Hi all,
Thought I'd post this up here.
I'm pretty sure I'm going to make the switch and go for a passive crossover for my current PA speakers, and wanted to document that journey.
They're in-use as side-fills here:
On top of a 15" subwoofer.
The drivers are as follows:
2x Faital Pro 10FH520
1x 18Sound ND1460
1x RCF HF94 horn
Currently, I'm using a DCX2496 to provide crossover duties, feeding into a Behringer NU6000 for mids and two channels of an NU4-6000 for highs. The other two channels are used for a couple of monitors, or bridged for subs on smaller gigs.
I'd like to consolidate my racks somewhat and take a step up in quality.
In order to do that I can...
1 - Find a 4-channel amplifier capable of driving the 10FH520s to their full potential (I'd like at least a kilowatt for each driver). It'd also need crossover functionality.
2 - Go for a passive crossover and use a stereo amplifier capable of driving the whole shebang.
For option 1, something like a Lab Gruppen PLM10000Q is an option. So is a Powersoft X4. However, those are both very expensive, and I wouldn't be using all of the potential power output - the high-frequency driver is very efficient, so the extra two multi-kilowatt channels are almost wasted.
Option 2 opens up a much wider range of amplifier options, with similar quality to the above amps available at a much lower cost, due to only needing stereo operation. It also bypasses a round of AD/DA conversion, which has to be a plus.
So, I measured the drivers and fired up XSim.
The lowpass section is very simple. The 10FH520 drivers have a fairly flat impedance curve, so it's an easy driver to work with. There's a bit of a peak around 2kHz, but this crossover shunts that ~20dB down. Once built, I might try adding a notch filter. I avoided that for the now, as any series inductors have to stand a lot of current.
The highpass is more complicated.
Going from the amp terminals, we have a shelving filter that cuts below 8kHz or so. The HF driver starts losing sensitivity up there, so that little bit helps flatten the response.
Next is a 4th order highpass. I went for that because I want to make sure the HF driver stays comfortable. They're probably not textbook values, but they work for this driver/horn combo. It's worth noting that the minimum recommended crossover is 800Hz and 12dB/octave, so at the moment I'm going a smidge higher, and using a steeper slope.
Lastly, there's an LCR network, which is for flattening the impedance of the compression driver, as seen by the crossover. The HF driver has an impedance curve that swings around quite a lot, reaching around 40ohm at 480Hz. There's also another peak at 1.1kHz. The network flattens the impedance in that region quite well. If you imagine a potential divider with the crossover network as one side and the driver the other, having the driver suddenly go up to 40ohm at 480Hz means it's going to receive quite a lot of signal there. The result is extra heat in the driver, as well as more excursion (ie, more distortion). The network cuts down the amount of work the driver is doing at 480Hz by around 9dB.
It's worth noting that impedance does drop low around 250Hz. The drivers themselves get to around 2.5ohm there, and the low-pass section drops it a little further. That's fine for the sort of amps I'll be using, but if anyone else is following along, that could be a consideration. You could put the drivers in series for a HiFi/low-power situation, but these cabinets are my primary main PA speakers, and I need them to go as loud as they can.
Last bit to look at is component power dissipation. These speakers are going to be on the receiving end of some fairly big amplifiers, so we need to make sure they're going to be fine with that sort of use on a long-term basis. I don't want the crossover to simply act as a series of fuses.
So, here goes:
Obviously, the midbass drivers are going to be dissipating most of the power. The compression driver is also seeing quite a lot.
R1, on the RC low-shelf filter is also dissipating a lot of power. What I might do there is bypass it with a switch, and then come up with a replacement EQ curve. That way, I can have a cabinet that's flat, or gain a little efficiency and EQ the louder bit down. Shorting that resistor alters the HF level by around 4dB, but the crossover summation stays good.
The lowpass inductor is seeing a lot of current, so I might series/parallel a few inductors to get the individual component dissipation down. That's additional cost and weight, but would make sure everything stays nice and cool.
That's all for now. Stay tuned.
Chris
Thought I'd post this up here.
I'm pretty sure I'm going to make the switch and go for a passive crossover for my current PA speakers, and wanted to document that journey.
They're in-use as side-fills here:
On top of a 15" subwoofer.
The drivers are as follows:
2x Faital Pro 10FH520
1x 18Sound ND1460
1x RCF HF94 horn
Currently, I'm using a DCX2496 to provide crossover duties, feeding into a Behringer NU6000 for mids and two channels of an NU4-6000 for highs. The other two channels are used for a couple of monitors, or bridged for subs on smaller gigs.
I'd like to consolidate my racks somewhat and take a step up in quality.
In order to do that I can...
1 - Find a 4-channel amplifier capable of driving the 10FH520s to their full potential (I'd like at least a kilowatt for each driver). It'd also need crossover functionality.
2 - Go for a passive crossover and use a stereo amplifier capable of driving the whole shebang.
For option 1, something like a Lab Gruppen PLM10000Q is an option. So is a Powersoft X4. However, those are both very expensive, and I wouldn't be using all of the potential power output - the high-frequency driver is very efficient, so the extra two multi-kilowatt channels are almost wasted.
Option 2 opens up a much wider range of amplifier options, with similar quality to the above amps available at a much lower cost, due to only needing stereo operation. It also bypasses a round of AD/DA conversion, which has to be a plus.
So, I measured the drivers and fired up XSim.
The lowpass section is very simple. The 10FH520 drivers have a fairly flat impedance curve, so it's an easy driver to work with. There's a bit of a peak around 2kHz, but this crossover shunts that ~20dB down. Once built, I might try adding a notch filter. I avoided that for the now, as any series inductors have to stand a lot of current.
The highpass is more complicated.
Going from the amp terminals, we have a shelving filter that cuts below 8kHz or so. The HF driver starts losing sensitivity up there, so that little bit helps flatten the response.
Next is a 4th order highpass. I went for that because I want to make sure the HF driver stays comfortable. They're probably not textbook values, but they work for this driver/horn combo. It's worth noting that the minimum recommended crossover is 800Hz and 12dB/octave, so at the moment I'm going a smidge higher, and using a steeper slope.
Lastly, there's an LCR network, which is for flattening the impedance of the compression driver, as seen by the crossover. The HF driver has an impedance curve that swings around quite a lot, reaching around 40ohm at 480Hz. There's also another peak at 1.1kHz. The network flattens the impedance in that region quite well. If you imagine a potential divider with the crossover network as one side and the driver the other, having the driver suddenly go up to 40ohm at 480Hz means it's going to receive quite a lot of signal there. The result is extra heat in the driver, as well as more excursion (ie, more distortion). The network cuts down the amount of work the driver is doing at 480Hz by around 9dB.
It's worth noting that impedance does drop low around 250Hz. The drivers themselves get to around 2.5ohm there, and the low-pass section drops it a little further. That's fine for the sort of amps I'll be using, but if anyone else is following along, that could be a consideration. You could put the drivers in series for a HiFi/low-power situation, but these cabinets are my primary main PA speakers, and I need them to go as loud as they can.
Last bit to look at is component power dissipation. These speakers are going to be on the receiving end of some fairly big amplifiers, so we need to make sure they're going to be fine with that sort of use on a long-term basis. I don't want the crossover to simply act as a series of fuses.
So, here goes:
Obviously, the midbass drivers are going to be dissipating most of the power. The compression driver is also seeing quite a lot.
R1, on the RC low-shelf filter is also dissipating a lot of power. What I might do there is bypass it with a switch, and then come up with a replacement EQ curve. That way, I can have a cabinet that's flat, or gain a little efficiency and EQ the louder bit down. Shorting that resistor alters the HF level by around 4dB, but the crossover summation stays good.
The lowpass inductor is seeing a lot of current, so I might series/parallel a few inductors to get the individual component dissipation down. That's additional cost and weight, but would make sure everything stays nice and cool.
That's all for now. Stay tuned.
Chris
One thing to keep in mind is that active speakers need less power ratings than passive.
Power is proportional to voltage squared. This is in addition to power losses due to resistors, crossover components, etc.
Also, make sure you benchmark your speakers now, so your final design measures similarly.
Best,
E
Power is proportional to voltage squared. This is in addition to power losses due to resistors, crossover components, etc.
Also, make sure you benchmark your speakers now, so your final design measures similarly.
Best,
E
I can't imagine that the passive approach will offer either the flexibility or ultimately the performance of the all active approach.
I can well understand the desire to simplify a bit, I just went from 3 way passive to 3 way active and I am still getting used to all of the electronics required to do it. (All my own design - amps and line level analog LR4 3 way XO) Despite six channels of amplification and the heat generated and space consumed I won't be going back. The difference in performance is fairly staggering despite using really good parts in the old passive XOs.
Perhaps more compact amplification is the answer..
Good luck!
Edit: I reread some of Erik's recommendations, excellent advice and very pragmatic. I used many of the same techniques in my original passive XO design. His other observations are on point if you feel you need to go this route.
I can well understand the desire to simplify a bit, I just went from 3 way passive to 3 way active and I am still getting used to all of the electronics required to do it. (All my own design - amps and line level analog LR4 3 way XO) Despite six channels of amplification and the heat generated and space consumed I won't be going back. The difference in performance is fairly staggering despite using really good parts in the old passive XOs.
Perhaps more compact amplification is the answer..
Good luck!
Edit: I reread some of Erik's recommendations, excellent advice and very pragmatic. I used many of the same techniques in my original passive XO design. His other observations are on point if you feel you need to go this route.
Hi all,
Wow, this got some replies quickly!
Okay, so in no particular order...
Kevin, while I take your point about the flexibility of active systems, there is still DSP available in the form of the output processing on the digital mixing desk. I think the quality improvement will be in going from a DCX2496 into iNukes to a single good amplifier. The ultimate would be something like a Powersoft X4 running all the crossover/EQ duties, but they cost an awful lot more than a stereo amp of similar quality and output. There's also the problem of eggs and baskets. If those speakers can only ever run on that amplifier with the right processing, I'm pretty knackered if the amp goes down. A passive speaker can be driven from anything - I always take a backup amp, so being able to re-patch some SpeakOn and get the main speakers up and running is really useful.
Erik, thanks for the hint on the cap on the lowpass section. I'll have a play around with that for sure. I appreciate that passive speakers require more power input, but that won't be a huge issue for me. At the moment, I have a Crown MA5002VZ that could be used, but I'm looking for something with more power and less weight. There's a huge range to choose from, though, so I'm not going to sweat that just yet.
For now, I'm going to get the crossover built as an external unit, and do a shoot out between the DCX & iNukes vs a Crown MA12000i and a passive crossover.
I'm betting the latter sounds better, simply because AD/DA steps have been eliminated, and the Crown is a better amplifier - more power, less distortion, doesn't care about HF impedance.
TBTL, I have some mics with those switches - Beyerdynamic M67. Absolutely love them.
Chris
Wow, this got some replies quickly!
Okay, so in no particular order...
Kevin, while I take your point about the flexibility of active systems, there is still DSP available in the form of the output processing on the digital mixing desk. I think the quality improvement will be in going from a DCX2496 into iNukes to a single good amplifier. The ultimate would be something like a Powersoft X4 running all the crossover/EQ duties, but they cost an awful lot more than a stereo amp of similar quality and output. There's also the problem of eggs and baskets. If those speakers can only ever run on that amplifier with the right processing, I'm pretty knackered if the amp goes down. A passive speaker can be driven from anything - I always take a backup amp, so being able to re-patch some SpeakOn and get the main speakers up and running is really useful.
Erik, thanks for the hint on the cap on the lowpass section. I'll have a play around with that for sure. I appreciate that passive speakers require more power input, but that won't be a huge issue for me. At the moment, I have a Crown MA5002VZ that could be used, but I'm looking for something with more power and less weight. There's a huge range to choose from, though, so I'm not going to sweat that just yet.
For now, I'm going to get the crossover built as an external unit, and do a shoot out between the DCX & iNukes vs a Crown MA12000i and a passive crossover.
I'm betting the latter sounds better, simply because AD/DA steps have been eliminated, and the Crown is a better amplifier - more power, less distortion, doesn't care about HF impedance.
TBTL, I have some mics with those switches - Beyerdynamic M67. Absolutely love them.
Chris
MA12000i? Budget doesn't seem to be a big issue to you. Guess that one can cook your Faitals in no time. Boy how I'd love to have all that resources available to play with!
Anyway - why not build 2 x UCD2K modules plus powerful PSU into a 19" case for the option 2? 1 x UCD 2K can give you >3KW of clean power into 4 Ohms - at a fraction of the Crown's cost - if the supply rail is stable and the cooling is adequate.
BTW: To get rid of a round of AD / DA is always a plus, but how the hell are you going to crossover between the 15" subwoofer and the 10"s?
Anyway - why not build 2 x UCD2K modules plus powerful PSU into a 19" case for the option 2? 1 x UCD 2K can give you >3KW of clean power into 4 Ohms - at a fraction of the Crown's cost - if the supply rail is stable and the cooling is adequate.
BTW: To get rid of a round of AD / DA is always a plus, but how the hell are you going to crossover between the 15" subwoofer and the 10"s?
I get the desire to simplify but going passive is a big downgrade, what you need to do is to make the boxes active with DSP driven plate amps and eliminate the amp rack completely. The iNukes are also well known for not sounding that great on anything besides subs so substituting just about anything else will improve SQ and SN ratios particularly in the mid/highs.
I actually got the MA12000i for a screaming bargain. The shop didn't have the power cord (20A IEC), so hadn't been able to test it. I haggled them down and managed to keep the returns policy and shop warranty intact. Lucky.
The UCD modules look nice. I'll do some research on pricing. It looks like they'll need heatsinks, fans, and the power supply will be non-trivial.
To cross between the subs and mains, I'll be going for Aux-fed subs from the desk. It can handle crossover and EQ duties just fine.
I've run things that way before and there are some benefits. Alternatively, I can put together a little analogue crossover in a box to sit between the desk and amps.
I'm not so sure passive crossovers are such a downgrade if they're done properly. I don't know if the crossover I've set out above is the best way to do things, but it looks pretty good to me.
Cabinets like the Danley SH46 are entirely passive, and get glowing reviews from users.
It won't be easy to make a passive crossover that good, and I plan a lot of testing to make sure I've got the active crossover beat.
I'm not keen on plate amps in speakers. Personal preference, but I prefer the flexibility and redundancy of seperate racks. I also like being able to monitor everything by looking in one place.
Point taken on iNukes. I could swap them out for nicer amps, but I'd still need more amps (and a processor) in the rack than if I went for a passive crossover. I'd rather spend the money on one good amplifier than trying to stretch the budget to two.
Anyway, the above is my reasoning for wanting to try the passive crossover route. I think it's possible to get good results from passive crossovers, and I think with enough work on my part I'll be able to get those results.
The only way to find out is to build the thing, so I'll be ordering some parts shortly.
Chris
DSP driven crossover gives you the advantage to change the settings with a mouse click instead of ordering new passive components and having to solder. Also, there is many more possibilities depending on the software, e. g. FIR filters, delay matching etc. what you cannot do with L R C. In both cases, passive as well as DSP, you should really understand what you are doing and it will be a lot of work to optimize in order to get the best result.
DSP disadvantage is the added conversion noise / distortion.
So that raises the question as to why you are not considering to simply get one of the best active speakers with dual range ADC, DSP and amp built in and call it a day.
DSP disadvantage is the added conversion noise / distortion.
So that raises the question as to why you are not considering to simply get one of the best active speakers with dual range ADC, DSP and amp built in and call it a day.
Hi Tom,
With regards to just buying a pair of active PA speakers, sure, I could do that. The RCF TTL-6A looks promising, and would probably go a couple of dB louder than the cabinets I have at the moment. They're also much bigger than my cabinets, and weigh twice as much.
This is DIY Audio, not Buy-It-From-RCF Audio.
I like the cabinets I have. You can do this: YouTube
with one per side, and the finish is just how I like it.
I'm not a complete amateur at this. I just think passive crossovers and a better amplifier will be the next system upgrade for me. Good passive crossovers are possible, and better amplification is certainly possible. What's the problem?
Chris
With regards to just buying a pair of active PA speakers, sure, I could do that. The RCF TTL-6A looks promising, and would probably go a couple of dB louder than the cabinets I have at the moment. They're also much bigger than my cabinets, and weigh twice as much.
This is DIY Audio, not Buy-It-From-RCF Audio.
I like the cabinets I have. You can do this: YouTube
with one per side, and the finish is just how I like it.
I'm not a complete amateur at this. I just think passive crossovers and a better amplifier will be the next system upgrade for me. Good passive crossovers are possible, and better amplification is certainly possible. What's the problem?
Chris
That is a superior speaker design though with mechanically optimized drivers .. they all feed a single horn to produce a coherent wavefront, that is a big part of what sets this speaker apart. I bet they would like to offer an active version but the speaker design doesn't lend itself to that and it would significantly impact cost and size of the product. No offence intended towards your work here I'm just comparing a standard PA design to a Synergy horn.
You have your work cut out for sure, speaking as someone that has come at this same problem from the other direction... I have some speakers(standard 12+1 reflex) that I designed as passive for simplicity(surprise) and I'm looking at ways to improve them with DSP and discrete amplification. I have tweaked the crossovers several times and have yet to come close to replicating what I can accomplish in 5 minutes with a DSP processor and a 2 channel power amp, I am an enthuastic DIYer but I'm no expert. I have other passive speakers that are powered and processed via a couple small racks of equipment and they produce vastly superior sound IMO but I don't want to go that route with these as they were designed with neo drivers to be light and easy to move, so a DSP equipped plate amp seems to be the most logical next step in my case. I also own a variety of modern brand name powered PA speakers that I use quite frequently and have come to appreciate the benefits they provide.
I think one of the things you may be overlooking is the size and weight of the passive components required for your project, I know the passive crossover added significant weight to my speakers.
All that said I do wish you luck, if nothing else it's a great learning experience, if you have more time and patience than I do maybe you will be able to match or exceed the current performance, it will take more time than you think though.
Hi Chris,
Passive? No way, my friend.
TD is a ******** genius at visualizing the entire electrical and acoustic interrelationship between passbands, driver selection, box construction, port locations in horn etc....
You gonna duplicate that without digging deep, starting over, all the way down?
And if you do design from ground zero, remember...you still have limiting to lay on top of the individual passbands....otherwise you're left simply triggering off the time interval implied by highest freq you need to protect, ....for the whole dang box ! ...
Seriously, screw passive....
Passive? No way, my friend.
TD is a ******** genius at visualizing the entire electrical and acoustic interrelationship between passbands, driver selection, box construction, port locations in horn etc....
You gonna duplicate that without digging deep, starting over, all the way down?
And if you do design from ground zero, remember...you still have limiting to lay on top of the individual passbands....otherwise you're left simply triggering off the time interval implied by highest freq you need to protect, ....for the whole dang box ! ...
Seriously, screw passive....
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.