I'm really enjoying reading some of the current posts today! Between this and a few other threads (like the Dan Wiggins vs Feandil) I've been laughing continuously since I got home from work. The entertainment value of this site has skyrocketed IMHO. I love the internet. Keep up the good work kids, err, fellas
cheers,
AJ
cheers,
AJ
Along with my friend I have listened extensively to the Accutons, several models including the C2-76, C2-88 and C2-92. Like all drivers they have some strengths and some limitations.
According to the waterfall plots there were some resonances, the cones seem more likely to flex between oval and round (x-y plane) than in the traditional sense where they flex in the z plane. In order to make them sound best we needed to include some narrow traps.
Their strenghts include amazing detail resolution leading you to believe they are very clear. Their biggest weakness IMHO was their lack of dynamics; i.e they only sounded best at one volume. I also did not like their overly analytical presentation. We debated adding filters to make them warmer but reasoned that the extra complexity added significant cost to drivers that weren't cheap to beging with and that the extra components might damage their main strength, resolution. In the end we used a polypropylene midrange, quite conventional for a cone material. This unit designed by the Skaanings has been praised at other points in this thread, known as the Quenze units.
In my opinion the ceramic drivers do offer some interesting alternatives, I am less sure that they represent the best possible answer for a midrange today. Their diamond tweeter sounded quite good, as to whether it sounds about $2,500 better than the $300 competition, I'll leave it up to you to decide.
According to the waterfall plots there were some resonances, the cones seem more likely to flex between oval and round (x-y plane) than in the traditional sense where they flex in the z plane. In order to make them sound best we needed to include some narrow traps.
Their strenghts include amazing detail resolution leading you to believe they are very clear. Their biggest weakness IMHO was their lack of dynamics; i.e they only sounded best at one volume. I also did not like their overly analytical presentation. We debated adding filters to make them warmer but reasoned that the extra complexity added significant cost to drivers that weren't cheap to beging with and that the extra components might damage their main strength, resolution. In the end we used a polypropylene midrange, quite conventional for a cone material. This unit designed by the Skaanings has been praised at other points in this thread, known as the Quenze units.
In my opinion the ceramic drivers do offer some interesting alternatives, I am less sure that they represent the best possible answer for a midrange today. Their diamond tweeter sounded quite good, as to whether it sounds about $2,500 better than the $300 competition, I'll leave it up to you to decide.
that is a very interesting post, hermany. i've yet to hear any accuton midranges, but i heard some of their tweeters and liked them. based on your observations they seem to also have that flexing problem, which is likely the cause of their weaknesses. how did you include the narrow traps? what made you think it flexed in the x-y plane?
their analytical presentation i don't think is a result of their rigidity, since the Dayton RS series woofers sound very warm in the midbass region despite being rigid metal cones. I think that analytical sound you describe is a result of high 3rd order distortion in the midbass region, which I experienced with the Adire Extremis 6.8.
anyway, i've had very very good results with the Peerless Exclusive 7" as my midrange and wonder if the $50 Peerless Exclusive 830882 would be the best midrange as of today. Tests show it's good to 3kHz. it uses a very good motor but cheap paper cone and highly dampened rubber surrounds that are big. i bet this midrange can be improved upon if we give it a diamond cone and make it surroundless.
their analytical presentation i don't think is a result of their rigidity, since the Dayton RS series woofers sound very warm in the midbass region despite being rigid metal cones. I think that analytical sound you describe is a result of high 3rd order distortion in the midbass region, which I experienced with the Adire Extremis 6.8.
anyway, i've had very very good results with the Peerless Exclusive 7" as my midrange and wonder if the $50 Peerless Exclusive 830882 would be the best midrange as of today. Tests show it's good to 3kHz. it uses a very good motor but cheap paper cone and highly dampened rubber surrounds that are big. i bet this midrange can be improved upon if we give it a diamond cone and make it surroundless.
The flexing is a guess, the cones are not conical in shape they are hemispehrical which would allow easier flexing in that plane where a conical shape I think would be more resistant. I'll bet you could see any flexing with a stroboscope and a felt marker (carefully!!!) on the cone face.
I you do decide you like the Accutons, remove that damn grille, it rings like a chime (yes, I know I've said this here before but it was so obviously a problem and they still make them that way. The ringing even seems to appear on their own waterfall plots, they blame it on the room) We bent round brass stock from a hobby shop and spanned the front of the driver. Inserted the bars under curved tension right into the baffle. The brass was pretty dead and didn't look bad either.
We had the use of MLSSA so designing traps wasn't hard, both of us are EEs so the values were easy to calculate. I suspect they would be somewhat enclosure and crossover value dependent so even if I could find the drawings I wouldn't publish them.
I you do decide you like the Accutons, remove that damn grille, it rings like a chime (yes, I know I've said this here before but it was so obviously a problem and they still make them that way. The ringing even seems to appear on their own waterfall plots, they blame it on the room) We bent round brass stock from a hobby shop and spanned the front of the driver. Inserted the bars under curved tension right into the baffle. The brass was pretty dead and didn't look bad either.
We had the use of MLSSA so designing traps wasn't hard, both of us are EEs so the values were easy to calculate. I suspect they would be somewhat enclosure and crossover value dependent so even if I could find the drawings I wouldn't publish them.
New 2006 version Seas Excel 5.5" Nextel mid:
http://seas.no/Product bilder/Art-pict.-M15CH002.opt..jpg
http://seas.no/PDF data excel 05-06/M15CH002E0043.pdf
What's not to like?
http://seas.no/Product bilder/Art-pict.-M15CH002.opt..jpg
http://seas.no/PDF data excel 05-06/M15CH002E0043.pdf
What's not to like?
infinia said:
in most DIY 3-way designs, mids are used between 150-3000Hz, so I don't see directivity being a problem. Also, subjectively lower midrange performance is more important than the upper midrange, so I would lean toward a larger midrange rather than a smaller one. this is not to say that larger midranges don't perform well up high, they can.
looks like it has a huge xmax for a dedicated midrange, and the xmax will not even be used up.
the older version using a similar (same?) motor but with a poly cone was tested by Mark K and was the best midrange he's tested on his website. Perhaps this one will test even better. Time will tell.
Here are some Görlich 13cm drivers.
http://cgi.ebay.de/PODSZUS-GORLICH-...ryZ21994QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
Any opinions about them?
http://cgi.ebay.de/PODSZUS-GORLICH-...ryZ21994QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
Any opinions about them?
Regarding Seas Excel M15CH-002
Directivity is my main issue. This driver would be used in a 3 way from 80-3000Hz at max SPL of 115 dB with a passive cross to a tweeter. For me to use this driver I would need at least a 3rd order LP filter along with a notch around 5 KHz. Even at the 3KHz cross this drivers directivity is 12 dB difference at 60 deg off axis. When I search for mid drivers the smooth high freq rolloff is the first thing I look at, for simple low distortion designs. I like to combine the natural low pass of the driver with very simple crossovers ideally a single inductor for a combined 3nd order LPF. Xmax is also just another indicaton of likely low distortion performance. I only have the specs for my initial decision to purchase an expensive item. This may or may not say anything about the sound of the driver in question. Also they claim low distortion but provide no data to back it up. Cotd please don't take it personally. I like to know what other people use for midranges in their systems and why. The midrange is the foundation of a well designed system. I wish I could listen to them all and even test before i made a choice. Peace
Directivity is my main issue. This driver would be used in a 3 way from 80-3000Hz at max SPL of 115 dB with a passive cross to a tweeter. For me to use this driver I would need at least a 3rd order LP filter along with a notch around 5 KHz. Even at the 3KHz cross this drivers directivity is 12 dB difference at 60 deg off axis. When I search for mid drivers the smooth high freq rolloff is the first thing I look at, for simple low distortion designs. I like to combine the natural low pass of the driver with very simple crossovers ideally a single inductor for a combined 3nd order LPF. Xmax is also just another indicaton of likely low distortion performance. I only have the specs for my initial decision to purchase an expensive item. This may or may not say anything about the sound of the driver in question. Also they claim low distortion but provide no data to back it up. Cotd please don't take it personally. I like to know what other people use for midranges in their systems and why. The midrange is the foundation of a well designed system. I wish I could listen to them all and even test before i made a choice. Peace
I have to agree that driver specifications have only a cursory relationship to sound quality. In my limited experience a driver with good out of band behavior expecially above its upper limit i.e. a relatively smooth break-up curve will give you something to work with for crossover design.
If a drivers behavior near its operational limits dominate the crossover requirements then final voicing becomes a more serious problem. As you listen to your design, if you are like me, you will want to make some fine adjustments in both crossover points and in the relative levels of the drivers. By using drivers that require steep filters to sound best many of these kind of options or adjustments dissapear or become painfully difficult to implement.
This doesn't mean those drivers can not be made to sound good, it's just so much work that it might be better left to a commercial developer who has time and manpower to make significant final design changes to accomodate a driver that is obstinate near its limits.
This task is much easier if you have access to powerful analysis and design tools. If you are doing this by ear, my advice would be to pick drivers that are well behaved at their limits and compatible with first order networks. Based on my experience this means that the exotic cone material drivers present significant design risk, it's not that they can't sound good, it's just that it may be quite technically difficult
If a drivers behavior near its operational limits dominate the crossover requirements then final voicing becomes a more serious problem. As you listen to your design, if you are like me, you will want to make some fine adjustments in both crossover points and in the relative levels of the drivers. By using drivers that require steep filters to sound best many of these kind of options or adjustments dissapear or become painfully difficult to implement.
This doesn't mean those drivers can not be made to sound good, it's just so much work that it might be better left to a commercial developer who has time and manpower to make significant final design changes to accomodate a driver that is obstinate near its limits.
This task is much easier if you have access to powerful analysis and design tools. If you are doing this by ear, my advice would be to pick drivers that are well behaved at their limits and compatible with first order networks. Based on my experience this means that the exotic cone material drivers present significant design risk, it's not that they can't sound good, it's just that it may be quite technically difficult
crossover design is easy these days, all of my DIY speaker friends know how to use speaker workshop or lspcad. these programs are very accurate, so i don't see using drivers with early breakup is a problem. some of the most detailed transparent drivers have nasty breakups but sound great once filtered.
infinia, it seems like you are looking for slow crossovers? why? steeper lower-frequency crossovers integrate better off-axis and have less distortion because it's for a far smaller portion of the frequency spectrum. If you're looking to play down to 80Hz, only mids that can do that are 6.5" and larger midbasses. No driver I know of for this size can go as high as 4kHz since like you said, there's directivity issues. dedicated midranges are generally better than midbasses for the midrange, IMHO. most midbasses have energy storage issues well before they reach 2kHz.
infinia, it seems like you are looking for slow crossovers? why? steeper lower-frequency crossovers integrate better off-axis and have less distortion because it's for a far smaller portion of the frequency spectrum. If you're looking to play down to 80Hz, only mids that can do that are 6.5" and larger midbasses. No driver I know of for this size can go as high as 4kHz since like you said, there's directivity issues. dedicated midranges are generally better than midbasses for the midrange, IMHO. most midbasses have energy storage issues well before they reach 2kHz.
Once you get more experience things will go from easy to difficult again. You have to evaluate many considerations besides flat on axis spl responses. Yes I mainly use 6.5" -7" midbass to 3KHz as i mention many times in this thread not 4KHz. Hint, look at phase and group delay of high order networks and consider the effects as drivers transition on their far field sound propagation, easy huh? look at testing impulse responses then the first order networks are fastest.
jdybnis said:
jdybnis: That is more than a bit over the top.
Crossover design is never simple, fancy software or not. If it were as easy as you pretend, there would hardly be the hundreds of different speaker designs on the market, each with certain strenghts and weaknesses. In spite of 60 years of mathematical engineering experience, the nice, simple, Linkwitz-Riley crossovers were not developed until the 1960 (70?) decade.
The best possible designs still have difficulties so advice that this might be more difficult than it seems is good valid advice.
Yes anyone with a little experience can throw togther a simple design and probably beat the mass market speakers at $1,000 or less.
This thread seems to be asking, what if I want more than that?
infinia said:
And in context it sounds even worse. Someone engages you in discussion. You respond by implying they are inexperienced and then you misrepresent their comments.
infinia said:
1) The poster cites off-axis response and distortion as reasons for using high-order filters. You respond by telling them to look at things other than on-axis response.
2) The poster says that their opinion is there are energy storage issues below 2k. You reply that you only use a midbass to "3kHz...not 4kHz" and you imply that the poster did not read the beginning of the thread.
3) You devalue the poster again by implying they are missing some trivial analysis. On top of that, you do some handwaving about phase, group delay, and far-field sound propagation without actually providing what that analysis is.
Easy huh?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.