(Beginner) An attempt to improve my speakers, help appreciated.

Vitus · 2020-02-13 7:43 pm

I want to improve the sound quality out of my speakers, yes - I know, a very generalized term, but then again I'm a complete beginner.

To start off, here's a picture of my speakers:
http://www.bilddump.se/bilder/20200213200134-46.195.138.78.jpg

Here's a drawing of the cabinet:
WeTransfer

The manufacturer will not help me in telling each drivers individual specifications, they're way too secretive, hence my doubt of the speaker not performing as per specifications.

Word has it this speaker performs very well in its price point, it's pretty much impossible to find speakers sounding this good for any lesser amount, and I can honestly agree, as long as you don't build the speakers from scratch yourself, nothing really beats it at this price level, which is $300.

Manufacturers and my take on the specifications:

Cabinet consists of 18mm MDF all around, and 15mm bracing internally.
MTM enclosure is 12,97 liters, woofer enclosure is 58,76 liters (with ports plugged from the inside)
4x 80mm 150mm long vented ports in the rear

1" Super Audio (lol) fabric tweeter, all plastic, neodymium magnet, 6 ohm
2 x 5.25" midrange, coated paper, shielded, steel basket, 80mm magnet, 4 ohm each
2 x 8" woofers, coated kevlar, full cast aluminium basket, 100mm magnet, 8 ohm each

Frequency Response 24-32000 Hz (- 3 dB)
Crossover Frequency 240/2500 Hz
Impedance 4-8 ohms
Sensitivity 94 db
RMS Power 190 W
Max Power 280 W

There is a terminal where you can change the midrange, -2 dB / 0 dB / +2 dB / +4 dB

Size (HxWxD) 1100 x 238 x 395mm
Weight 28 kg per tower

I don't know how to measure the drivers to calculate the RMS wattage of each individual driver, the impedance I could easily measure, but the sticker on each driver should do just fine.

There is no ampere figure on the drivers, the wattage is therefor completely unknown.

Word has it the tweeter easily blows if you turn the speaker up too loud, common issue.

To my plan:

I plan to focus on the weaknesses of these speakers, I find the cabinet to be of my taste, and it's decently sized for my room, fits perfectly with all of my other speakers too.

Obviously the tweeter needs to be changed, it's probably very weak, either that or the crossover is just poorly designed, I would guess a mixture of the too, but more the tweeter itself being the issue.

Also the midrange is made out of cheap coated paper, by the looks of it it's a very cheaply made driver, only 80mm magnet, and the overall quality inspection and feel of the driver is very cheap, there is no quality inspection during the fabrication process it feels like.

The woofers however, they're not that critical to replace, they're made of kevlar (supposedly) and it looks very much like it at the back of the cone, however the magnet is only 100mm, most midrange drivers have this size and that makes me question these woofers.

I also plan to seal these cabinets with foam plugs internally, so that I reduce the cabinet volume as much as possible (the new midrange speakers I found does require a bit less volume) so nothing wrong in that.

While doing these improvements to the speakers, I also cover all the inside walls with dampening material (Zachry Q12) since I seal the cabinets fully, I don't fill them fully. just cover up the walls.

Is there a better way to dampen the cabinets, again - please look at the cabinet drawing and suggest improvements.

First however, let's focus on tweeter and the midrange.

I know I will need to design a new crossover for these improvements and changes, I've fiddled around somewhat in XSim, planning on getting a UMIK-1 to measure the current speaker as a whole and each individual driver.

I found this tweeter to be a perfect match, physically, and it seems really nice by looking at the specifications as well, people have reported this driver to be perfect with a 2k crossover, and it's made out of fabric, which my atmos upfiring speakers are too, so that's good I guess.

SB Acoustics SB26STAC-C000-4 - 1.1" Dome Tweeter

As for midrange, it's quite tough finding speakers with the very strange frame/basket shape that these speakers have, it's a square shape with rounded corners.

Anyway, Aurum Cantus reportedly seems to offer pretty nice midrange speakers, I have no clue what my current midrange are rated at in terms of power, or pretty much any specification really, other than the magnet being 80mm, and the 4 ohm rating.

Aurum Cantus offer many variations, but I thought that the AC-130MK2 looked pretty nice, it's very even in the frequency response, and the 2k crossover fits perfectly with the SB26STAC-C000-4 from what I can tell as a beginner.

Aurum Cantus AC-130MK2 - 5.5" Woofer

Right, my questions:

1) Can I measure the RMS wattage closely to the manufacturer specifications (if I had them) by measuring with a True RMS volt/ampmeter?

2) If yes on the above, tell me step by step how to connect everything, because I'm unsure if the drivers can be connected directly to a PC for example, you would need some sort of amplifier I guess, but is a normal 2-channel amplifier (like a Yamaha) overkill, is it easier to have a USB soundcard at hand?

3) What kind of test tone would you output, is there a general rule of thumb, 1k? 2k? What kind of SPL, just about any would do? Speakers vary quite a bit in voltage and amps (hence also watts) that's why I ask.

4) Does my suggested tweeter/midrange combo look like they would match and yield great performance if the crossover is designed correctly?

5) How would I go about matching the tweeter/midrange to my already existing woofers, do I need to measure them and create/generate a FRD/ZMA file somehow, with the UMIK-1 maybe?

I've read some stuff on the forums, but I have a hard time connecting everything together, I appreciate all the help you can get me, we all start somewhere, this is where I start.

Explain to me in a greater noob fashion that you think you need to, because eventhough I might of touched on some delicate topics, I'm really just a beginner.

And oh yeah, I plan to measure the speakers assembled together (once upgraded) in the room where they'll be used, so that I can measure the room strength/weaknesses and implement and compensate for those in the crossover.

I have no issues at all having to re-design the crossover if moving the speakers to a different room.

Thanks again!
Vitus

Feel free to toss my thread around, anywhere it can go for the right answers and I'm more than glad.

Maybe some real gurus could hop in and have a go at it, I need all the help I can get.

I'm used to be an electrician as a profession, so I'm not too shabby with electrical terms and how all that plays out.

I would give you something if I could, thank you.

jamesblonde · 2020-02-13 8:32 pm

I wouldn't recommend tampering with these. Best to sell them and build a new pair from scratch if you want to get your hands dirty.

dazcwl · 2020-02-13 8:41 pm

A few people have replied since writing my main post. I would agree with them, probably not worth messing with an “affordable” commercial speaker.

Having said that;

Building a new crossover for those speakers may be worth a try even just for the learning experience, I cannot recommend changing drivers. You would need to modify the enclosure, changing the midrange driver will almost certainly require a different enclosure volume. At most I would consider changing the tweeter.

I do not want to stop you from trying but when I first started in DIY speakers I considered modifying commercial speakers before building a new design. If you are going to invest in the measurement gear and put the time in you may as well just start with decent drivers and build your own enclosure?

I would highly recommend getting the UMIK-1 as the frequency and impedance response is unknown.
You will also need something like DATS to take an impedance measurement, your post states you can measure impedance not sure if this is with a multimeter?

I would try to get a measurement of the speakers as they are before deciding on any new drivers.

Sorry I cant help you with “measuring RMS wattage”, I would instead measure for harmonic distortion on each driver at increasing levels and go from there.

Also I would Not try to account for the room frequency response within a passive crossover. Not sure if that is what you meant or not?

Sonce · 2020-02-14 12:15 am

Vitus said:
I want to improve the sound quality out of my speakers...
Obviously the tweeter needs to be changed, it's probably very weak, either that or the crossover is just poorly designed, I would guess a mixture of the too,

Start with the obvious - if the tweeter is too weak (not loud enough), try to modify the crossover (tweeter's high-pass filter). Draw the crossover schematic and post it here. If there is a serial resistor to the tweeter, change it with a smaller value. If there is a resistor parallel to the tweeter, increase it's value. If it doesn't help, change the serial capacitor with a slightly higher value, say if it is 3.3 microF, than try 4.7.
DO NOT try to measure the speaker RMS wattage, there is no need for that! It is certainly 100 Wrms at least, so you can safely use a 100 W amplifier. You don't need anything more powerful than that.
Cheap look of the drivers doesn't "guarantee" low sound quality, but skimpy woofer magnet may produce boomy bass. Put more damping material in the cabinet and block vents with foam - it should tighten the bass.

If all these doesn't help, only then you can try to replace drivers with the better ones. For that you need UMIK-1, REW or ARTA measurement software and sim software (VituixCAD or similar). Prepare for many hours of learning... and fun!

nigelwright7557 · 2020-02-14 12:26 am

jamesblonde said:
I wouldn't recommend tampering with these. Best to sell them and build a new pair from scratch if you want to get your hands dirty.

I agree.

You need to sit down and decide what you want fro ma set of speakers and what you didn't like about your ones.

A simple approach I use is using full range drivers line Fane 12-250TC which goes from low to 17KHz. Some love it and some hate it but I have had good results with a pair. With full rangers you get rid of the pain and cost of crossovers.
If you want a bit more bottom end and size isnt a problem then Fane 15-TC300 is an option.

Moondog55 · 2020-02-14 5:28 am

I will assume that is $300- USD
If so they do actually look like good value.
I wouldn't bother yet with any driver substitutes, instead i would ask how long you have owned them, how hard you have driven them and what sort of music you listen to.
Secondly have you taken the drivers out and had a look in the boxes interior?
Padding or no padding? If padding what is it?
If no padding add some decent fibreglass, wool wadding or good synthetic fibre; especially to the mid-range
Thirdly some cheap speakers can sound a lot better by adding in a good pair of subwoofers and taking all of the signal below 80 / 120 Hz
EDIT
We need pictures of the midrange drivers from the rear Ditto the tweeter I don't think they are sealed back drivers because of the separate driver compartment but best to be sure

system7 · 2020-02-14 6:03 am

As Sonce says, you really must work out what the filter is and how the speakers are wired.

Since this is a WWMTM, you need to know if the drivers are wired series or parallel.

I'd guess the mids are wired series at 4 ohms each, to make 8 ohms. The woofers, I can't guess.

This is a glorified 8" plus 5" plus 1" 3 way, like Troels Gravesen's fine SEAS 3 way Classic:
SEAS-3-Way-Classic

The theory:

646927d1511229072-suppose-shallow-vs-steep-argument-duelund-2-83-jpg

The practice:

424984d1403466206-troels-gravesen-time-aligned-3-published-3-classic-jpg

Before you go running around looking for new drivers, the greatest improvement will come from a good crossover. It is easy to convert a given circuit for multiple drivers and that should be your approach.

Job 1: Trace the existing crossover and find out if the basses are series or parallel wired. Note the polarity of the drivers too. 3 ways often mix it up.

Vitus · 2020-02-14 2:19 pm

jamesblonde said:
I wouldn't recommend tampering with these. Best to sell them and build a new pair from scratch if you want to get your hands dirty.

Right along, lesson taken, thanks for your input, I appreciate it (really, I mean it).

Althought I'm still interested to find out to which extent I could possibly improve the audio coming from these, so I'll give it a shot and see how far I can come.

Thanks again.

Temaad said:
It is a very difficult job to shoe horn different speakers into a established cabinet. Yes changing the tweeter is the easiest. The cabinet size & shape has been developed to suit the other speakers & most likely will not suit another speaker from a different supplier. One way to upgrade the sound is to ditch the X-Over & go for tri-amping with an electronic X-over. It will mean buying a bass amp (around $100, class D TPA3255) & tweeter amp (around $35, Class D TPA3118)
However if this does not suit, then you need to buy or build new speakers.

I understand, I would very much still like to give it a shot, I'm not against trying to modify the cabinet, we'll see what it takes, ultimately.

As for the X-change, I could very well remove it if designing a completely new crossover, I mean why bother with a X-change when you can design a crossover for my room and my room alone?

Tri-amping is out of the question I think, I need a passive crossover to go along with my STR-DN1080 130W/channel home cinema receiver, I don't have the space for any more amplifiers.

Yes, that's right, I want to improve these speakers for movies first hand, second comes music, I'll just chime in on FLAC/DSD at times.

dazcwl said:
Building a new crossover for those speakers may be worth a try even just for the learning experience, I cannot recommend changing drivers. You would need to modify the enclosure, changing the midrange driver will almost certainly require a different enclosure volume. At most I would consider changing the tweeter.

I do not want to stop you from trying but when I first started in DIY speakers I considered modifying commercial speakers before building a new design. If you are going to invest in the measurement gear and put the time in you may as well just start with decent drivers and build your own enclosure?

I would highly recommend getting the UMIK-1 as the frequency and impedance response is unknown.
You will also need something like DATS to take an impedance measurement, your post states you can measure impedance not sure if this is with a multimeter?

I would try to get a measurement of the speakers as they are before deciding on any new drivers.

Sorry I cant help you with “measuring RMS wattage”, I would instead measure for harmonic distortion on each driver at increasing levels and go from there.

Also I would not try to account for the room frequency response within a passive crossover. Not sure if that is what you meant or not?

Wow, long reply, I'll try to cover it all.

Yes, the tweeter is certainly the best change here, however I am still very curious about finding out of I can improve the midrange too, I see modifying the enclosure/enclosures as a part of the challenge to improve the audio if needs be.

I don't want to build my own enclosures for now, maybe next time.

Yes, when I talked about impedance measuring I refered to using a multimeter, maybe not the best way to do it, better just refer to the nominal impedence rated by the manufacturer.

UMIK-1 I will certainly buy, DATS is a nice recommendation, I see that DATS V3 just got released, I'll look into it, thank you!

By accounting for the room frequency response in a passive crossover I would get the best possible audio out of the speakers in the very room I decide to measure in, no?

Otherwise I would have to use more software based EQ and that's worse than doing the EQ with hardware in the crossover, yes/no?

Sonce said:
Start with the obvious - if the tweeter is too weak (not loud enough), try to modify the crossover (tweeter's high-pass filter). Draw the crossover schematic and post it here. If there is a serial resistor to the tweeter, change it with a smaller value. If there is a resistor parallel to the tweeter, increase it's value. If it doesn't help, change the serial capacitor with a slightly higher value, say if it is 3.3 microF, than try 4.7.
DO NOT try to measure the speaker RMS wattage, there is no need for that! It is certainly 100 Wrms at least, so you can safely use a 100 W amplifier. You don't need anything more powerful than that.
Cheap look of the drivers doesn't "guarantee" low sound quality, but skimpy woofer magnet may produce boomy bass. Put more damping material in the cabinet and block vents with foam - it should tighten the bass.

If all these doesn't help, only then you can try to replace drivers with the better ones. For that you need UMIK-1, REW or ARTA measurement software and sim software (VituixCAD or similar). Prepare for many hours of learning... and fun!

Thanks, very useful response!

I wouldn't really modify the current crossover to accomodate for an improved audio with the current tweeter, because the tweeter itself is probably very poor, as even stated by the reseller (it often blows with high volume).

I would much rather change the tweeter, and then adjust in the crossover for that, no point spending time adjusting for something that's not worth adjusting, IMHO.

Draw the crossover schematic and post it here

Yes, this is very tricky, I took one of the speakers apart, and I tried to follow the circuits by continuity, but it's rather hopeless as the PCB's are multi-layered and you don't really know if some components are wired in parallel to ruin the measurements.

I certainly tried to measure, but it's hopeless, it's pretty much impossible to trace, you would need to un-solder every component and open the PCB to really find out how it's wired, you can't see any wiring paths on the PCB.

However, I took note of the components.

There are two PCB's, one for HF, one for LF.

The HF PCB has the following components:

1x CYC CD71 100V 100uF
1x CYC CD71 100V 5.6uF
1x CYC MKT J250V 4.7uF
1x Solid-Core Inductor 1.5mH 0.45 Ohm (Midrange, manufacturer says)
1x Air-Core Inductor 0.25mH 0.95 Ohm (Tweeter, manufacturer says)

The LF PCB has the following components:

1x CYC NP 100V 6,8uF
1x Solid-Core Inductor 3mH 0.48 Ohm (Woofers, manufacturer says)
1x 5W4R7J Resistor

The X-change terminal has the following components:

1x 5W2R2J Resistor
1x 5W1RJ Resistor

When not using the X-change, i.e. unplugged, both of these resistors are utilized and affects the 4-20kHz range (i.e the tweeter most likely, haven't checked physically in the crossover).

For a +2dB boost in the X-change, the 5W2R2J is bypassed.
For a 0dB difference, the 5W1RJ is bypassed.

Quick sketch:

[IMGDEAD]http://www.bilddump.se/bilder/20200214154812-46.195.138.78.PNG[/IMGDEAD]

How would I go about reverse engineering the crossover?

It's pretty much impossible to figure out how it's wired?

I'm using the Sony STR-DN1080 home cinema receiver, as I will mostly dedicate these speakers towards movies, occasionally I would put on some FLAC/DSD.

It's rated 130W per channel, 6-16 Ohms.

I more dampening I understand is a pretty simple fix, the speaker is originally based a vented design, thus the lack of dampening material on certain inside walls, more dampening is required when using a sealed design I take it.

Below you'll find what's dampened and not, currently - sketch is from the side of the speaker.

[IMGDEAD]http://www.bilddump.se/bilder/20200214145219-46.195.138.78.PNG[/IMGDEAD]

Green is where there is dampening, red is where there's not.

+ Side Walls in GREEN means there is dampening on the side walls.
+ Side Walls in RED means there is no dampening on the side walls.

Plugging the ports definitely tightened up the bass, I much prefer this sound, and leave the lower bass handling to my active subwoofers which also are sealed, hence why I am looking to redesign these speakers for a sealed operation.

UMIK-1 I will get, thanks for the recommendation.

I tried VituixCAD, I much prefer Xsim due to the larger crossover schematic window, I'll look into REW and ARTA, and I will most likely purchase DATS V3.

Thank you!

nigelwright7557 said:
I agree.

You need to sit down and decide what you want fro ma set of speakers and what you didn't like about your ones.

A simple approach I use is using full range drivers line Fane 12-250TC which goes from low to 17KHz. Some love it and some hate it but I have had good results with a pair. With full rangers you get rid of the pain and cost of crossovers.
If you want a bit more bottom end and size isnt a problem then Fane 15-TC300 is an option.

I will not redesign these for full range drivers, I like the approach of a 3-way design, hence why I posted this thread in Multi-Way.

I also like the idea of fiddling with the sound with a crossover.

Thanks for your input!

Moondog55 said:
I will assume that is $300- USD
If so they do actually look like good value.
I wouldn't bother yet with any driver substitutes, instead i would ask how long you have owned them, how hard you have driven them and what sort of music you listen to.
Secondly have you taken the drivers out and had a look in the boxes interior?
Padding or no padding? If padding what is it?
If no padding add some decent fibreglass, wool wadding or good synthetic fibre; especially to the mid-range
Thirdly some cheap speakers can sound a lot better by adding in a good pair of subwoofers and taking all of the signal below 80 / 120 Hz
EDIT
We need pictures of the midrange drivers from the rear Ditto the tweeter I don't think they are sealed back drivers because of the separate driver compartment but best to be sure

Yes, $300USD, good value indeed.

For padding/dampening, look at the below picture, it's a side view of the cabinet:

[IMGDEAD]http://www.bilddump.se/bilder/20200214145219-46.195.138.78.PNG[/IMGDEAD]

Green is where there is dampening, red is where there's not.

+ Side Walls in GREEN means there is dampening on the side walls.
+ Side Walls in RED means there is no dampening on the side walls.

The current dampening material is synthetic I believe.

I have dual 10" sealed active subwoofers from 30-180hz.

Crossover possible from 40-180hz.

The midrange drivers are definitely not sealed, they're open-back drivers, so are the woofers, only the tweeter is sealed back.

Midrange drivers are in a separate MTM enclosure, just as you mentioned.

Could I improve the dampening here and add dampening to top and bottom of the MTM compartment?

Why did the manufacturer leave top and bottom naked?

system7 said:
As Sonce says, you really must work out what the filter is and how the speakers are wired.

Since this is a WWMTM, you need to know if the drivers are wired series or parallel.

I'd guess the mids are wired series at 4 ohms each, to make 8 ohms. The woofers, I can't guess.

Before you go running around looking for new drivers, the greatest improvement will come from a good crossover. It is easy to convert a given circuit for multiple drivers and that should be your approach.

Job 1: Trace the existing crossover and find out if the basses are series or parallel wired. Note the polarity of the drivers too. 3 ways often mix it up.

I took my lesson, did my due diligence, took one of the speakers apart, below is what I found out:

WIRING:

Tweeter - Parallel (6 Ohm driver)
Midrange - Series (4 Ohm drivers)
Woofer - Parallel (8 Ohm drivers)

This leaves the tweeter at 6 Ohm, not accounting for the crossover (see the crossover components in the FOURTH quote in this post).

Midrange is wired in series to 8 Ohm.

Woofers are in parallel to eachother and makes for a 4 Ohm load, there is also a resistor here (5W4R7J) however I'm unsure what to make of that as the crossover design and schematics are unknown, again see the FOURTH quote in this post).

Could you ellaborate on how to reverse engineer and determine how the crossover circuitry looks like?

I tried measuring continuity but it's impossible since you don't know which components might be wired in parallel to others.

There are two PCB's, HF/LF, both have multi-layered PCB design so the circuit paths are impossible to see, you would have to open up the PCB (impossible) to see the wiring.

About the polarity of the drivers, could you expain some more?

I'm much interested in finding out, but since the crossover is hard to trace, I would need some further advice on this point, thank you.

Further on I have some questions about the internal wiring to the speakers, this is what I could see inside the cabinet:

Tweeter: 20AWG, non-golden cable shoes, ordinary copper wiring.

Midrange: 18AWG, non-golden cable shoes, ordinary copper wiring.

Woofers: 18AWG, non-golden cable shoes, ordinary copper wiring.

Is there an audio quality increase from using a different gauge wiring for the drivers, if so, which and why?

Would there be a benefit of using Oxygen-Free-Copper (OFC) cables or just better HiFi cables/wiring internally?

Gold-plated cable shoes, or better off just soldering directly with 9,5% silver solder?

All speaker terminals are non-gold plated.

Thanks for your reply!

jReave · 2020-02-14 5:54 pm

If you're willing to learn, what you are proposing looks doable to me. This thread should help: So you want to design your own speakers from scratch

Probably most important is to learn how to properly take frequency and impedance measurements because if you can do this right, it's pretty easy for others to use those files and help you design the xo's.

For FR measurements - White paper: Accurate in-room frequency response measurements

How to find the relative acoustic centers is vital as well.

To start with, you can actually use an AV receiver's measurement mic with free REW software if your computer's soundcard allows it (mostly yes for older desktops but no for laptops - laptops usually require a separate soundcard). A better mic can be purchased later if desired.

For impedance DATS will certainly work but you can also use a little homemade jig with REW to get the job done as well.

Given all of that, I would indeed practice on your existing speakers and perhaps correct any glaring xo problems to start off with. You may actually be able to improve those quite a bit. Or maybe not. Regardless, you must have your woofers' FR and impedance measurements in the cabinets in order to work out any simulations with new drivers.

But you've got 13L for your mids to work in and that should be fine for 2 of the AC 130-MKII's. Trouble is that that driver's FR starts falling fairly early at about 1600Hz which doesn't leave a lot of room to play with at the xo point. I might instead choose the AC-130F1 which has a more extended FR and should be easier to work with. It's also used in the well known Continuum speakers by Jeff Bagby.

Potentially, your biggest problem is matching the existing woofer's sensitivity with the new MTM you want to build. Likely though that if your current speaker's sensitivity is below 90dB, that shouldn't be a problem. More than likely, you'll instead have to pad 2 of those mids in parallel down a fair amount to match the woofers instead.

Personally for this kind of learning project, I might choose less expensive drivers but I'm assuming you are looking at the AC's because they will fit your existing cutouts. There are ways to possibly get around that but if your budget allows for the AC's you are looking at, well then by all means proceed. With the caveat of course, it's always best to simulate the whole speaker response before buying any new drivers.

jReave · 2020-02-14 6:11 pm

Also - about your existing tweeter.

The fact that it's known to blow could be inherent to the tweeter but it may also depend on the xo filter, ie. on the implementation. If the xo point is too low and/or if the xo slope isn't steep enough the same problems can happen. So it may be possible to make improvements with a re-designed xo. Or not. Could just be a cheap tweeter in need of replacement in the end.

lowmass · 2020-02-14 6:16 pm

1-get a measurement system and measure each speaker with stock crossover
2- then measure each drivers response in cabinet seperately without crossover connected.
3- now u have the info u need to even start thinking about what u might do

until u do this u really are shooting in the dark

my guess is that with a decent cheap tweeter and a complete crossover re design guided by the measurements, you likely could end up with a better sounding speaker thats about 3 db less sensitive than stock and what u learn following these simple rules will pay big dividends in the future. Keep it simple and you will not fall into chasing your tail

you would be surprised how good even cheap drivers can sound when all is balanced well and the experience of doing that is worth a bunch

system7 · 2020-02-14 7:36 pm

I've been reverse-engineering computers and loudspeakers for years.

I enjoy it.

There are two PCB's, one for HF, one for LF.

The HF PCB has the following components:

1x CYC CD71 100V 100uF
1x CYC CD71 100V 5.6uF
1x CYC MKT J250V 4.7uF
1x Solid-Core Inductor 1.5mH 0.45 Ohm (Midrange, manufacturer says)
1x Air-Core Inductor 0.25mH 0.95 Ohm (Tweeter, manufacturer says)

The LF PCB has the following components:

1x CYC NP 100V 6,8uF
1x Solid-Core Inductor 3mH 0.48 Ohm (Woofers, manufacturer says)
1x 5W4R7J Resistor

The X-change terminal has the following components:

1x 5W2R2J Resistor
1x 5W1RJ Resistor

This is a WWMTM.
We also know we have two 4 ohm 5" mids in series. Making 8 ohms.
And two 8 ohm 8" woofers in parallel, making 4 ohms.
A 6 ohm 1" tweeter with an ohm or two of attenuation.

It's not going to be a million miles from this off-the-shelf crossover, except missing the 0.68mH mid rolloff:

HW 3/130 NG - 8 Ohm | Visaton

I think I'll just Boxsim this one with some similar drivers: Software | Visaton

Bass filter must be 3mH, with a 4.7R and 6.8uF shunt.
Tweeter must be 4.7uF and 0.25mH and 2 ohms resistor in front.
Mid filter must be 100uF with a 1.5mH and 5.6uF shunt.

We can do this!

Moondog55 · 2020-02-14 8:08 pm

First off try filling the midrange compartment with fibre, any fibre will do for the first experiment, even an old woollen jumper
Less damping means lower manufacturing cost. Most talk of and about wiring tends to snake oil after a while, what the manufacturer uses is probably fine. Wiring length inside a cabinet is short
XO values are probably correct and 100V capacitors mean that reasonable care has been taken Re quality but pictures of the actual XO would help. If these are for movies only then L&R quality on voice is less important as most dialogue comes from the centre speaker, what are the details of the centre speaker?

Vitus · 2020-02-14 8:44 pm

jReave said:
If you're willing to learn, what you are proposing looks doable to me. This thread should help: So you want to design your own speakers from scratch

Probably most important is to learn how to properly take frequency and impedance measurements because if you can do this right, it's pretty easy for others to use those files and help you design the xo's.

For FR measurements - White paper: Accurate in-room frequency response measurements

How to find the relative acoustic centers is vital as well.

To start with, you can actually use an AV receiver's measurement mic with free REW software if your computer's soundcard allows it (mostly yes for older desktops but no for laptops - laptops usually require a separate soundcard). A better mic can be purchased later if desired.

For impedance DATS will certainly work but you can also use a little homemade jig with REW to get the job done as well.

Given all of that, I would indeed practice on your existing speakers and perhaps correct any glaring xo problems to start off with. You may actually be able to improve those quite a bit. Or maybe not. Regardless, you must have your woofers' FR and impedance measurements in the cabinets in order to work out any simulations with new drivers.

But you've got 13L for your mids to work in and that should be fine for 2 of the AC 130-MKII's. Trouble is that that driver's FR starts falling fairly early at about 1600Hz which doesn't leave a lot of room to play with at the xo point. I might instead choose the AC-130F1 which has a more extended FR and should be easier to work with. It's also used in the well known Continuum speakers by Jeff Bagby.

Potentially, your biggest problem is matching the existing woofer's sensitivity with the new MTM you want to build. Likely though that if your current speaker's sensitivity is below 90dB, that shouldn't be a problem. More than likely, you'll instead have to pad 2 of those mids in parallel down a fair amount to match the woofers instead.

Personally for this kind of learning project, I might choose less expensive drivers but I'm assuming you are looking at the AC's because they will fit your existing cutouts. There are ways to possibly get around that but if your budget allows for the AC's you are looking at, well then by all means proceed. With the caveat of course, it's always best to simulate the whole speaker response before buying any new drivers.

I love you!

Very good documents, I'll certainly take a read and come back with more knowledge on the subject, thank you.

I won't spend time building custom tools and/or microphones, I'd much rather spend the time on correcting the speakers and putting the time where it matters the most, I can take a bit of money spending on this project, not the biggest concern.

My focus is on DATS V3 and UMIK-1, I'll definitely need these to measure the woofers like you mentioned, as I don't really plan to replace these as the dimensions of these drivers are very hard to come by in the aftermarket.

I had an idea once of making an MDF speaker spacer with my router, or even using 3D CAD to create a drawing and send the files off to waterjet cutting, I've done this with an archway for a door/wall once, super easy, but then I'd have to go down to a 7" woofer, I think I'll definitely measure what I have to start with, as they might very well be up to the task.

The thing that worries me though is that the magnet of the 8" woofers are 100mm, much the same diameter as the AC-130F1's or MK2's, it makes me doubt the 'woofer' capability of the speakers, especially considering I'm sealing the cabinet and these are tuned for the vented operation.

Measurements are key, I'll follow up on that, for sure.

Again, money is not the biggest concern here, the speaker themselves are $300USD alone, spending another $300USD or so could still yield a better speaker than what's available for purchase, very interesting for sure.

And you're correct - Aurum Cantus is the perfect match, physically.

Thanks man, I appreciate your post!

P.S. Please scroll down and have a look at the very last quote in this post, a forum member on here asked for which center speaker I am matching these speakers too as the whole project is targeted for home cinema, your information and knowledge could be very useful there, it's the next speaker up in line for improvement!

jReave said:
Also - about your existing tweeter.

The fact that it's known to blow could be inherent to the tweeter but it may also depend on the xo filter, ie. on the implementation. If the xo point is too low and/or if the xo slope isn't steep enough the same problems can happen. So it may be possible to make improvements with a re-designed xo. Or not. Could just be a cheap tweeter in need of replacement in the end.

Yes, could be both, but then again why would the manufacturer design a bad crossover when the speaker has had several revisions up until now, the speaker series has been around for years.

I think it's more about the tweeter itself being limited by the budget, basically they must of thought the tweeter was up to the task, to moderate listening levels - and called it a day.

I mean, who would buy a speaker for $300USD with the aim of listening to reference levels?

lowmass said:
1-get a measurement system and measure each speaker with stock crossover
2- then measure each drivers response in cabinet seperately without crossover connected.
3- now u have the info u need to even start thinking about what u might do

until u do this u really are shooting in the dark

my guess is that with a decent cheap tweeter and a complete crossover re design guided by the measurements, you likely could end up with a better sounding speaker thats about 3 db less sensitive than stock and what u learn following these simple rules will pay big dividends in the future. Keep it simple and you will not fall into chasing your tail

you would be surprised how good even cheap drivers can sound when all is balanced well and the experience of doing that is worth a bunch

Noted, I will take the list going forward when doing my measurements, thank you.

A question though, I take it you're trying to tell me that a less sensitive speaker sounds better, why is that? Sorry for the noob question.

Yes, otherwise I agree, knowledge and smartness is key, expensive is not always best, very much on point here.

About the drivers I chose, I basically picked them because they fit the measurements of my cabinet without any modifications, apart from some mounting holes having to be re-drilled, some cutouts made larger for the magnets, but I have a router, no worries.

Thanks for posting!

system7 said:
I've been reverse-engineering computers and loudspeakers for years.

I enjoy it.

This is a WWMTM.
We also know we have two 4 ohm 5" mids in series. Making 8 ohms.
And two 8 ohm 8" woofers in parallel, making 4 ohms.
A 6 ohm 1" tweeter with an ohm or two of attenuation.

It's not going to be a million miles from this off-the-shelf crossover, except missing the 0.68mH mid rolloff:

HW 3/130 NG - 8 Ohm | Visaton

I think I'll just Boxsim this one with some similar drivers: Software | Visaton

Bass filter must be 3mH, with a 4.7R and 6.8uF shunt.
Tweeter must be 4.7uF and 0.25mH and 2 ohms resistor in front.
Mid filter must be 100uF with a 1.5mH and 5.6uF shunt.

We can do this!

I'm not getting it, you basically came up with the same crossover as the one that's in place right now, by using similar drivers in BoxSim?

What's a 'shunt', is it the same as a slope generated by a capacitor?

You never tried to BoxSim with my suggested midranges and tweeter?

Thanks!

Moondog55 said:
First off try filling the midrange compartment with fibre, any fibre will do for the first experiment, even an old woollen jumper
Less damping means lower manufacturing cost. Most talk of and about wiring tends to snake oil after a while, what the manufacturer uses is probably fine. Wiring length inside a cabinet is short
XO values are probably correct and 100V capacitors mean that reasonable care has been taken Re quality but pictures of the actual XO would help. If these are for movies only then L&R quality on voice is less important as most dialogue comes from the centre speaker, what are the details of the centre speaker?

Very interesting topic you made, about filling the compartement completely with fibre, I have a question about this and it concerns filling the MTM enclosure when the drivers are open-back?

Wouldn't it cause interference with the spider and cone (possibly voice coil) during operation?

For this very reason I had original plans of just using dampening materials along the sides of the cabinet internals.

About wires, it just feels good to have them replaced I guess, like a quality assurance, but most likely snake oil, yes, as an electrician I would almost certainly agree, however there are such variables involved with creating audio and sound that one might always wonder what's best and what's overkill.

I have quality wires outside the cabinet already, I will most likely fall into the pit and spend money on quality wires internally as well, so there you go, we live in a materialist society.

I don't see how a simple picture of the crossover can make someone figure out how it's wired, or to what use such a picture might be, but I could certainly snap a picture and have it your way, expect it tomorrow, I'll buzz you.

You make another very interesting topic about what center speaker I plan to match these speakers to, and that's the next speaker in line for an improvement (as per my original plans).

You see this center speaker is matched to the front towers from the manufacturer, it consists of the same tweeter and midranges as the fronts, added to that is two woofers that serves the lower frequencies in the cabinet.

I wasn't planning on discussing that speaker until I've settled with the fronts, but what the heck, might be better of planning/investigating/measuring this speaker too and see what drivers might be of good use if deciding that a driver replacement is the key for an improved sound.

I've taken center speaker apart as well, but I did not investigate the crossover as I did with the fronts today.

By eyeballing, the woofers taking care of the lower frequencies look EXACTLY like the midrange speakers, exept from the dust cap which is larger than on the midranges, the voice coil, cone, surrounds, magnet, well - everything else is exactly the same.

I think there's just marketing going on here, because the manufacturer offered a center speaker without the additional low frequency woofers before this center speaker got introduced.

Old center speaker:

New center speaker (the one I got, and plan to improve):

Of course, they might of fiddled around in the crossover and introduced a dedicated speaker for the low frequencies, but to which extent is there an improvement over the old design?

The new design also implements two ventilation ports on the rear.

It might of been tuned to a lower frequency, and the addition of two extra drivers is an addition in dB output?

How exactly is more bass achieved?

My intent was not to discuss this speaker until later, but there you go, we might as well start now.

Thanks for asking, very interesting from your side!

Lojzek · 2020-02-14 8:58 pm

Can you post a photograph of the woofer taken out of cabinet from the magnet side?

Moondog55 · 2020-02-14 9:20 pm

Adding more damping material will have no effect on the driver unless you use expanding foam [ joke] but it will have an effect on the midrange performance, how much effect you can't tell but expect a little more clarity. Adding more stuffing is a simple cheap tweak. Actual pictures of the XO can tell us a lot about the quality of the components Photos of the box construction can tell us if there have been shortcuts taken with bracing and sealing although from the drawings it seems to me this has been well thought out.
From my perspective those centre speakers have the usual but incorrect orientation and for HT all speakers in the system should use the same [ or very similar] construction and components and "Yes" tweeters are often the weak link in value for money systems and as already mentioned a tweeter upgrade might be your best value although you may want to try a conjugate Zobel addition first, such small additional circuits often help a cheap tweeter

jReave · 2020-02-14 10:19 pm

Re your woofers - forget for a moment about the size of the magnets. It's how they sound that matters. If you are happy with how they sound sealed then there is no need to change them. On the other hand, if you feel there is room for improvement....... maybe you want to replace those as well.

Can't find a size match? Well then it's maybe time to think about cutting the old front baffles right off the old speakers and putting on new ones. That should allow you more freedom in driver choices as well.

The problem with an MTM style CC is that you get comb filtering in the horizontal plane. This means the FR around the xo point is going to vary (sometimes dramatically) depending on where you are sitting on the couch. If you happen to be the only person doing the listening then this might not be such a big deal.

To avoid the problem in the 1st place, you need to go WT/MW where the tweeter and the mid are one above the other. The problem with that is that it means a higher speaker height which can make placement more of an issue. And for you, you'd need to build a whole new CC cabinet to make that work. Using the same drivers as your front mains is usually a good idea too, as Moondog55 suggested.

You're going to have so much to figure out with your towers, I would suggest worrying about the CC is best left until later. Your choice of course.

Vitus · 2020-02-15 3:04 pm

Lojzek said:
Can you post a photograph of the woofer taken out of cabinet from the magnet side?

Moondog55 said:
Adding more damping material will have no effect on the driver unless you use expanding foam [ joke] but it will have an effect on the midrange performance, how much effect you can't tell but expect a little more clarity. Adding more stuffing is a simple cheap tweak. Actual pictures of the XO can tell us a lot about the quality of the components Photos of the box construction can tell us if there have been shortcuts taken with bracing and sealing although from the drawings it seems to me this has been well thought out.
From my perspective those centre speakers have the usual but incorrect orientation and for HT all speakers in the system should use the same [ or very similar] construction and components and "Yes" tweeters are often the weak link in value for money systems and as already mentioned a tweeter upgrade might be your best value although you may want to try a conjugate Zobel addition first, such small additional circuits often help a cheap tweeter

Thanks for sharing your experience, I found a rather good dampening material that I'm considering covering all blank areas that are missing dampening currently, your advice on more midrange clarity with added dampening is nice to hear.

For woofers I guess you add dampening to decrease "boominess/bloating/mudding".

I took some photos of the crossover/crossovers, HF/LF cards, see below:

HF Filter:

http://www.bilddump.se/bilder/20200215153702-46.195.138.78.JPG

http://www.bilddump.se/bilder/20200215153712-46.195.138.78.JPG

http://www.bilddump.se/bilder/20200215153722-46.195.138.78.JPG

http://www.bilddump.se/bilder/20200215153730-46.195.138.78.JPG

http://www.bilddump.se/bilder/20200215153738-46.195.138.78.JPG

LF Filter:

http://www.bilddump.se/bilder/20200215153620-46.195.138.78.JPG

http://www.bilddump.se/bilder/20200215153635-46.195.138.78.JPG

http://www.bilddump.se/bilder/20200215153646-46.195.138.78.JPG

Here is a photo of the insides of the cabinet and the bracing, you've already seen the drawings, unfortunately I didn't remember to take photos of the inside of the cabinet, fortunately however - someone else uploaded a great photo that might be useful, see below:

http://www.popular-hifi.com/projects/nad_stereo/m85ex/m85ex_bild7.jpg

I know a center channel should be exactly the same as the F/R channels in a home theater setup, but this is in a living room and I'm unable to put a speaker that big behind the screen, that's why I purchased the center speaker I mentioned.

I can't do magic, but I could certainly possibly improve the performance out of it.

Thanks!

jReave said:
Re your woofers - forget for a moment about the size of the magnets. It's how they sound that matters. If you are happy with how they sound sealed then there is no need to change them. On the other hand, if you feel there is room for improvement....... maybe you want to replace those as well.

Can't find a size match? Well then it's maybe time to think about cutting the old front baffles right off the old speakers and putting on new ones. That should allow you more freedom in driver choices as well.

The problem with an MTM style CC is that you get comb filtering in the horizontal plane. This means the FR around the xo point is going to vary (sometimes dramatically) depending on where you are sitting on the couch. If you happen to be the only person doing the listening then this might not be such a big deal.

To avoid the problem in the 1st place, you need to go WT/MW where the tweeter and the mid are one above the other. The problem with that is that it means a higher speaker height which can make placement more of an issue. And for you, you'd need to build a whole new CC cabinet to make that work. Using the same drivers as your front mains is usually a good idea too, as Moondog55 suggested.

You're going to have so much to figure out with your towers, I would suggest worrying about the CC is best left until later. Your choice of course.

You triggered a very interesting idea in my head, about removing the current baffles and manufacturing my own ones, but that idea came to an end when I further investigated the possibilities with my center channel.

You see, I have Dolby Atmos upfiring speakers too, and these use a bigger tweeter than the one I have been looking at, and the one that's currently in my F/R channel speakers, it's 30mm instead of 25mm, also uses a waveguide, which I think could further improve the off-axis response from my towers and center channel.

I opened my center channel up and unfortunately there's not enough room to place the midrange/woofers more closely to the outer edges of the cabinet (with a new baffle) to accomodate for the larger tweeter that is 115mm instead of the original 100mm.

So the idea of creating my own baffles sort of died on me, I see no point when I can't fit the drivers that I want to, instead I'll just have to refer to the 100mm limitation of the tweeter and focus on finding the best possible (SB Acoustics SB26STAC-C000-4) and narrow down my area of improvement to the midrange and center channel woofers.

I'll have to measure my current F/R channel woofers and see how they perform, I have a very difficult time finding a replacement for these, but then again, as some of you have mentioned, maybe they don't need replacing, I could just optimize the cabinet and crossover and maybe they'll be more than fine, really.

UMIK-1 and DATS V3 is a must.

About comb filtering and the center channel, yes, I know, but it's unavoidable, I don't have room for any other center channel, and I can't really live without it, sort of a tricky situation, I don't think I'll complain, in either case.

All systems have their limitations and constraints, these are what ultimately forms the limits of these sorts of projects.

I think I'll just leave the CC as is, just change the drivers if it looks promising.

Building a new cabinets will have to be the next project I guess.

Thanks a lot!

Vitus · 2020-02-15 3:12 pm

system7 said:
I think I'll just Boxsim this one with some similar drivers: Software | Visaton

Bass filter must be 3mH, with a 4.7R and 6.8uF shunt.
Tweeter must be 4.7uF and 0.25mH and 2 ohms resistor in front.
Mid filter must be 100uF with a 1.5mH and 5.6uF shunt.

We can do this!

Vitus · 2020-02-15 5:18 pm

How do you actually use WinISD?

I've entered all the parameters, but there's loads to fill in that just aren't used eventhough you try to fill out as many as possible in the driver specifications, I've tried using both auto-calculation and without.

Is the final Qtc all that matters?

I select 0.707 Qtc for a sealed enclosure (my MTM enclosure), should my box volume match as closely to 0.707 as possible, that's the goal here?

There's a plotting curve that shows up, what to look for in that?

Thanks!

(Beginner) An attempt to improve my speakers, help appreciated.

Member

Member

Member

Member

Member

diyAudio Member

Member

Member

Member

Member

Member

Member

diyAudio Member

Member

Member

diyAudio Member

Member

Member

Member

Member