Project advances. On Wednesday I'll go on a work trip to nearby city and pick up a MiniDSP 2x4 HD and UMIK-1 from a local retailer. My wallet cries blood for their cost, but it's about time I got one. And it will be absolutely necessary to simulate and plan the correct crossover with proper tools if I plan to run the synergy horn with passive crossovers.
Small update. The more I've thought of this project, the less viable the synergy horn seems, even though it's a very interesting take on the matter. It's perfectly doable with MiniDSP, but on my application the source would only be able to EQ one channel at the time, rest would be passive EQ. The total dealbreaker was that I got a super deal on a pair of Peerless SLS-12 drivers second-hand, so went for them. They are very musical and quite efficient, so I can easily battle the dips and peaks in room mode when I level match the drivers.
So basically, I'm back to square one with an idea of a FAST system. The FR candidates are SB65WBAC25-4, Visaton B80 and Scanspeak 10F/8414G10 and would be used in a short dagger TL. The woofer would be either tall sealed enclosure, or even a TL enclosure. Not sure how viable a TL would be considering that it's going to be corner-mounted on an enclosure with only 3 side panels.
Another thing that keeps me wondering is the choice of amplifiers. I've got Sanwu TDA8932 blocks, but I wonder if they have enough oomph to drive the 12" Peerless. Maybe a TDA7498 would be a better alternative? This would only be used to drive the woofer, so an amp which gives the best bass response would be the best for the application. For the FR, it's either the JLH 1969 clone I'm building from an eBay kit or some class AB amp. I'm open for suggestions since my experiences with amps are pretty limited 🙂
So basically, I'm back to square one with an idea of a FAST system. The FR candidates are SB65WBAC25-4, Visaton B80 and Scanspeak 10F/8414G10 and would be used in a short dagger TL. The woofer would be either tall sealed enclosure, or even a TL enclosure. Not sure how viable a TL would be considering that it's going to be corner-mounted on an enclosure with only 3 side panels.
Another thing that keeps me wondering is the choice of amplifiers. I've got Sanwu TDA8932 blocks, but I wonder if they have enough oomph to drive the 12" Peerless. Maybe a TDA7498 would be a better alternative? This would only be used to drive the woofer, so an amp which gives the best bass response would be the best for the application. For the FR, it's either the JLH 1969 clone I'm building from an eBay kit or some class AB amp. I'm open for suggestions since my experiences with amps are pretty limited 🙂
Last edited:
Congratulation the super deal on a pair of Peerless SLS-12 drivers, also like your mid/tweet candidates you have good taste : )
Metal verse paper verse fibre glass cone, are you going to try out all three before decision for final build.
SB65 should reach the highest HF point then B80 and then 8414, but from what i have seen about on axis natural HF roll off slope (uncorrected out of box) 8414 probably gives cleanest impulse response then B80 and then SB65. Have SB65 and 8424 myself and from own measurements it looks like SB65 have some high Q peaks up very high in frq that maybe could be corrected but it will need DSP running 88.2kHz or more to have bandwith for correction up there. Will have to try a pair of paper B80 down the road because at two of xrk971's comparison threads i actual picked it as my favor sound clip.
Metal verse paper verse fibre glass cone, are you going to try out all three before decision for final build.
SB65 should reach the highest HF point then B80 and then 8414, but from what i have seen about on axis natural HF roll off slope (uncorrected out of box) 8414 probably gives cleanest impulse response then B80 and then SB65. Have SB65 and 8424 myself and from own measurements it looks like SB65 have some high Q peaks up very high in frq that maybe could be corrected but it will need DSP running 88.2kHz or more to have bandwith for correction up there. Will have to try a pair of paper B80 down the road because at two of xrk971's comparison threads i actual picked it as my favor sound clip.
I'll keep listening to xrks clips from the comparison thread until I can decide between B80 and 8414. Might order SB65 anyhow as it's not very costly, I'll figure use for it later 🙂
For now, the project is still on hold and I'm just looking for hot deals. I was supposed to move at the beginning of this month, but the previous tenants keep delaying. Right now it seems at earliest I'll move at the beginning of October. On the other hand, scored another major deal. Bought 30pcs of Tangband W4-657sc from a builder who never got around finishing his line array project, 5€ / driver. Might just build a two-way line array after all 😉
For now, the project is still on hold and I'm just looking for hot deals. I was supposed to move at the beginning of this month, but the previous tenants keep delaying. Right now it seems at earliest I'll move at the beginning of October. On the other hand, scored another major deal. Bought 30pcs of Tangband W4-657sc from a builder who never got around finishing his line array project, 5€ / driver. Might just build a two-way line array after all 😉
Don't make the same mistake many have made, just slapping together a bunch of drivers is no guarantee for a good sounding line array 🙂.
I'd recommend making a test box for one driver and see how that does. It's no coincidence that most arrays on here use that one particular driver. 😉
Look around for line arrays using other drivers and see how successful they are. I've only found a few others that were pleased with their end result.
So do look into that driver before doing all that work.
I'd recommend making a test box for one driver and see how that does. It's no coincidence that most arrays on here use that one particular driver. 😉
Look around for line arrays using other drivers and see how successful they are. I've only found a few others that were pleased with their end result.
So do look into that driver before doing all that work.
I have experience with this driver, and I'm currently making a BR box with them for a friend. It has excellent bass for a wideband driver, but problems in high register. It has horrible peak due to cone breakup at 6-8kHz, so it would need a crossover to a tweeter, effectively making it a multi-way design. What immidiately popped into mind would be a Harsch XO with -48db/oct crossover for the woofers, cutting the cone breakup out of the game. Tweeter would likely be happy with -12db/oct when correctly time aligned. Review of the driver from a german magazine attached. My german is too rusty to make anything out of it aside from the graphs, but they seem be happy with it's capabilities as a mid-driver.
In any case, the original thought why I ditched the line array idea was that I couldn't squeeze the drivers close enough for comb filtering not to be a problem at my listening distance. If I don't even try to run a FR line array and settle for about 30Hz to 4-5kHz band on the widebands and fill the rest in with tweeter or a tweeter array, I don't think comb filtering will become such a problem.
In any case, the original thought why I ditched the line array idea was that I couldn't squeeze the drivers close enough for comb filtering not to be a problem at my listening distance. If I don't even try to run a FR line array and settle for about 30Hz to 4-5kHz band on the widebands and fill the rest in with tweeter or a tweeter array, I don't think comb filtering will become such a problem.
Attachments
Last edited:
So you're on it, good to know! Personally I wanted to rid myself of using crossovers to reach my specific goal. And even then it wasn't all that easy. Hope you get to build it and show us your journey 🙂.
Listening distance is simply too close for a line array of any arrangement......the advantages of an array will be lost to the inappropriate setting.
A FAST design is certainly better, but the coherence of the mid tweeter is better presented in the nearfield where it's advantage accells.
A true, well executed two way gives you the best chance of success as the drivers have ample distance to sum, a directivity match is obtainable and the drivers can best utilize their effective pass band.
Consider a RAAL ribbon tweeter in an MTM with 7-8" damped cone midwoofers or a large format compression driver (1.4") in a suitable waveguide mated to a 15" woofer in an EBS alignment.
A FAST design is certainly better, but the coherence of the mid tweeter is better presented in the nearfield where it's advantage accells.
A true, well executed two way gives you the best chance of success as the drivers have ample distance to sum, a directivity match is obtainable and the drivers can best utilize their effective pass band.
Consider a RAAL ribbon tweeter in an MTM with 7-8" damped cone midwoofers or a large format compression driver (1.4") in a suitable waveguide mated to a 15" woofer in an EBS alignment.
That is most likely the case still. Oh well, I trashed a few old drivers, so might as well stock up on new ones 😉 If nothing else, I'll make Karlsonator HT systems out of the W4-657SCs for family.
What bugs me about tweeters is that you can't physically get them inside the required 1/4 wavelenght to blend seamlessly. The closest I've seen so far is the new Genelec speaker in the image below. This is much easier with FR, as the XO can be on much lower frequency.
What I don't like is coaxial drivers. They never sound quite right to me. Haven't heard a coax design I'd want to live with, and I've heard some really high-end stuff.
What bugs me about tweeters is that you can't physically get them inside the required 1/4 wavelenght to blend seamlessly. The closest I've seen so far is the new Genelec speaker in the image below. This is much easier with FR, as the XO can be on much lower frequency.

What I don't like is coaxial drivers. They never sound quite right to me. Haven't heard a coax design I'd want to live with, and I've heard some really high-end stuff.
Corner supported 2 piece cabinets
Line Source: What is the name of this speaker you are showing pictures of? It is very interesting. Is this one of your personal projects, or a build I can read about? I'm interested in where the OP is going with this project, and the various options he is considering. I found his comments about built-ins to be particularly interesting. I have a new room to finish up that is 23'W x 21'L x 9'H. I'm considering building something very efficient, possibly into a corner, or into a wall. One of the 23' walls opens into a much larger space, and would be ideal for infinite baffle.
Any details on the cabinet you show would be greatly appreciated!
Line Source: What is the name of this speaker you are showing pictures of? It is very interesting. Is this one of your personal projects, or a build I can read about? I'm interested in where the OP is going with this project, and the various options he is considering. I found his comments about built-ins to be particularly interesting. I have a new room to finish up that is 23'W x 21'L x 9'H. I'm considering building something very efficient, possibly into a corner, or into a wall. One of the 23' walls opens into a much larger space, and would be ideal for infinite baffle.
Any details on the cabinet you show would be greatly appreciated!
Alright, after another week of digging through the forums like a 19th century coal miner, my plans have become a bit clearer.
If I start from beginning, what are the starting points of the project?
Room and audio source:
1) Room is 3150x3640cm when including cabinets, which will have doors on them. It might be actually beneficial to have the listening position on the side wall instead of before the doors, as it will de-clutter the room greatly. I adjusted the speaker shape to this.
2) The source will still be Raspberry Pi 3 with Hifiberry DAC+ Pro. It can only give out 2 channels of information, so all the crossover functions will happen on analog level before the amplifier. I can't get a steep enough crossover on passive line-level, so I need to do active line-level. For this I will use Sallen-Key topology (unity gain) high quality opamps. I have stocked up on these: NE5532, LM4562, LME49860MA, OPA4134UA, OPA2113PA. I will decide which one is the most suitable by listening tests. The exact value of crossover will be tuned with miniDSP 2x4HD, then build as analog circuit. Currently thinking it depends on how much I can help the phase problems with computer.
3) Whilst the Raspberry Pi 3 and Hifiberry DAC+ Pro are only capable of stereo output, they are very capable of high-quality FIR filters. I haven't had time to read up on these first, I've currently just concentrated on the speaker side. How much can I do with the FIR filtering? Can I for example help with the phase rotation that is inherent to 4th order RL filters?
4) Acoustic treatment is a must. I will build a DIY acoustic panel behind the listening chair. We are also have going to have a nice big fat carper on the floor where me and my wife can sit and play board games with our friends while listening to music.
Speakers
1) The bass driver I will use is the Peerless SLS12 woofer. It will cross over quite low, between 300 to 400Hz. This is because it has an increase in 3rd order distortion higher than 300Hz. It will put some strain to the FR, but nothing that it shouldn't handle. In the graph you can see the distortion at about 100db level. I'm a bit worried about the distortion though. How beneficial would it be to maybe buy 2 more of these and make the cabs taller? I was thinking of a maximum cabinet height of 2m. I currently own a pair. Keep in mind high-tap FIR filtering will also be done, so cabinet design can be modified.
2) The full range will be Scanspeak 10F/8424G00. It seems like a very good driver for this application with excellent off-axis response. I will also build a waveguide for these out of plywood, althought currently still have very little idea on how to do it. This is to time align the acoustic centers of the drivers. On horizontal level, the waveguide will be quite steep to keep the drivers withing 1/4 wavelenght of crossover frequency. Test data of the driver:
http://www.scan-speak.dk/datasheet/reviews/10f8424g00-KlangTon2012-2.pdf
3) The cabinet that currently interests me for that room arrangement is a wedge-shaped very shallow cabinet that you can see in the room picture. It has a small footprint with wide baffle. A closed cabinet recommended by Peerless for the SLS 12 is 90l, but with FIR filtering I believe I can push it a little lower. The enclosure volume for 10F/8424 will be a small triangular TL, "Dagger".
Current concerns:
1) How much can I do with FIR filtering?
2) Is the woofer suitable?
3) What XO should I use between the drivers?
4) How is the waveguide constructed for the full range?
If I start from beginning, what are the starting points of the project?
Room and audio source:
1) Room is 3150x3640cm when including cabinets, which will have doors on them. It might be actually beneficial to have the listening position on the side wall instead of before the doors, as it will de-clutter the room greatly. I adjusted the speaker shape to this.
2) The source will still be Raspberry Pi 3 with Hifiberry DAC+ Pro. It can only give out 2 channels of information, so all the crossover functions will happen on analog level before the amplifier. I can't get a steep enough crossover on passive line-level, so I need to do active line-level. For this I will use Sallen-Key topology (unity gain) high quality opamps. I have stocked up on these: NE5532, LM4562, LME49860MA, OPA4134UA, OPA2113PA. I will decide which one is the most suitable by listening tests. The exact value of crossover will be tuned with miniDSP 2x4HD, then build as analog circuit. Currently thinking it depends on how much I can help the phase problems with computer.
3) Whilst the Raspberry Pi 3 and Hifiberry DAC+ Pro are only capable of stereo output, they are very capable of high-quality FIR filters. I haven't had time to read up on these first, I've currently just concentrated on the speaker side. How much can I do with the FIR filtering? Can I for example help with the phase rotation that is inherent to 4th order RL filters?
4) Acoustic treatment is a must. I will build a DIY acoustic panel behind the listening chair. We are also have going to have a nice big fat carper on the floor where me and my wife can sit and play board games with our friends while listening to music.
Speakers
1) The bass driver I will use is the Peerless SLS12 woofer. It will cross over quite low, between 300 to 400Hz. This is because it has an increase in 3rd order distortion higher than 300Hz. It will put some strain to the FR, but nothing that it shouldn't handle. In the graph you can see the distortion at about 100db level. I'm a bit worried about the distortion though. How beneficial would it be to maybe buy 2 more of these and make the cabs taller? I was thinking of a maximum cabinet height of 2m. I currently own a pair. Keep in mind high-tap FIR filtering will also be done, so cabinet design can be modified.
2) The full range will be Scanspeak 10F/8424G00. It seems like a very good driver for this application with excellent off-axis response. I will also build a waveguide for these out of plywood, althought currently still have very little idea on how to do it. This is to time align the acoustic centers of the drivers. On horizontal level, the waveguide will be quite steep to keep the drivers withing 1/4 wavelenght of crossover frequency. Test data of the driver:
http://www.scan-speak.dk/datasheet/reviews/10f8424g00-KlangTon2012-2.pdf
3) The cabinet that currently interests me for that room arrangement is a wedge-shaped very shallow cabinet that you can see in the room picture. It has a small footprint with wide baffle. A closed cabinet recommended by Peerless for the SLS 12 is 90l, but with FIR filtering I believe I can push it a little lower. The enclosure volume for 10F/8424 will be a small triangular TL, "Dagger".
An externally hosted image should be here but it was not working when we last tested it.
Current concerns:
1) How much can I do with FIR filtering?
2) Is the woofer suitable?
3) What XO should I use between the drivers?
4) How is the waveguide constructed for the full range?
Attachments
Last edited:
Why do you need anything fancier than a sealed box for the full-range if you are having an XO freq much higher than driver Fs? And why do need a physical XO if you have a miniDSP?
I want to use the dagger sealed TL to spread the internal resonances as evenly as possible. It basically is a sealed box, just different shape. MiniDSP is excellent, but the limitation is that it can do a maximum of 2048 tap for FIR filter. On Raspberry Pi, I can do something around 64k taps. Plus, I want to keep the miniDSP as a general experimentation tool without committing it to one project.Why do you need anything fancier than a sealed box for the full-range if you are having an XO freq much higher than driver Fs? And why do need a physical XO if you have a miniDSP?
Well, you can also do a cross-over on RPi, but then you will need a usb sound card with multiple outputs. This way you can easily tune it later.
I know, but I already have the Hifiberry DAC+ at hand 🙂 I'll give it a go first, and if I can't get satisfactory results with that I'll switch to other means. Currently it'll mean splitting and filtering the signal on analog domain.Well, you can also do a cross-over on RPi, but then you will need a usb sound card with multiple outputs. This way you can easily tune it later.

Just bought a pair of these little gems from a local DIY builder. They are currently residing in my pat. pending Secular Harmonics Optimized Enhancer Box, or a SHOEBox enclosure filled with polyfill 😎
I'm actually surprised that they so much bass. Crossing at 300Hz should be no problem with these. At the same breath, I have to admit these are overall the best full range driver I've heard. The Philips 9710 is better with human voice, but these beat them if you listen to a variety of audio outside simple vocal music.
They are indeed 🙂 Had a hard time selecting between 8414, 8424 and B80, but it ultimately tipped my boat that I could obtain these second-hand locally.
Small update: I'm moving to my new apartment at the start of November, and then this project can truly begin.
The current plan is a trapezoidal enclosure about 1m tall where the bass driver (Peerless SLS-12) will reside as high as dimensions allow. The enclosure itself will be about 50cm wide at the bottom and 30cm wide at the top point, along with 30cm bot depth to 20cm top depth. So each side will be slanted 10cm / 1m. The back baffle will be straight. It will be constructed of 24mm BB plywood with good bracing. It will be stuffed full of rockwool. Internal volume will be about 77l, which is a bit small for the SLS-12, but nothing that I shouldn't be able to fix with FIR filtering.
The top enclosure will be a 30x20x20cm (width/height/depth) and constructed with BBC/Harbeth -style thin wall cabinet design. It will have an elliptical waveguide for the full range (Scanspeak 10F/8424) which main purpose is to match the acoustic centers with the Peerless driver. It will be constructed of 4mm BB plywood with 8mm bitumen on insides. This cabinet will not have a bottom at all; instead it will be attached to the top of the bass cabinet with elastic silicone based glue which will effectively absorb vibrations and form an airtight seal between the boxes.
The main idea is to brace the bottom cabinet so much that the resonances in the cabinet move so high that the bass driver won't excite them (over 400-500Hz area). Similarly, the idea of the top cabinet is to drop the resonances so low that the FR won't excite them (to about 100-200Hz). The elastic glue layer will separate the cabinets from each other and prevent the vibrations from travelling between them. The baffle edges will be rounded to minimize diffractions.
The crossover will be active LR4 filter made with LME49860 OP amps. It is the best sounding op amp that I've tried on my DAC (Audioquest Prodigy Cube, allows the end user to switch op amps easily). The exact crossover frequency will be determined with measurements done with MiniDSP 2x4 HD along with UMIK-1. The amplifier will most likely be the Sanwu TDA7498, 1 per speaker. Room correction and other DSP will be done as FIR filtering with Raspberry Pi 3 fitted with Hifiberry DAC+ Pro.
For room treatment I will construct a DIY sound absorber behind the listening chair. The speakers will have their back baffle right at the back wall, same with the listening chair. The floor will have a heavy carpet between the speakers and the chair, but no other treatment is planned. If it has bad ringing, I'll build bass traps on the corners next to windows. It's a rental apartment, so not worth heavily investing on 🙂
The current plan is a trapezoidal enclosure about 1m tall where the bass driver (Peerless SLS-12) will reside as high as dimensions allow. The enclosure itself will be about 50cm wide at the bottom and 30cm wide at the top point, along with 30cm bot depth to 20cm top depth. So each side will be slanted 10cm / 1m. The back baffle will be straight. It will be constructed of 24mm BB plywood with good bracing. It will be stuffed full of rockwool. Internal volume will be about 77l, which is a bit small for the SLS-12, but nothing that I shouldn't be able to fix with FIR filtering.
The top enclosure will be a 30x20x20cm (width/height/depth) and constructed with BBC/Harbeth -style thin wall cabinet design. It will have an elliptical waveguide for the full range (Scanspeak 10F/8424) which main purpose is to match the acoustic centers with the Peerless driver. It will be constructed of 4mm BB plywood with 8mm bitumen on insides. This cabinet will not have a bottom at all; instead it will be attached to the top of the bass cabinet with elastic silicone based glue which will effectively absorb vibrations and form an airtight seal between the boxes.
The main idea is to brace the bottom cabinet so much that the resonances in the cabinet move so high that the bass driver won't excite them (over 400-500Hz area). Similarly, the idea of the top cabinet is to drop the resonances so low that the FR won't excite them (to about 100-200Hz). The elastic glue layer will separate the cabinets from each other and prevent the vibrations from travelling between them. The baffle edges will be rounded to minimize diffractions.
The crossover will be active LR4 filter made with LME49860 OP amps. It is the best sounding op amp that I've tried on my DAC (Audioquest Prodigy Cube, allows the end user to switch op amps easily). The exact crossover frequency will be determined with measurements done with MiniDSP 2x4 HD along with UMIK-1. The amplifier will most likely be the Sanwu TDA7498, 1 per speaker. Room correction and other DSP will be done as FIR filtering with Raspberry Pi 3 fitted with Hifiberry DAC+ Pro.
For room treatment I will construct a DIY sound absorber behind the listening chair. The speakers will have their back baffle right at the back wall, same with the listening chair. The floor will have a heavy carpet between the speakers and the chair, but no other treatment is planned. If it has bad ringing, I'll build bass traps on the corners next to windows. It's a rental apartment, so not worth heavily investing on 🙂
Last edited:
- Status
- Not open for further replies.
- Home
- Loudspeakers
- Full Range
- Hi-End Full-Range project