|
Posted by Murataltuev on
08-12-2015
|
|
fiogf49gjkf0d
I would like to share my 5-way horn project.
The idea was to have full range system, so the first horn built was Tapped Horn.
I got design from Volvotreter.de and was really impressed with performance of this monster.
16hz is just true thunder in my room  (I'm using them in HomeCinema system also).
For the midrange I'm now using 2m length horn with CD which is capable to go down to 300Hz, so for the midBass (70-250hz in my case) I've designed straight 2m length with 1x1m month for 12" Supravox with help of 3D acoustic modeling software. For the midrange 2m length is may be too long, so I'm going to experiment with LeCleac'h 660mm length, 600mm mouth and 1.5" throat. HornResp model shows perfect curve from 250Hz for TAD2001. I have Thiele-Small parameters of this driver, but I'm still thinking about right driver for this range and going to decide after experimenting with say everything. I'm not limited in budget and going to try JBLs, Vitavox S2 and WE555 clone from LM.
For the upper range I'm now experimenting with JBL LE085 and like the sound, but I'm going to try other drivers as well. And for the 12k+ range I'll use RAAL Lazy Ribbon 9" as this is the only option to get highs on 110+db speakers.
At this stage this is the design of concept, so I appreciate any comments and suggestions.
Thank you,
Murat
|
|
|
|
Posted by Romy the Cat on
08-12-2015
|
|
fiogf49gjkf0d Murat,
I am not qualified to talk credibly
about Tapped Horns. I know that they get a lot of traction now around audio
people but I am, without listening a good example of it, am not so warm to the Tapped
Horn idea. The comments of yours that your horn is “thunder in room” do not help ether… I do not think
that you shall pay attention to anybody feelings regarding the Tapped Horns. If you set your hart to this way to
get lower bass then it is what it is and let see what design idea you have.
The configuration you outline is relatively close to
Macondo, so I know this setting very well. A few comments I would like to
share. It might be juts bad drawing but the MF horn shall not be offset on the
right and has to be on the same vertical line as the rest of the drivers.
Now, the main kink of the installation you propose is the
use of slow opening upper-bass horn and integration of Tapped Horn into the speaker frame. The exponential upper-bass
horn is fine. It is a noble attempt to try but you need to understand the design
consequences. Anyhow, if it is what you want then I would advise to beef up a transition
between frame and upper-bass horn and to make the bottom upper-bass horn
supporting platform much smaller and perhaps moving on rolls. Regarding the integration
of Tapped Horn into the speaker frame I am very much
opponent of this idea. If you invest time and money into building something like
this then why you couple the bass and MF channels in the same assembly? What
would you do it you let say left loudspeaker in your left room will create -14dB
suck out in your room? And it will!!!! The whole point to have in more advanced
acoustic system the bass and MF channels separated become you will have a change
to put MF in the room in respect where you would like to be and LF in the room
in respect where your room demands it. If your LF and MF are physically
integrated then you play lottery with your room. I hope you understand that at
50-60hz there is no correction that you will be able to make.
So, if you decided to separate
LF and horns into a separate physical entity then you can address the worse
thing (in my view) in your design idea: the front baffle. You see, it does not
sound intuitive but horns are very sensitive, very much alike the direct
radiators to surface of front baffle. I
always advocate to keep the horns away from any horizontal surface behind them.
In your case the upper MF horn and the tweeter are sitting on the baffle and it
is not good (again, it is in my view).
If you do insist in this front-baffle
configuration then here is what I would advise you to do. The horn in time
alight position will not be aligned in vertical plane but they will be curved forward
with the curve more if you sit closer. So you can make atop baffle not flat but
introduce sort of like bi-relief sculpture, sort of semi-trapezoid form of baffle.
Think about the shape of Russian tank T34 with its sloped armor. You can
do the same sloped baffle.
If you look how the tweeter mounded in Avalon Eidolon then it would give you an
idea. Anyhow, whatever you do I would never mount tweeter in the way you
propose. Make the sloped baffle and cover the slops with acoustic foam to assure
that it will be no acoustic fog coming from the frame
surface.
Rgs, Romy the Cat
|
|
|
|
Posted by Paul S on
08-12-2015
|
|
fiogf49gjkf0d Murataltuev, I am not qualified to advise about the fine points of horns, per se, but I am guessing the big "slab" in the middle of each horns tower will create more problems than it solves, at least with respect to sound. Honestly, unless you are OK with giving the final sound to "pro-level" DSP, I think you might never be able to overcome the sonic effects of those slabs as you depict them. Certainly, you will not get the aggregate HornResp prediction you're aiming for at your listening chair. Perhaps there is another way you can bring the tapped horn LF into the room, so the horns themselves can "breathe"? Have a look at the Avantgarde Duo Mezzo. This is a "minimal" horn/hybrid speaker system that "works".
Best regards,
Paul S
|
|
|
|
Posted by Murataltuev on
08-13-2015
|
|
fiogf49gjkf0d Thank you for comments.I agree with front-baffle concerns.I'll redesign it.But I'm still thinking to keep Tapped Horn integrated somehow to the whole setup and may be use RoomCorrection, but only for Tapped Horn (20-50Hz).Another problem with Tapped Horn is that it needs 4-8 order filter at 50Hz, so going digital for this range looks rational.
|
|
|
|
Posted by Romy the Cat on
08-13-2015
|
|
fiogf49gjkf0d  Murataltuev wrote: | Thank you for comments.I agree with front-baffle
concerns.I'll redesign it.But I'm still thinking to keep Tapped Horn integrated
somehow to the whole setup and may be use RoomCorrection, but only for Tapped
Horn (20-50Hz).Another problem with Tapped Horn is that it needs 4-8 order
filter at 50Hz, so going digital for this range looks rational.
|
|
Well, Murat, if you are willing
to use RoomCorrection then why do you need to build anything excessive, time aligned
or anything else “pushy” at this point? The whole point of designing of intelligent
installation is not use any artificial means and they do bring a lot of own problems.
It is like making food tasty and excited without using MSG or like making a woman
happy and thrilled without use of kinky toys.
One more thing. The fact
that you need a super sharp filter atop of Tapped Horn is a large warning sign.
Yes, it would not be hard to render it on digital level but the structure of lower
bass is what all those digital toys destroy very first. I do not know you and I
might be way off the marks but I feel that you (along with 99.9999999%) of the
rest audio people has a bit mistaken expectations of how bass shall be shaped
by audio installations. Spend some time to listen live and not amplified
classical music. You will not be able to get the same bass but you shall be
able at least to set some own expectations where you are going with all of it.
|
|
|
|
Posted by Murataltuev on
08-13-2015
|
|
fiogf49gjkf0d
I understand all disadvantages going digital.
I strongly believe that minimal influence to the signal is the most important aspect.
And influences in digital domain are usually terrible.
But I'm not sure what other options available.
I like tapped horn, because in working range 20-50hz single 15" diaphragm displacement is in 2mm range...and sensitivity is enough to drive it by SET amp. And nothing can be compared with SET driving even lowest bass!
I had an experiment in my room comparing the best 2000W D-class and 500W AB-class transistor amps with 18W SET and even on cinema tracks choice was obvious.
I don't clearly understand how you get your 20-50hz range driving by SET. Difficult to believe that relatively small drivers in sealed enclosures can do it.
I'll learn now through your site.
Hope to listen to your system one day if it is ever possible
|
|
|
|
|
|
|
Posted by scooter on
08-13-2015
|
|
fiogf49gjkf0d Hi Murat,
1. Please take a look around this site for some intelligence on how detrimental digital room correction is and why. Romy and Paul are being excessively polite today regarding DSP, perhaps because the weather is nice. I'll be more direct - If you are taking the time to design and build a good system, why risk killing it with some half-baked DSP "solution"? Why not just design the system properly out of the box?
DSP can be super useful for testing and system integration. Just put it back on workbench with your hammer and your chainsaw after you are "done."
2. How many systems have you heard that properly integrate LF? Think carefully about why that is.
There are only a handful of key elements, but you need to get them all right. Secondary elements are important, but secondary. LF is a different and extremely problematic beast because the room matters so much. Think carefully about why electronic DSP "solutions" could struggle with LF room nodes. Comments littered through this site can be a great guide.
3. Also, I would like to address your "thunder" comment. A properly integrated system should be integrated. Seamless (well, as seamless as possible). Thus, if you notice "true thunder," your LF solution is not properly integrated (unless you are listening to recorded thunder). If that concept seems idiotic or trite or simplistic or childish, in your mind's eye, imagine how good music is realized in a good hall.
4. Before moving forward, I think it would be a fun and useful exercise to spend a few weeks trying to get your current LF solution properly integrated. That is like taking one step back and three steps forward. I strongly believe the results will influence your new project.
5. I would like to concentrate on one element, emphasizing that by separating the LF physically, you will have an excellent "manual DSP"
solution, just put the LF boxes on sliders and shuffle them around the
room to find "optimal" placement. Once you find the "best" placement, you can easily adjust gain. I think that is a good "brute first step" for proper integration. Some people may not need to do this type of testing, but do you honestly think you have the ability to model your room's LF response without physical experimentation with absolute certainty? I certainly don't have that ability. And what if your wife redecorates the room, remodels the room, or moves you to another house? Regardless, the option to move LF around for "free" is quite a valuable "option" in the experimental phase.
Anyways, good luck!
|
|
|
|
Posted by Murataltuev on
08-13-2015
|
|
fiogf49gjkf0d My 20-50hz TH is not very sensitive. HornResp shows 95db for Ang = 2.0 x Pi.So, it needs power. My 18W GM-70 SET can drive them without clipping, but I prefer to keep amp working at 10% of max power to keep distortions as low as possible, so for TH I'm building huge 100W GU-81 amp.I'm not sure about full active approach (6 amps!), but for TH it is the only way.
|
|
|
|
Posted by Romy the Cat on
08-13-2015
|
|
fiogf49gjkf0d | Murataltuev wrote: | | My 20-50hz TH is not very sensitive. HornResp
shows 95db for Ang = 2.0 x Pi.So, it needs power. My 18W GM-70 SET can drive
them without clipping, but I prefer to keep amp working at 10% of max power to
keep distortions as low as possible, so for TH I'm building huge 100W GU-81
amp.I'm not sure about full active approach (6 amps!), but for TH it is the
only way. |
|
Hm… Why to keep an amp 10%
of max power? You need to keep it in A1 that all that matters. If you have GM-70
and you feel that it drives your TH then do DSET with GM-70. Get rid of your expensive
sectional GM-70 transformer and introduce single section but super high inductance
iron slab. Optimizing currant and voltage you shall be able to get good 50W out
of a single GM-70 for such a low bandwidth. You can easily factor in a 2nd
- 4th order filter right in the amp. You even if you have excessive
power can ease the GM-70 load and you will lose some of distortions. If I believe
that my 18W GM-70 SET can drive whatever it drives without clipping then that would
be exactly what I would do. I am not sure why you do the GU-81 is you feel that
18W is enough to you. BTW, the GM-70 is DH and can easy work very much atop of
A1. If you have enough current in your driver then GM-70 enters A2 very gently.
If you feel that 18W was enough for you then you have a LOT of options.
|
|
|
|
Posted by Murataltuev on
08-13-2015
|
|
fiogf49gjkf0d | scooter wrote: |
1. Please take a look around this site for some intelligence on how detrimental digital room correction is and why. Romy and Paul are being excessively polite today regarding DSP, perhaps because the weather is nice. I'll be more direct - If you are taking the time to design and build a good system, why risk killing it with some half-baked DSP "solution"? Why not just design the system properly out of the box? DSP can be super useful for testing and system integration. Just put it back on workbench with your hammer and your chainsaw after you are "done."
|
|
I think it is still debatable what is worse: 50Hz 4th order passive filter with huge transformers or FIR filter with separate high-quality DAC.I'm now using huge transformers, but willing to try FIR especially considering the fact that my music library is digital, because my favorite performers do not publish vinyl.So, it is already digital source and while it is digital why not to LowPass in digital ?I never suggesting it for vinyl.
2. How many systems have you heard that properly integrate LF? Think carefully about why that is.
There are only a handful of key elements, but you need to get them all right. Secondary elements are important, but secondary. LF is a different and extremely problematic beast because the room matters so much. Think carefully about why electronic DSP "solutions" could struggle with LF room nodes. Comments littered through this site can be a great guide.
|
|
I agree. LF is most difficult to implement. And this site has really unique content.But, you know, in my expirience, better quality LF is easier to integrate.Before Tapped Horns I used huge 21" driver in box from PHC (http://proaudiotechnology.com/assets/lfc_21sm_3_view_drawing.pdf)That was really difficult to integrate
3. Also, I would like to address your "thunder" comment. A properly integrated system should be integrated. Seamless (well, as seamless as possible). Thus, if you notice "true thunder," your LF solution is not properly integrated (unless you are listening to recorded thunder). If that concept seems idiotic or trite or simplistic or childish, in your mind's eye, imagine how good music is realized in a good hall.
|
|
No, when I say "true thunder", I mean that movie scenes with thunder is real!I use the same setup in my home cinema to impress my friends
4. Before moving forward, I think it would be a fun and useful exercise to spend a few weeks trying to get your current LF solution properly integrated. That is like taking one step back and three steps forward. I strongly believe the results will influence your new project.
|
|
I understand what you mean.I actually don't like my LF currently, but I know what is wrong. I've followed some suggestions which turned to be false.Soon I'll get 2m upper bass horn ready and change drivers in TH then hope to evaluate and share the result.
5. I would like to concentrate on one element, emphasizing that by separating the LF physically, you will have an excellent "manual DSP"
solution, just put the LF boxes on sliders and shuffle them around the
room to find "optimal" placement. Once you find the "best" placement, you can easily adjust gain. I think that is a good "brute first step" for proper integration. Some people may not need to do this type of testing, but do you honestly think you have the ability to model your room's LF response without physical experimentation with absolute certainty? I certainly don't have that ability. And what if your wife redecorates the room, remodels the room, or moves you to another house? Regardless, the option to move LF around for "free" is quite a valuable "option" in the experimental phase.
|
|
One thing is important to consider, that TH (as any other LF, I believe) works much more effectively in the corner.So, placing two of them in the corners is rational. HornResp gives +10db for my TH when I change Ang from 2xPi to 0.5xPi.But in my case, corners will be occupied by speakers, that is why I think to integrate them.May be it is good idea to build another two TH in parallel with first pair and move them around back or side walls to eliminate modes.
|
|
|
|
Posted by anthony on
08-13-2015
|
|
fiogf49gjkf0d Hi Murat,
Personally, I would re-think the tapped horns being physically incorporated with the rest of the system. To start, from what I understand (I don't use a tapped horn) because they use both the front and the rear of the driver for sound, they have a time alignment that wanders a little with frequency. I concede that it is low bass and that we are supposedly not that sensitive to bass time of arrival in small rooms but it is what it is. But that little thing is nothing compared to the elephant in the room: it does not matter how good your bass solution is if it not placed into the room properly.
The room OWNS the bass below 200Hz or so. There is no way around this as the wavelengths of the frequencies become larger than the room dimensions so we get room nodes and so on and so forth. You can be fortunate like Romy and have a lower bass line array that fits into his room at the speaker position...but that is not very common. Most people have to move their lower bass around in the room so the the room nodes do not occur at the listening position/s. This may include just one bass source or three or more with them placed at odd spots in the room. The thing is that you just don't know until you put your bass source in the room and experiment.
If you have a regular rectangular room, you may be tempted to try a Double Bass Array. It will require DSP to run properly, but it does get rid of room nodes which may go a long way toward a sound that can be tailored to your requirements. Have a look at the first post here for a brief explanation of the concept. How does it sound? Don't know, but I have a feeling that as we take the room further out of the equation we are more free to work towards our own requirements for sound.
Regards,
Anthony
|
|
|
|
Posted by Murataltuev on
08-14-2015
|
|
fiogf49gjkf0d Ok, I give up...I'll keep TH separated.DBA concept is interesting and I don't understand why it is not common in Home Cinema or studio installations.In my case I'll consider to make double TH - one pair in front and position at 1/3 and 2/3 of wall width, and another pair at back.But I'm still thinking that TH is capable to produce bass more precisely because of very small diaphragm excursion in compare with any other type of sub.
|
|
|
|
Posted by anthony on
08-14-2015
|
|
fiogf49gjkf0d | Murataltuev wrote: | | Ok, I give up...I'll keep TH separated.DBA concept is interesting and I don't understand why it is not common in Home Cinema or studio installations.In my case I'll consider to make double TH - one pair in front and position at 1/3 and 2/3 of wall width, and another pair at back.But I'm still thinking that TH is capable to produce bass more precisely because of very small diaphragm excursion in compare with any other type of sub. |
|
Personally, I am probably going to do a line array like Romy's using the same drivers but twice as many...so eight each side. I have the drivers. This configuration should give somewhere around 109dB @ 20Hz from 15w in room with cone movement of about +-3.5mm. That is still low excursion, and they are excellent drivers. But first I am going to make some test boxes and get them in room to see how the bass works in my space. Fingers crossed I can keep them in a time aligned position with the horns, but I am not holding my breath.
DBA is difficult to setup because you need drivers at both ends of the room and time to measure things to get the delayed cancellation correct. Plus you need a symmetrical shaped room. So lots of drivers to mount on the walls at specific spots with amplification and DSP to drive them all. Not easy! Plus as far as I am aware it is a relatively recent idea.
|
|
|
|
Posted by Murataltuev on
08-14-2015
|
|
fiogf49gjkf0d | anthony wrote: | | This configuration should give somewhere around 109dB @ 20Hz from 15w in room with cone movement of about +-3.5mm. That is still low excursion, and they are excellent drivers. |
|
Well, in this case, I also have to try this configuration.109db is what I need to try to keep system running with single amp.What crossover slope are you going to use ?How are you implementing upper bass ?
|
|
|
|
Posted by anthony on
08-14-2015
|
|
fiogf49gjkf0d | Murataltuev wrote: | | anthony wrote: | | This configuration should give somewhere around 109dB @ 20Hz from 15w in room with cone movement of about +-3.5mm. That is still low excursion, and they are excellent drivers. |
|
Well, in this case, I also have to try this configuration.109db is what I need to try to keep system running with single amp.What crossover slope are you going to use ?How are you implementing upper bass ? |
|
The maths for the 109dB/15W goes as follows. According to the datasheet the driver is around 85dB sensitive at 20Hz. Double the number of drivers three times to 8 and SPL goes to 94dB/1W. Add the second speaker and we get to 97dB/1W. Multply 1W four times to get 16W of output and that makes 109dB/16W. Then we have to add room gain at those frequencies and subtract for the listening distance. Room gain is unpredictable but we are going to lose 6dB every doubling of distance to the listening position so if we listen at 4m we might lose 12dB which makes the speaker able to produce a maximum of 97dB at a 4m distance with 16W of power applied. Then there is room gain to add to this number which may take it to 100dB as the listening chair if we are lucky.
Anyway, that may or may not be enough SPL for the size of your room and the size of your amplifiers. Those particular drivers will take more power than 16W but I guess will probably start to be less than optimal with maybe 80W put through them (Romy would have better insight than I) which would raise the SPL by about 7dB. Is this loud enough in my case? I think it just might be. Is it good enough in your case? Depends on the size of your room and how far away you sit from the lower bass sources. I am counting on some room gain in those lower frequencies to raise the SPL droop of the driver and sealed boxes but how much I will get I won't know until I do it, but hopefully I won't have to resort to DSP to get a decent frequency response.
Regarding the other questions you asked, I am basically duplicating Romy's system as my starting point with horns. If I spend enough time on it I am hoping for a good result. Personally, I really dig the concept of his DSET amplification and the specialisation of the amplifier to match the frequency range and transducer and sound quality objectives.
|
|
|
|
Posted by Romy the Cat on
08-14-2015
|
|
fiogf49gjkf0d | anthony wrote: | | Personally, I am
probably going to do a line array like Romy's using the same drivers but twice
as many...so eight each side. I have the drivers. This
configuration should give somewhere around 109dB @ 20Hz from 15w in room with
cone movement of about +-3.5mm. |
|
The absolute amount of cone movement is irrelevant
as it does not take in consideration many other factors: like type of suspension,
the ay how driver deal with inertia, the type of magnetic system, the effective
frequency, the positioning of VC to the gap, the way how code attached to the suspension,
the driver “loading” into the box, the way how the driver is damped by the amp
and many other parameters. In some cases a few mm is too much and in some case
a few cm is fine.
| anthony wrote: | | If you have a
regular rectangular room, you may be tempted to try a Double Bass Array. |
|
I never heard this concept and after looking online I need to say that it does
look very slick and very smart. There is a problem with DBA however. The DBA is
a very nice way to deal with the room modes and I presume that if the DBA is properly
implemented then it will be no room problem. However, it says nothing about the
quality of bass itself. In my estimation the complexity of making a playback to
have good bass are 70-80 about bass structure and the rest is about the room response.
Room might have horrible nodes but ironically that might be corrected by bass
output if you run narrow octave bass module. However, if the bass has wrong structure
then there is nothing there might be done. An excessive bass volume is
something that a person get accustomed and mind tunes it off and correct it eventually.
The wrong bass structure is not brain-correctable and will be always there as
some kind of alien force.
|
|
|
|
Posted by anthony on
08-14-2015
|
|
fiogf49gjkf0d | Romy the Cat wrote: |
I never heard this concept and after looking online I need to say that it does
look very slick and very smart. There is a problem with DBA however. The DBA is
a very nice way to deal with the room modes and I presume that if the DBA is properly
implemented then it will be no room problem. However, it says nothing about the
quality of bass itself. In my estimation the complexity of making a playback to
have good bass are 70-80 about bass structure and the rest is about the room response.
Room might have horrible nodes but ironically that might be corrected by bass
output if you run narrow octave bass module. However, if the bass has wrong structure
then there is nothing there might be done. An excessive bass volume is
something that a person get accustomed and mind tunes it off and correct it eventually.
The wrong bass structure is not brain-correctable and will be always there as
some kind of alien force. |
|
This is my main issue with DBA as well. It looks promising on paper but has been taken up mainly by home cinema people as far as I can tell and there has not been a discussion about the quality of the bass. Plus I have never heard it or even know where there would be one to go and listen to in a two channel setup. It won't work in my room so I won't be trying it.
|
|
|
|
Posted by scooter on
08-14-2015
|
|
fiogf49gjkf0d I never heard of DBA but it looks to be an interesting approach. It looks like DSP is used as a cancellation device at the back wall; maybe in this role, it does not have such a bad impact on sound. But who really knows how this will sound without trying.
The DBA approach is so different, so cheap and so easy to implement (at least at a basic level), that if I had an appropriately sized room, I would like test it out for fun and learning purposes. No expectations.
My gut tells me this would be just another (radical) experiment destined for the bin, but honestly I have no idea how this might work in the real world.
|
|
|
|
Posted by Murataltuev on
08-15-2015
|
|
fiogf49gjkf0d | anthony wrote: |
Regarding the other questions you asked, I am basically duplicating Romy's system as my starting point with horns. |
|
Yes, as you can see my approach is also very similar.
Two major differences: TH and Fundamental frequency range.We already discussed TH vs line array and there are pros and cons of both ways, but also interesting to discuss fundamental frequency range.I still belive that this range should be as wide as possible - something like 200-300 to 3000-4000Hz.I've found interesting thread here about experiments with this range, but they ended up with nothing.As I understand - because of the size of horn capable to do 300Hz.
Romy, do you still plan some experiements in this field ?Interesting to try GOTO SG 555PS or JBL2490 to see if 200-500Hz range is better to get from compression driver in proper horn.
I got some promising measurements of Line Magnetic 555W (WE 555 clone) and thinking why not to try it ?
|
|