Hi. I'm a (small) studio owner and enthusiast working on a lot of audio related DIY projects. I'd like to get my feet wet with three way active speakers (i.e. active crossover and separate amplifiers) built on ATC drivers. Something like a wilmslow system, but activated.
What I'd like to ask is if anyone knows how the active crossovers are done in, say, activated SCM100/SCM150 and what kind of amps are specified for that kind of design.
I didn't find any public info shared here nor did anyone at ATC respond to my pleas..
The project is some months away, but I really like to read up on projects before starting them. In any case, if you have some info and are willing to share, please do
Cheers,
B.
What I'd like to ask is if anyone knows how the active crossovers are done in, say, activated SCM100/SCM150 and what kind of amps are specified for that kind of design.
I didn't find any public info shared here nor did anyone at ATC respond to my pleas..
The project is some months away, but I really like to read up on projects before starting them. In any case, if you have some info and are willing to share, please do
Cheers,
B.
Hi baadc0de
I am currently doing a similar project using ATC drivers. Though i don't have any specific information to offer about how ATC do things in their commercial designs, i can tell you my thoughts and conclusions so far.
I originally intended to built an active SCM-50A clone type speaker but using a ribbon for the tweeter. Having got hold of some 12" studio bass units, i started thinking instead to do a SCM-100A clone, and am now actually thinking of a four way. As yet no speaker is actually built, i just have all the parts. (mid range is the ATC "super").
I don't know how much experience you have building speakers. Myself, it's my first go. For crossovers, i didn't see much point going analogue active as ATC do. Digital crossovers are very flexible for DIY and quite cheap. I bought a Behringer DCX2496 which can do 3-way. I have one with the Jan Didden volume control. But since i started thinking 4-way, i actually preferred the Minidsp solution and have recently bought two of those, and the minidigi input boards. This kit will give you lots of flexibility to tailor the crossovers and get the sound you want. Coupled with room measurement software and a calibrated mic, you should be able to get good results - i say 'should' - i haven't done it yet. In the ATC designs, i believe there is some equalisation employed to allow the mid-dome to be used as low as they do, something like 380Hz. With a digital crossover you can experiment with all that. You could go sealed and use a linkwitz transform now available on minidsp.
For power amps, i built a number of them from a design in a book. But it takes a lot of time. There are many good modules out there from various manufacturers, its a no brainer to buy those if time is a constraint. Connexelectronic worth considering. Depending on budget, Hypex modules another option. I think these are going to compete very well with anything ATC puts in (which i think are class A/B). Perhaps a bridged stereo amp for the bass and another stereo amp for mid and treble. Digital amps typically need less heatsinking.
In terms of cabinets, i think wilmslow will allow you to buy just the CAD drawings. Speaking with them a while ago, they said the stated 'liter' volume on their website includes a compartment for housing the crossovers. So in reality, less volume is needed.
As i said, i don't know your experience with speaker building, perhaps all this is already well known to you.
Good luck with your project, please post some pictures if you have some success.
Also, have you seen this project with ATC drivers?:
project 10
I am currently doing a similar project using ATC drivers. Though i don't have any specific information to offer about how ATC do things in their commercial designs, i can tell you my thoughts and conclusions so far.
I originally intended to built an active SCM-50A clone type speaker but using a ribbon for the tweeter. Having got hold of some 12" studio bass units, i started thinking instead to do a SCM-100A clone, and am now actually thinking of a four way. As yet no speaker is actually built, i just have all the parts. (mid range is the ATC "super").
I don't know how much experience you have building speakers. Myself, it's my first go. For crossovers, i didn't see much point going analogue active as ATC do. Digital crossovers are very flexible for DIY and quite cheap. I bought a Behringer DCX2496 which can do 3-way. I have one with the Jan Didden volume control. But since i started thinking 4-way, i actually preferred the Minidsp solution and have recently bought two of those, and the minidigi input boards. This kit will give you lots of flexibility to tailor the crossovers and get the sound you want. Coupled with room measurement software and a calibrated mic, you should be able to get good results - i say 'should' - i haven't done it yet. In the ATC designs, i believe there is some equalisation employed to allow the mid-dome to be used as low as they do, something like 380Hz. With a digital crossover you can experiment with all that. You could go sealed and use a linkwitz transform now available on minidsp.
For power amps, i built a number of them from a design in a book. But it takes a lot of time. There are many good modules out there from various manufacturers, its a no brainer to buy those if time is a constraint. Connexelectronic worth considering. Depending on budget, Hypex modules another option. I think these are going to compete very well with anything ATC puts in (which i think are class A/B). Perhaps a bridged stereo amp for the bass and another stereo amp for mid and treble. Digital amps typically need less heatsinking.
In terms of cabinets, i think wilmslow will allow you to buy just the CAD drawings. Speaking with them a while ago, they said the stated 'liter' volume on their website includes a compartment for housing the crossovers. So in reality, less volume is needed.
As i said, i don't know your experience with speaker building, perhaps all this is already well known to you.
Good luck with your project, please post some pictures if you have some success.
Also, have you seen this project with ATC drivers?:
project 10
Thanks for the responses guys.
No, this is my first DIY speaker project.. it's to be built as a stereo reference pair to have in a control room of a project studio, so I'd like to get close to the baseline spec of an active ATC system and then tweak from this 'known good' configuration..
That said, a behringer digital xover could to the trick, I might do that actually. It's available off the shelf, is configurable and once I'm happy with the settings I can just make myself an analogue counterpart. DIYing a few class A huge power amplifiers shouldn't be too much of a problem either (I like the challenge and have built tons of studio gear, even guitar tube amplifiers).
ATC "super" - that's not the one included (or available) from the wilmslow site, right?
Cheers!
B
No, this is my first DIY speaker project.. it's to be built as a stereo reference pair to have in a control room of a project studio, so I'd like to get close to the baseline spec of an active ATC system and then tweak from this 'known good' configuration..
That said, a behringer digital xover could to the trick, I might do that actually. It's available off the shelf, is configurable and once I'm happy with the settings I can just make myself an analogue counterpart. DIYing a few class A huge power amplifiers shouldn't be too much of a problem either (I like the challenge and have built tons of studio gear, even guitar tube amplifiers).
ATC "super" - that's not the one included (or available) from the wilmslow site, right?
Cheers!
B
Baadc0de, just build it up with the DCX either leaving it there or doing the analogue clone later as you said. If ATC have specific filters or EQ designed into their electronics you won't ever know unless you measure their implementation and copy it.
Just build what you want, measure them and tweak to taste.
I heard an active set of ATC about 10 years ago, don't recall the model # now, but 15" 3 way based and I thought they were excellent, but too much for my wallet. This was in one of the now defunct audio stores here and I thought they were better than many of the much more expensive systems on offer there.
I will enjoy watching a build thread for these.
Just build what you want, measure them and tweak to taste.
I heard an active set of ATC about 10 years ago, don't recall the model # now, but 15" 3 way based and I thought they were excellent, but too much for my wallet. This was in one of the now defunct audio stores here and I thought they were better than many of the much more expensive systems on offer there.
I will enjoy watching a build thread for these.
Latency analogue in-out is specified at <1ms but it can set >500ms from memory.
Could be similar to the/a passive xover. But probably is not, because there's not the same issues with phase as you have in passives vs. amps. Only the phase from the point source/driver to driver and not from passive xover components.
http://www.flickr.com/photos/tanais/262821257/in/set-72157600272689589/
Last edited:
Unfortunately Wilmslow don't sell the "super" mid range dome (SM75-150S). I wonder though how much difference there really is between the "normal" version they sell and the super. And the super is a lot more money too. If you do want the super, you might be able to get it direct from ATC. I'm not sure what they're policy is, i know they restrict distribution of their Super Linear Bass drivers but not sure about the mid range. (they quoted me around £470 inc VAT per unit if i remember correctly).
There are some SM75-150 currently on ebay UK:
ATC SM75-150 midrange Mitteltöner pair speaker on eBay (end time 26-Oct-10 16:07:17 BST)
Just remember if you go the DCX route, and want to use its digital input, you'll need some sort of ganged volume control for it (e.g. Jan Didden design). Otherwise, you can just use the analogue ins with a stereo pre-amp before the DCX.
The DCX has numerous filter curves. If you were to prototype using the LR24 curves then it should be quite straight forward to transfer it over to the Elliot Sounds LRX424 analogue active crossover PCBs at a later stage:
Project 125
There are some SM75-150 currently on ebay UK:
ATC SM75-150 midrange Mitteltöner pair speaker on eBay (end time 26-Oct-10 16:07:17 BST)
Just remember if you go the DCX route, and want to use its digital input, you'll need some sort of ganged volume control for it (e.g. Jan Didden design). Otherwise, you can just use the analogue ins with a stereo pre-amp before the DCX.
The DCX has numerous filter curves. If you were to prototype using the LR24 curves then it should be quite straight forward to transfer it over to the Elliot Sounds LRX424 analogue active crossover PCBs at a later stage:
Project 125
Browsing on the 'net it seems that ATC restrict usage of their 'SL' series drivers to their own speakers and 'S' are hard to get (and as mrwireless pointed out, they're not cheap at all). I'd still like to get as close to the SL using line as possible for starters and then tweak from there, so if anyone has a suggestion, don't hesitate to voice it.
The idea of prototyping on a DCX and then transferring that to lrx424 is exactly what I had in mind. How fortunate that there's a compatible PCB project ready made.
Thank's a lot for the suggestions made thus far
The idea of prototyping on a DCX and then transferring that to lrx424 is exactly what I had in mind. How fortunate that there's a compatible PCB project ready made.
Thank's a lot for the suggestions made thus far
I have worked and developed product around ATC drivers and "cloned" the SCM100 albeit with a ribbon tweeter.
As a loudspeaker developer I have to say I would not go the route of ATC drivers, particularly the bass drivers..I think it is way past time we moved on from heavy cones and reflex loading!! Low efficiency as well as limited power handling by todays standards.
Anyway as for the lower efficiency SCM 75 dome..(is the high efficiency one still available with the rarity of alnico these days? Perhaps they are using neodymium), it has it's resonance at around 350Hz from memory and they suggest it is well damped and not audible enabling a crossover of 380Hz or so.
Well, it is clearly audible in my view, not as a response hump although there is one, but as a nasal quality. I would suggest two courses of action. Firstly try using the loudspeaker manager at 48dB/Octave as this may do the trick or build an RCL network across the driver to take out the impedance resonance. Back in the day I found I could apply a passive network with great success once the dome impedance was levelled..sounded far more transparent and open as well!
As a loudspeaker developer I have to say I would not go the route of ATC drivers, particularly the bass drivers..I think it is way past time we moved on from heavy cones and reflex loading!! Low efficiency as well as limited power handling by todays standards.
Anyway as for the lower efficiency SCM 75 dome..(is the high efficiency one still available with the rarity of alnico these days? Perhaps they are using neodymium), it has it's resonance at around 350Hz from memory and they suggest it is well damped and not audible enabling a crossover of 380Hz or so.
Well, it is clearly audible in my view, not as a response hump although there is one, but as a nasal quality. I would suggest two courses of action. Firstly try using the loudspeaker manager at 48dB/Octave as this may do the trick or build an RCL network across the driver to take out the impedance resonance. Back in the day I found I could apply a passive network with great success once the dome impedance was levelled..sounded far more transparent and open as well!
Since you are designing for a studio, I will assume that you have an ASIO analog in/out 24/96 interface available. Take a look at Frequency Allocator/Arbitrator from Transient Perfect loudspeaker DSP tools . Windows DSP speaker crossovers. It is a VST plug in that performs XO/EQ and phase adjustment to make transient perfect crossovers.
I've owned it a while and use an M-Audio FW410 for output operating at 24/96. Allocator allows you to import .frd response measurements for each driver and then add HP/LP filters up to 7th order, up to 4 PEQ and two shelving filters per frequency band. Each band's attenuation and global delay are adjustable. It took me just a few minutes to dial in the crossover and measure the result. A couple iterations for power response and the results were significantly better than my analog active crossover in a quarter of the time it takes just to stuff the crossover boards. Clear out the .frd files and you get a graphic representation of the transfer function if you want to go analog.
The arbitrator (phase adjustment) portion has a ~1.85 msec latency IIRC. The Allocator light version has very low latency, but you give up the transient perfect capability. I'm not sure if I hear the benefits of transient perfect designs.
There are free demos available, and the price is negligible if you already have an ASIO I/O. Running it on a 3 GHz Celeron with only a Gig of Ram I haven't had any glitching while simultaneously running a media player, browser, email and word processor.
I'm just a big fan of the software, unrelated.
I've owned it a while and use an M-Audio FW410 for output operating at 24/96. Allocator allows you to import .frd response measurements for each driver and then add HP/LP filters up to 7th order, up to 4 PEQ and two shelving filters per frequency band. Each band's attenuation and global delay are adjustable. It took me just a few minutes to dial in the crossover and measure the result. A couple iterations for power response and the results were significantly better than my analog active crossover in a quarter of the time it takes just to stuff the crossover boards. Clear out the .frd files and you get a graphic representation of the transfer function if you want to go analog.
The arbitrator (phase adjustment) portion has a ~1.85 msec latency IIRC. The Allocator light version has very low latency, but you give up the transient perfect capability. I'm not sure if I hear the benefits of transient perfect designs.
There are free demos available, and the price is negligible if you already have an ASIO I/O. Running it on a 3 GHz Celeron with only a Gig of Ram I haven't had any glitching while simultaneously running a media player, browser, email and word processor.
I'm just a big fan of the software, unrelated.
Why are we now talking about computers?!
I know I went off a bit tangentially as well from the OP but my point was not regarding the accuracy of DSP or analogue active crossovers or the computer analysis of performance although this does raise interesting issues.
My work with ATC drivers for my own speakers was in the late 80's. I had the pleasure of visiting the ATC factory and meeting the designer of the ATC drivers back in the late 70's when they were fledglings. I understand the dome has been under continuous development probably ever since!!
Because it is not about how flat you can get it by measurement and eq or prediction in software or RTA or gated RTA or anything else you care to mention. Mechanics plays an important role. You can flatten a response all you like but if the dome resonance is actually causing problems in terms of an amplifier's ability to damp oscillation or instability, it doesn't matter how flat it looks with response correction.
Removing the dome resonance with a passive RCL ensures that you can drive the dome pretty much to it's free air resonance. Traditionally it was always considered that you should cross a driver over at least two octaves from it's resonance..this of course would make the dome pretty useless! An RCL should or was often employed with dome tweeters, but fluid coil damping generally got around the problem to a large extent.
Listening tests with the RCL revealed a more natural midrange that now blended to create a cohesive three dimensional sound stage no longer showing up a colouration and slight honk in the lower mid particulary effecting male vocal. Listening with ears and not a computer prediction!!
I was even able to create a full blown 4th order passive (final response) 3 way network!! OK, the component cost was astronomical and it just fitted a 4u high rack case!
But it was the 80's and the ATC outboard active crossover was 'orrible as it sat there and hissed at you..plus of course without the RCL on the dome it did nothing to combat the resonance/damping issue and nor of course did it have EQ, because as pointed out some shelving was necessary and fixed active crossover points don't do that!
For the active set up they considered their market to be pro who would expect to have to EQ the system anyway?
And I only had one £4000 Threshold Stasis (still do!). It was clearly obvious to all that listened that the Threshold/Passive combination sounded like music and the analogue active + three ok stereo power amps (Yamaha P series) sounded artificial and entirely less rewarding. The current owner of the passive set still has a smile on his face some 15 years after buying them from me. Such is the load and correct tuning that he is able to drive the system with one Quad 405 and I can concur that the sound is just open, detailed relaxed and accurate.
It's what it is all about, isn't it?
Further proof of the low efficiency dome issue was was that the high efficiency dome reduced this problem substantially by having a much more powerful magnet assembly. Looking at their original plots the dome rolled off smoothly in response through resonance, a sign of optimised components used in the manufacture and very much one of the black arts of actual driver design. Coil weight, flux, amount of damping fluid on the fabric dome, the cross section of the dome material, dome weight are just a few of the criteria.
These days it is true to some extent that managing a loudspeaker in DSP is a good method but you must take into account you are still relying on an A/D unless you have digital in, and a D/A on the way out thus not enabling you to use a "boutique" quality D/A should you choose. You are stuck with the D/A converters in the unit.
Most loudspeaker management systems are designed to be implemented in live or location set up for which they are superb, but it negates your ability to go "high end" and select a separate D/A converter for each section driven...Benchmark, Sonifex, Apogee, RME, Mutec and the humble but superb Cambridge DAC Magic to name a few.
Only the top drawer DBX loudspeaker managers have the ability to go digital out per channel..there may be others but I am not aware of them?
Incidentally, whilst we are on the subject I have to confess to being one of the first to introduce a studio monitor loudspeaker which was managed digitally. It was the Omen Ra 1812 and those wishing to read about it may like to find the Hugh RobJohns review in "Line Up", Aug/Sep 2004..there is a whole lot more to loudspeaker design then what a computer can tell you!! Suffice it to say, with enormous pride, my design was very favourably compared to ATC, PMC and Genelec, Good company to keep I think.
Sadly the company I designed for did nothing to actually sell product so it has disappeared into history! But time has not stood still and I have continued to research and create product for my own operation..this time with a difference. We have moved on from just stand alone designs and now re-appraised some of the problems of the "hot seat" and how to create even response throughout different listening areas..those interested in cardioid, asymmetric dipoles, run times and predicted..(yes, even me!) polar response may wish to "watch this space...."!
As a final aside..our review pair of 1812s submitted to Hugh RobJohns was delivered with two DSP loudspeaker management systems..the BSS366T Omnidrive and the Behringer Ultracurve at 1/10th of the cost. The two units were set up identically. It would be churlish to suggest you go and read the review to find out his comments regarding the subjective quality of the two units, so I will put you out of your misery. With this set up, the Behringer won!!
All is not what it should be in the world of sound reproduction..whatever the computer may tell you!
I know I went off a bit tangentially as well from the OP but my point was not regarding the accuracy of DSP or analogue active crossovers or the computer analysis of performance although this does raise interesting issues.
My work with ATC drivers for my own speakers was in the late 80's. I had the pleasure of visiting the ATC factory and meeting the designer of the ATC drivers back in the late 70's when they were fledglings. I understand the dome has been under continuous development probably ever since!!
Because it is not about how flat you can get it by measurement and eq or prediction in software or RTA or gated RTA or anything else you care to mention. Mechanics plays an important role. You can flatten a response all you like but if the dome resonance is actually causing problems in terms of an amplifier's ability to damp oscillation or instability, it doesn't matter how flat it looks with response correction.
Removing the dome resonance with a passive RCL ensures that you can drive the dome pretty much to it's free air resonance. Traditionally it was always considered that you should cross a driver over at least two octaves from it's resonance..this of course would make the dome pretty useless! An RCL should or was often employed with dome tweeters, but fluid coil damping generally got around the problem to a large extent.
Listening tests with the RCL revealed a more natural midrange that now blended to create a cohesive three dimensional sound stage no longer showing up a colouration and slight honk in the lower mid particulary effecting male vocal. Listening with ears and not a computer prediction!!
I was even able to create a full blown 4th order passive (final response) 3 way network!! OK, the component cost was astronomical and it just fitted a 4u high rack case!
But it was the 80's and the ATC outboard active crossover was 'orrible as it sat there and hissed at you..plus of course without the RCL on the dome it did nothing to combat the resonance/damping issue and nor of course did it have EQ, because as pointed out some shelving was necessary and fixed active crossover points don't do that!
For the active set up they considered their market to be pro who would expect to have to EQ the system anyway?
And I only had one £4000 Threshold Stasis (still do!). It was clearly obvious to all that listened that the Threshold/Passive combination sounded like music and the analogue active + three ok stereo power amps (Yamaha P series) sounded artificial and entirely less rewarding. The current owner of the passive set still has a smile on his face some 15 years after buying them from me. Such is the load and correct tuning that he is able to drive the system with one Quad 405 and I can concur that the sound is just open, detailed relaxed and accurate.
It's what it is all about, isn't it?
Further proof of the low efficiency dome issue was was that the high efficiency dome reduced this problem substantially by having a much more powerful magnet assembly. Looking at their original plots the dome rolled off smoothly in response through resonance, a sign of optimised components used in the manufacture and very much one of the black arts of actual driver design. Coil weight, flux, amount of damping fluid on the fabric dome, the cross section of the dome material, dome weight are just a few of the criteria.
These days it is true to some extent that managing a loudspeaker in DSP is a good method but you must take into account you are still relying on an A/D unless you have digital in, and a D/A on the way out thus not enabling you to use a "boutique" quality D/A should you choose. You are stuck with the D/A converters in the unit.
Most loudspeaker management systems are designed to be implemented in live or location set up for which they are superb, but it negates your ability to go "high end" and select a separate D/A converter for each section driven...Benchmark, Sonifex, Apogee, RME, Mutec and the humble but superb Cambridge DAC Magic to name a few.
Only the top drawer DBX loudspeaker managers have the ability to go digital out per channel..there may be others but I am not aware of them?
Incidentally, whilst we are on the subject I have to confess to being one of the first to introduce a studio monitor loudspeaker which was managed digitally. It was the Omen Ra 1812 and those wishing to read about it may like to find the Hugh RobJohns review in "Line Up", Aug/Sep 2004..there is a whole lot more to loudspeaker design then what a computer can tell you!! Suffice it to say, with enormous pride, my design was very favourably compared to ATC, PMC and Genelec, Good company to keep I think.
Sadly the company I designed for did nothing to actually sell product so it has disappeared into history! But time has not stood still and I have continued to research and create product for my own operation..this time with a difference. We have moved on from just stand alone designs and now re-appraised some of the problems of the "hot seat" and how to create even response throughout different listening areas..those interested in cardioid, asymmetric dipoles, run times and predicted..(yes, even me!) polar response may wish to "watch this space...."!
As a final aside..our review pair of 1812s submitted to Hugh RobJohns was delivered with two DSP loudspeaker management systems..the BSS366T Omnidrive and the Behringer Ultracurve at 1/10th of the cost. The two units were set up identically. It would be churlish to suggest you go and read the review to find out his comments regarding the subjective quality of the two units, so I will put you out of your misery. With this set up, the Behringer won!!
All is not what it should be in the world of sound reproduction..whatever the computer may tell you!
Cool. Schematic looks like pretty textbook LR2 with phase adjustment. Where's the magic?
Only brought up computers because OP was asking for studio use, which is often computer based these days. Windoze vs. Mac seems to be pretty pertinent there, too if programs don't support windows operation and I proposed a windows solution.
I wholeheartedly agree that flat on axis response is hardly the last word. Same with measurements. I just wanted to point out how quickly I got to the point where I didn't feel like fiddling any more and wanted to listen to my entire library.
Only brought up computers because OP was asking for studio use, which is often computer based these days. Windoze vs. Mac seems to be pretty pertinent there, too if programs don't support windows operation and I proposed a windows solution.
I wholeheartedly agree that flat on axis response is hardly the last word. Same with measurements. I just wanted to point out how quickly I got to the point where I didn't feel like fiddling any more and wanted to listen to my entire library.
Last edited:
Just curious... is the RCL across the dome primarily for improving tube-amp setups, or would there also be benefits with solid-state ?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.