed my ancient interface,

Talk is cheap and I thought of dropping couple of quick and dirty ones of a Marchand XM9

Test rig is E-MU 0204 ADC>Marchand XM9 with its x-over freq at 240Hz>E-MU 0204 DAC,the rest of the settings are all visible in the pics:



60Hz



..same with REW for comparison (but at 48Khz,forgot to change that but it makes no difference)

Bonus a 20Hz one where a lot of x-overs struggle (yep,I have seen lots) :



..and the proof that my ancient interface is maxed,so a loopback of it's I/O at the exact same configuration:





Marchand seem to improve it!

Not bad at all in my book.

 

[Theta]

Do we have measurements on any analog crossovers?
I guess they would have an impact on the signal in terms of SINAD, frequency response, etc.
If so, then that would be yet another disadvantage

Regarding volume control: the only place I control volume is in Jriver (internal 64-bit volume control) using its Volume Protection feature - that way the volume can never be maxed accidentally (for that matter it is not even easy to max it even deliberately)

Yes Amir did a review on a DBX analog crossover, the numbers were surprisingly good.