G.Kleinschmidt said:
Hi,
Your woes are caused by trying to put the cart before the horse 😉.
Chucking money at the problem does really help either, a proper
understanding of engineering compromises is required, usually
trying to make an "uber-speaker" simply falls at the first hurdle.
To get back to basics you have a 100W / 8 ohms amplifier in
terms of voltage and consequently this with the midrange /
treble sensitivity will determine overall maximum levels.
Over-engineering the bass end to be capable of producing levels
simply not required in practise is not good design, and FWIW all
really "good" speakers are carefully designed and voiced above
100Hz, bass extension is just icing on a (very tasty) cake.
WinISDpro (or Basta) is useful for analysing maximum capabilities
in the bass, noting that for most music run at clipping level the
peak levels in the bass are typically 6dB to 10dB down, i.e. the
equivalent of 10W to 25W into 8 ohms.
That is until you bring out your bass "demonstration" tracks.
Building an "uber" speaker first time round is a bad idea, building
a very good value speaker optimised to the current requirements
IMHO would be far more rewarding and in the end not a lot worse.
I would consider a high value Watt / Puppy clone. I'd base it around
the Seas ER18RNX. Data is available at Zaphaudio. It will hit over
90dB/2.83V in the midrange. Choose high value tweeter / twin 4
ohm bass units to suit, all of which each should cost less than
the ER18RNX, if you want to keep it a well balanced design.
(You should have noticed by now drivers for sealed boxes are rare.
Play with WinISDpro :
http://www.linearteam.dk/default.aspx?pageid=winisdpro
It is not accurate in all respects, but it does show low tuned vented
boxes is the most sensible approach for the conbination of bass
extension with power handling capability at low frequencies.)
Basta! is a good investment for illustrating all the theory, its only
http://www.tolvan.com/basta/
limitation is it cannot accept frd and zma files. It is rather good
at nearly everything else, especially good at crossover / driver
interactions, attenuation affecting Qts, F box impedance etc. ......
You will still need the FRD tools to use real response data.
If you are not careful you will end up with a pointless 400W or
so subwoofer matched to a top end and amplifier that most
of the time will not remotely exercise the subs capabilities.
🙂/sreten.
http://www.humblehomemadehifi.com/Humble_Statement.html
Hi, One way of chucking money at it, 2 x 6 ohm bass, 🙂/sreten.
Hi, One way of chucking money at it, 2 x 6 ohm bass, 🙂/sreten.
Glen, it's clear to me that you're completely capable of doing as complex of a system as you'd like. You might want to make it modular, a head unit and one or more bass units to keep each box managable and allow replacement if issues arise. You might also want to make the baffle boards removable in case you want to try different driver layouts. Northcreek came up with a clever system for this, and I believe the PE boxes use it now also where you can read about it.
We've discussed the bass section mostly, here's some food for thought for the rest of the system. I particularly like the PSB Stratus Golds as a high value system:
http://www.kef.com/history/1970/model105.asp
http://www.stereophile.com/floorloudspeakers/704/
http://www.stereophile.com//loudspeakerreviews/96/
You've been tossing many ideas around which is good to see and understand the limitations of each. However, it's difficult to choose drivers without some firm requirements; I'm sure you'll get there as your preferences gel in your mind.
Pete B.
We've discussed the bass section mostly, here's some food for thought for the rest of the system. I particularly like the PSB Stratus Golds as a high value system:
http://www.kef.com/history/1970/model105.asp
http://www.stereophile.com/floorloudspeakers/704/
http://www.stereophile.com//loudspeakerreviews/96/
You've been tossing many ideas around which is good to see and understand the limitations of each. However, it's difficult to choose drivers without some firm requirements; I'm sure you'll get there as your preferences gel in your mind.
Pete B.
PB2 said:
Hi,
Sometimes less is more, and sometimes assumptions are incorrect.
Arbitrary complexity does not guarantee quality in the end result.
Quoting Zaph :
I'll repeat that trying to make the ultimate speaker first time out
is pretty much doomed to failure IMO. Trying to make it as good
as possible whilst being as simple as possible (this is complex)
and suitable for the requirements at moderate cost is IMO a
more sensible and rewarding way to approach the problem.
FWIW B&W's white papers on speakers are also very interesting :
http://bwmedia.keycast.com/download/Libraries/3/800s_l2_w0_h0_2.pdf?dm=633108447502000000
🙂/sreten.
sreten said:
😱
Do you actually follow any of my posts in detail, or just revert to negative trolling on autopilot?
Yes, good bass drivers for compact sealed boxes are indeed rare, but I have found one that works exceptionally well (the Scan-Speak unit) – see my previous posts.
The only down side of this driver (as with all such drivers) is low sensitivity (83dB SPL 2.83V/1m.
Also, I have absolutely no intention of “over designing” the bass and three of Scan-Speak drivers in parallel will give me a sensitivity to match (in terms of 1W SPL, max SPL and linear cone excursion – all simulated in WinISD and UniBox) the midrange and treble drivers I am currently looking at.
Fortunately, with my amplifier, I can get away with that.
Also, I fail to see what the heck is so complex about what I am proposing here – which is now pretty much just a standard 3-way speaker with 3 bass drivers of modest dimensions (9”) instead of just one larger unit, along with high quality midrange and treble drivers.
Hi Glen.
I think you fail to see the complexity because you have not done it before.
While it is easy to put drivers together and get better than "store bought" at the bottom end of the market i.e: what is avaliable in chain stores, getting everything to meld into a harmonius whole is IMHO very very difficult.
People are trying to help you here, to do what you want isn't easy and mistakes with quality drivers should be avoided if at all possible.
Obviously you feel able to tackle the project and have the funds to allow you to make mistakes and correct them.
please remember that ALL of the information and advice you have received has been provided by accomplished and very smart people, but as you should appreciate we all have opinions, different tastes in music and different listening environments that colour our interprtation of how to accomplish the end result.
If I had to give any advice at this stage it would be to " Suck it and see"
Build what you think would be acceptable and then listen to it for a month or two; on a variety of music, female jazz to hard rock and classical. And IF it all sounds good, clear and uncoloured, with sufficient dynamic range to reproduce a symphony orchestra at full throttle I'd say you got it right.
How-ever personal experience says that you may have to go through 3 to 6 iterations of the cross-over and driver combination before that happens.
Regards Ted
I think you fail to see the complexity because you have not done it before.
While it is easy to put drivers together and get better than "store bought" at the bottom end of the market i.e: what is avaliable in chain stores, getting everything to meld into a harmonius whole is IMHO very very difficult.
People are trying to help you here, to do what you want isn't easy and mistakes with quality drivers should be avoided if at all possible.
Obviously you feel able to tackle the project and have the funds to allow you to make mistakes and correct them.
please remember that ALL of the information and advice you have received has been provided by accomplished and very smart people, but as you should appreciate we all have opinions, different tastes in music and different listening environments that colour our interprtation of how to accomplish the end result.
If I had to give any advice at this stage it would be to " Suck it and see"
Build what you think would be acceptable and then listen to it for a month or two; on a variety of music, female jazz to hard rock and classical. And IF it all sounds good, clear and uncoloured, with sufficient dynamic range to reproduce a symphony orchestra at full throttle I'd say you got it right.
How-ever personal experience says that you may have to go through 3 to 6 iterations of the cross-over and driver combination before that happens.
Regards Ted
Moondog55 said:
Hi Ted
But a million other people have and the art of 3-way design is documented to death. I am not stepping into uncharted territory here.
I am however, in a relatively “unique” position with the amplifier I have under construction.
Take sretens third to last post for example. He suggests a Watt/Pupply clone - this is essentially what I am already proposing.
The only major difference in terms of drivers is that I intend to use three 4-ohm bass drivers instead of just two, in a sealed box instead of a vented box.
I have my reasons for wanting to stick with a sealed bass enclosure with a Qtc of 0.7.
I have found a 4-ohm bass driver that, in a 35L sealed enclosure, will linearly handle my amplifiers voltage swing limited power output and will do an f3 of 31Hz.
This driver makes it practical to build a +90dB bass end in a relatively modest size enclosure, and my amplifier is more than capable of driving three of them.
They are expensive, but I guess that is the price you have to pay for the performance (I have simmed a lot of drivers, and these are by far the best I've found).
Two cheap bass drivers in a vented enclosure sounds a compromise to me.
Cheers,
Glen
I agree with your sentiments,regarding sealed boxes. Just not all of us are in the position to buy the best drivers and I have to compromise.
bawling:
Looking forward to the build thread when you get started:
I don't think a million other people succeeded at first try though.
bawling:
Looking forward to the build thread when you get started:
I don't think a million other people succeeded at first try though.
PB2 said:
Hi Pete.
Some good suggestions here. The bass unit is the most straight forward part of the design.
I am actually considering building the bass enclosure as a separate unit. Basically, a rectangular sealed box of a little over 100L internal volume with three of those Scan-Speak 9” bass drivers, crossed over at 250Hz.
This will give me a healthily low F3 of approximately 30Hz with a Qtc of 0.7, approximately 92dB 2.83V/1m sensitivity.
Then I can freely experiment with the midrange / treble enclosure to mount on top.
One thing I am considering (although my simulations tell me it is probably a bit over the top) is an MMT top using two identical 8-ohm midrange drivers connected in series (mounted as close together as possible) to share the power handling and to keep the cone excursion really low.
Cheers,
Glen
three (smallish) bass drivers sounds workable.G.Kleinschmidt said:
Two large mid range drivers sounds wrong.
What excursion does an 8inch require to produce 120dB @ 1m from 400Hz and upwards? Why two?
If you were using a pair of 5 or 6inch mids, then I could see some logic in your experiments.
AndrewT said:
I said 8-ohm, not 8 inch. I won't go over 6" for the mid range. Prefer to stick with 5"
Cheers,
Glen
G.Kleinschmidt said:
6" might be a good compromise. As we've been referencing good commercial designs, it might be interesting to point out that all B&W models above about £2000 use a 6" surround-less mid. Clearly they don't expect much excursion to be taking place. They cross over 2nd order at 200hz, and 4khz to the tweeter. I like to mention this because I love B&W speakers, apart from the treble, which has always been poor on the ones I've heard. You're not limited to metal tweeters though 😎
Simon
G.Kleinschmidt said:
Hi,
Good point that you can apply to yourself. I will leave you to it.
By all means spend $2.5K on bass units with capabilities you
do not need in reality nearly all of the time. Your assumptions
are somewhat flawed, which I have pointed out, but ignore this.
G.Kleinschmidt said:
All good design is a compromise. To me there no such thing as
"uncompromising" design, just different ways of not doing it
intelligently, you are of course welcome to differ.
🙂/sreten.
G.Kleinschmidt said:
Glen,
As inflammatory as Streten's posts can come across, I think he's laid out some fantastic advice here recently.
WRT the number of drivers, why don't you keep things relatively straightforward and use a normal tweeter, a normal high quality 6" midrange and cross over to the bass where baffle step losses occur? At 200-300hz crossover it isn't going to be terrible to use several bass drivers.
Alternatively, a couple of 4-5" mids, and baffle step done there, then cross these at 150-200hz to the woofers. This solution might have better dispersion. A 3.5 way. This really appeals to my (albeit amateur) thinking.
I'd want to use a tweeter and midrange driver combination that someone like John Krutke or Tony Gee (prolific DIY speaker designers) has used before and copy their crossover exactly. The 0.5 section will be based on the baffle step frequency and will be less critical. The crossover to bass should be easy enough, as there'll be a considerable range of overlap of the twin 5" drivers and triple 8" drivers' abilities.
(I'm just reiterating some of the ideas that have come up in this thread so far, not really anything new). As a bonus I can't see why this approach would need any expensive drivers to offer high end sound. A further benefit to you would be a potentially demanding load for the amp, looking after all those crossover parts and 6 drivers per side.
Simon
SimontY said:
Hi
What you are suggesting here is pretty much along the lines of what I have planned (at this point in time).
The bass enclosure will be 0.45m wide, producing the baffle step at the bass-mid crossover of 250Hz.
I prefer to do the baffle step at the bass crossover because that keeps things simple and affords me ~6dB headroom in the SPL requirement for the mid-range driver and tweeter.
Maybe some of my words were a bit too strong, but I do not intend to use overly expensive mid-range and treble drivers, just sensible.
The 5" 90dB Seas mid range driver I am looking at will work nice from 250Hz to 3.5kHz (actually the recommended operating frequency range). The 3.5kHz tweeter crossover is high enough to avoid having to use some extravagant thingie that can go much lower (such as in a 2-way system).
I’ll probably go for one of the Morel 1” dome units that I can conveniently order from a local distributor.
Cheers,
Glen
sreten said:
😱
Deja Vu
Great, so now that you’ve run out of reasons why my current proposal “won’t work” the argument for the negative shifts to the “capabilities” I “do not need in reality nearly all of the time”, as if there was ever an argument there.
Also, I never said that I was attempting an "uncompromising" design - there are plenty of compromises with just about any design.
And what difference does it make to you if I spend $25, $250 or $2500 on the bass drivers?
Is that really what this is all about?
G.Kleinschmidt said:
Hmmm.......
😱 is certainly true.
Not much wrong with the Seas midrange driver for what is.
Never seen any tests that have a good thing to say about Morel tweeters.
Mating the above with > $2K worth of bass drivers ? Yawn .....
It makes no difference to me other than it ceases to be interesting.
I have not run out of "reasons", I've given up trying to help (almost).
You do not seem to want to understand my posts, so what is the point .....
🙂/sreten.
- Status
- Not open for further replies.