AndrewT said:
Okkayyyyy......
I'll google Wilson when I get the time. Being overloaded with links here 🙂
Why the preference for 16 ohm drivers?
My main priority at the moment (besides a spectrun analyzer design) is to finish the construction of my K800AB amp which delivers 1600W into 0.5 ohms (800W into 1 ohm continuous, class A bias for 200W into 4 ohms).
No problem driving a difficult load here.
Cheers,
Glen
Hi,
paired 16ohm drivers give an 8ohm effective impedance.
All/most amplifiers perform better into higher impedance loads.
Using 8ohms drivers will give a 4ohm speaker and if you use 4ohm bass drivers to compensate for lower sensitivity you will end up with a speaker that is 2ohm to 4ohm.
I would much rather have 4 to 8ohm for my speakers than 2 to 4ohm.
paired 16ohm drivers give an 8ohm effective impedance.
All/most amplifiers perform better into higher impedance loads.
Using 8ohms drivers will give a 4ohm speaker and if you use 4ohm bass drivers to compensate for lower sensitivity you will end up with a speaker that is 2ohm to 4ohm.
I would much rather have 4 to 8ohm for my speakers than 2 to 4ohm.
G.Kleinschmidt said:
Hi,
This is only of real interest in the bass, where you will have 4ohms
and 2 parallel drivers, setting your nominal senstivity, which BTW
for full BSC = the sensitivity of one of your bass drivers.
For the rest of the range being 4 ohm and two drivers is pointless.
Single 8 ohm drivers with the right senstivity will be fine. The fact
you are not drawing full power midrange / treble is irrelevant.
It does not matter as long as the senstivity matches the bass end.
FWIW the c/o components in a 4-way for the lowest c/o point
will be hideously expensive, and very easy to get very wrong.
(Why most manafacturers go for active bass 4-ways)
IMO you should consider a "standard" 3.5 way, nearly 4-way.
If you are not a volume fiend then high power suggests a low
efficiency design, which leads to very good bass extension
for the size of the speaker cabinets - basic physics.
Size vs. efficiency vs extension is always a balance for the bass.
Is not for the rest of the range, which needs to match SPL levels,
not power handling levels or drawn power.
🙂/sreten.
Re: Re: Parallel driver interaction in a 4-way system.
3.5way with 12" woofers, I would like to hear more about how you will do that
I understand that the use of multiple drivers would be fore better powerhandling...well, fore midbass I personally would use smallish PRO drivers which can be found with moderate sensitivity...B&C, AE etc.
But with "punkrk" in mind I think I am a bit reluctant with this
Its wrong to say its impossible, but with what has been suggested until now I think it will be very difficult
You cant be explained such matters but will have to completely understand how it works...in other words, "if you have to ask, then you cant do it"
http://www.geocities.com/cc00541/Statements.html
Add a subwoofer and you have your 4way 😀
Personally I would look closer at Gedlees Summa
sreten said:
3.5way with 12" woofers, I would like to hear more about how you will do that
I understand that the use of multiple drivers would be fore better powerhandling...well, fore midbass I personally would use smallish PRO drivers which can be found with moderate sensitivity...B&C, AE etc.
But with "punkrk" in mind I think I am a bit reluctant with this
Its wrong to say its impossible, but with what has been suggested until now I think it will be very difficult
You cant be explained such matters but will have to completely understand how it works...in other words, "if you have to ask, then you cant do it"
http://www.geocities.com/cc00541/Statements.html
Add a subwoofer and you have your 4way 😀
Personally I would look closer at Gedlees Summa
Okkkaaayyyyyy......
Let me put my load requirements another way. My amp will clip at not much above 40V peak. It needs a 4 ohm load minumum just to get pushed out of its 200W rms class A operating region.
Peak current is limited to over 10 times that at 120A and I want to make use of a good proportion of it.
Cheers,
Glen
EDIT:
BTW, I am all ears to anyone in the know who can suggest how to make a decent 3 or 4 way HiFi speaker with a nominal impedance of around 2 ohms (dips to as low as 0.5 ohms are OK) and a ~400Wrms power handling capability.
Let me put my load requirements another way. My amp will clip at not much above 40V peak. It needs a 4 ohm load minumum just to get pushed out of its 200W rms class A operating region.
Peak current is limited to over 10 times that at 120A and I want to make use of a good proportion of it.
Cheers,
Glen
EDIT:
BTW, I am all ears to anyone in the know who can suggest how to make a decent 3 or 4 way HiFi speaker with a nominal impedance of around 2 ohms (dips to as low as 0.5 ohms are OK) and a ~400Wrms power handling capability.
this seems like approaching a brief with the mentality of a Car Audio eejit.G.Kleinschmidt said:
You are better than that.
AndrewT said:
Great, but I'm starting to care a little less about what you have to say.
Input from anyone who can concentrate solely on the specification at hand and suggest possible approaches for meeting the engineering challenge are still invited.
Cheers,
Glen
Hmmm ........
Seems the cart and horse are decidedly the wrong way round.
Designing a speaker to exercise the capabilities of an amplifier seems
to me a fairly pointless endeavour, but it is not particularly difficult.
Try finding some Apogee Scintilla's, the 1 ohm nominal version.
🙂/sreten.
Seems the cart and horse are decidedly the wrong way round.
Designing a speaker to exercise the capabilities of an amplifier seems
to me a fairly pointless endeavour, but it is not particularly difficult.
Try finding some Apogee Scintilla's, the 1 ohm nominal version.
🙂/sreten.
http://www.gedlee.com/Summa.htm
Probably the best and cheapest way to avoid clipping at loud levels with low distortion
But there is nothing fancy about its look, which I like but many wont
But I dont understand why you wanna push your amp out of its highish classA idling, when you can build something that makes it stay in classA...if thats what you are suggesting, or else I dont understand, unless you like clipping amps
Probably the best and cheapest way to avoid clipping at loud levels with low distortion
But there is nothing fancy about its look, which I like but many wont
But I dont understand why you wanna push your amp out of its highish classA idling, when you can build something that makes it stay in classA...if thats what you are suggesting, or else I dont understand, unless you like clipping amps
sreten said:
tinitus said:
The amp can deliver 1600W into 0.5 ohms in class AB without clipping.
20 pairs of BJT output devices were required to meet the dissipation requirements of the 200W into 4 ohms classs A specification. A byproduct of this is an output stage capable of delivering huge output current in class AB into low impedances.
Cheers,
Glen
Hi Glen
You are underestimating the complexity of speaker systems. I live in both worlds (electronics and acoustics) and I can tell that getting a 2-way system to perform right is already a more complex task than the development of any linear amplifier, particularly understanding crossover design and acoustical phenomenta like comb filtering, lobing, diffraction, baffle step, spatial loading, physical resonances or just natural speaker driver response.
You should not attempt to start with an unreasonably high complexity. The highest complexity at which you should aim is probably dual bass, midrange and tweeter, and this is already going to result in many headaches.
Also, regarding crossovers, it's the acoustical response what should be considered and 1st order acoustical crossovers doesn't really exist in practice. Furthrermore, 1st order filters result in very poor power handling and high distortion. Speaker drivers are bandpass devices with inherent 2nd order roll-offs at both ends. Thus the minimum acoustical crossover slopes are 18dB/oct (1st order filters) unless unrealistic overlap is employed (and real drivers are not that good). For proper power handling you should aim at 12dB/oct filters at least, resulting in a 24dB/oct arrangement (12dB from the crossover plus 12dB from the driver at each end, LR24 recommended).
Another thing that you learn once you start to design speakers is that, in an amplifier with 40V output, 120A current capability is completely pointless and not to be taked advantage of by any speaker, 14A is already completely adequate. It's the wrong design criteria, you can only design 40V 120A amplifiers when you don't understand speakers. Standard impedances are between 4 and 16 ohms, not between 0.4 and 1.6 ohms. Current is always limited by voice coil resistance (that increases by >25% due to heating when playing real loud) plus other phenomena (nonsense myths apart).
The numbers must be actually the opposite in order for it to be real loud: 120V and 40A. Speakers are "voltage hungry", not "current hungry". You are never going to get very loud with 40V, particularly if you go for low efficiency drivers (the ones with the wider frequency response thus allowing 1st order filters and great overlap). On the other hand, higher efficiency drivers (and horns) have much narrower bandwidth and require higher order filters (like 24dB/oct acoustical as mentioned previously). This results in a compromise as usual.
Finally, using more than one driver for mid-high and high frequencies is strongly not recommended due to comb filtering and strong lobing, unless you go for a line array. You should read about and understand these phenomena first.
I hope most of your misconceptions to fall under its own weight once you start to face the facts about speaker design.
You are underestimating the complexity of speaker systems. I live in both worlds (electronics and acoustics) and I can tell that getting a 2-way system to perform right is already a more complex task than the development of any linear amplifier, particularly understanding crossover design and acoustical phenomenta like comb filtering, lobing, diffraction, baffle step, spatial loading, physical resonances or just natural speaker driver response.
You should not attempt to start with an unreasonably high complexity. The highest complexity at which you should aim is probably dual bass, midrange and tweeter, and this is already going to result in many headaches.
Also, regarding crossovers, it's the acoustical response what should be considered and 1st order acoustical crossovers doesn't really exist in practice. Furthrermore, 1st order filters result in very poor power handling and high distortion. Speaker drivers are bandpass devices with inherent 2nd order roll-offs at both ends. Thus the minimum acoustical crossover slopes are 18dB/oct (1st order filters) unless unrealistic overlap is employed (and real drivers are not that good). For proper power handling you should aim at 12dB/oct filters at least, resulting in a 24dB/oct arrangement (12dB from the crossover plus 12dB from the driver at each end, LR24 recommended).
Another thing that you learn once you start to design speakers is that, in an amplifier with 40V output, 120A current capability is completely pointless and not to be taked advantage of by any speaker, 14A is already completely adequate. It's the wrong design criteria, you can only design 40V 120A amplifiers when you don't understand speakers. Standard impedances are between 4 and 16 ohms, not between 0.4 and 1.6 ohms. Current is always limited by voice coil resistance (that increases by >25% due to heating when playing real loud) plus other phenomena (nonsense myths apart).
The numbers must be actually the opposite in order for it to be real loud: 120V and 40A. Speakers are "voltage hungry", not "current hungry". You are never going to get very loud with 40V, particularly if you go for low efficiency drivers (the ones with the wider frequency response thus allowing 1st order filters and great overlap). On the other hand, higher efficiency drivers (and horns) have much narrower bandwidth and require higher order filters (like 24dB/oct acoustical as mentioned previously). This results in a compromise as usual.
Finally, using more than one driver for mid-high and high frequencies is strongly not recommended due to comb filtering and strong lobing, unless you go for a line array. You should read about and understand these phenomena first.
I hope most of your misconceptions to fall under its own weight once you start to face the facts about speaker design.
G.Kleinschmidt said:
Hi,
A pointless byproduct to exercise for a class A amplifier.
There is no point in running Class A into low impedance loads.
You might as well design for a 32 ohm load and use 8 ohms.
🙂/sreten.
tinitus said:
Then you can "vaste" your money on competent subs, which will be hard enough
You seem to completely lack all the basic skills needed to pull off a 4way project like that, or even a 3way fore that matter
You have already got the advise from competent people
If you really want to fool around with a 4way, do like Eva suggest, 1 driver fore each passband and if you fancy 6db so much you will need wideband drivers or drivers that roll off smoothly by themselves, but it will most likely be 12db anyway...and be prepared to spend the next 5 years to make it work, if you are lucky
😱
Let's cover a few minor points -
First -WOW- , I actually agree with Sreten. As much as I respect him, even he will admit our agreeing is a rare occurrence.
Second, you have a slight misinterpretation -
Though there are some technical complexities that I am ignoring for the moment. Speakers are Voltage or Signal driven device; we feed them Voltage/Signal and they consume power. Restated, they are not driven by power, they consume power.
If I can add a slightly flawed analogy, my car is not driven by gasoline; it is driven by an engine that consumes gasoline. (It's the best I can do on short notice.)
So, my point is that you can drop the impedance of the load as low as you want and all you do is consume more power. BUT, and this is a big BUT, the ceiling doesn't move (if you are lucky). You say yours is 40 volts, well it is 40 volts regardless of whether you are consuming 200watt or 1000watts.
Now, as to your design concept, even considering your original post, is still pretty vague, though it is becoming more clear.
I don't think you necessarily want more drivers, though that is a valid solution, I think what you want is better drivers. As some one already suggested, look a Pro quality drivers (not cheap). They will handle enormous power compared to standard stereo speakers.
Next, despite your amp being able to handle very low loads, you still need to consider impedance of the final system. There is a reason why the standard speaker load is between 4 ohms and 16 ohms.
Remember consuming more power to no end is kind of pointless. Better to consume power wisely and get the best benefit.
Next, a characteristic of midrange and tweeters, though more so tweeters. At equal measured volume, high frequencies will naturally sound louder. Also consider that tweeters only have to make a tiny movement and move a tiny mass relative to woofer to create the same sound level. Consequently, you can have a much lower powered tweeter in your system and still have it hold up to your woofers.
Although pretty fast and loose, a guideline might be 100:50:25. For a 100 watt system use a 50 watt Mid and a 25 watt tweeter, and functionally, you will have a system fully capable of handling 100watts.
Consider the SPL (sound level) rating of tweeters vs woofers. If you use horns, tweeters can be well above 100db, keeping in mind that every 3db means doubling the power. And many non-horn tweeters approach 100db. Woofers are lucky if they break 90db.
ALso, for a given signal level, you need to dial-back (reduce the volume of or attenuate) the mids and tweeters to balance them with the woofer. So, despite the signal level being fed to the tweeter network, a much smaller signal actually reaches the tweeter. Sometimes 1/4th the power or less.
Next, let's do a reality check on your amp. What you really have is a massively stable 100 watt amp (to 8ohms) that has (pardon the expression) one hell-of-a power supply.
Again, remember, that simply dropping the impedance of the speaker system only consumes more power. There is no guarantee that it is actually delivering more acoustical power.
To go back to the automobile analogy, lower the impedance for no good purpose is like driving your car around all day in LOW gear. You have tons of power available and tons of power being consumed, but what purpose does it serve?
Also, consider this. Build a Sub with two 12" subwoofers in one cabinet. Then build a separate 3-way cabinet in some configuration that pleases you. A straight forward 3-speaker 3-way with a 10" woofer would kick some serious butt when placed along side the Subwoofers (in this case, I'm thinking two subwoofers, one for each front stereo speaker).
So, the 3-way could be WMT, or WWMT, or WWMTM.
I think both these design could be found on the internet or reasonably conceived. That would virtually be a wall of sound, and a lot easier to move around.
As to combining them, that is, connecting them to your amp. If at all possible, keep the impedance as high as possible. If you can get Pro Subs in 16 ohm and parallel them to make 8 ohms, that would be preferred. Then I think about the best you can do for a complex front tower would be 4 ohms (possibly 6 ohms, but doubtful) unless you went with a straight WMT configuration which could be at 8 ohms.
I actually think there is someone here who has a configuration similar to this, I've seen the photo in his signature.
So, now we have 8 ohms (Sub) in parallel with likely 4 ohms (tower) for a resulting 2.67 ohms. That would be bad for a normal amp, but if your amp is as good as you say, it should tolerate it.
Dividing the audio spectrum in a conventional 4-way IS more difficult than you think, and needs sharp drop offs, 12db or higher at each crossover. Building a standard 3-way, even if it is a multi-woofer tower design, is a far easier. Then just add a subwoofer to the 3-way and the windows will be rattling, plaster will be falling, and the neighbors will be calling the cops.
As a side note, most conventional 4-way systems are not Sub, bass, Mid, high; they are bass, mid, high, and super-high. See the JBL 4-ways.
So, again, expand your design goals. Yeah, I get it, you want huge 4-ways that will eat tons of power. But to what end, for what purpose? Home theater? Stereo Music? Public Address? DJ music for hire?
So, a 3-way with a separate Sub, or the 3.5-way with a Sub equivalent built-in, is going to be easier, and give better results than a full spectrum standard 4-way system.
Really, the difference is not much more than semantics, but from a practical perspective 3.5 trumps 4.0.
EDITED:
Also, you still haven't mentioned budget, unless I missed it, even if it is only fantasy ballpark budget. Budget matters. I can pull this off for $500 to $1000 or it can be done for $1000 to $5000; or if you prefer, $5000 to $10,000. But those are very different approaches with very different results.
That's enough for now.
Steve/bluewizard
First -WOW- , I actually agree with Sreten. As much as I respect him, even he will admit our agreeing is a rare occurrence.
Second, you have a slight misinterpretation -
Though there are some technical complexities that I am ignoring for the moment. Speakers are Voltage or Signal driven device; we feed them Voltage/Signal and they consume power. Restated, they are not driven by power, they consume power.
If I can add a slightly flawed analogy, my car is not driven by gasoline; it is driven by an engine that consumes gasoline. (It's the best I can do on short notice.)
So, my point is that you can drop the impedance of the load as low as you want and all you do is consume more power. BUT, and this is a big BUT, the ceiling doesn't move (if you are lucky). You say yours is 40 volts, well it is 40 volts regardless of whether you are consuming 200watt or 1000watts.
Now, as to your design concept, even considering your original post, is still pretty vague, though it is becoming more clear.
I don't think you necessarily want more drivers, though that is a valid solution, I think what you want is better drivers. As some one already suggested, look a Pro quality drivers (not cheap). They will handle enormous power compared to standard stereo speakers.
Next, despite your amp being able to handle very low loads, you still need to consider impedance of the final system. There is a reason why the standard speaker load is between 4 ohms and 16 ohms.
Remember consuming more power to no end is kind of pointless. Better to consume power wisely and get the best benefit.
Next, a characteristic of midrange and tweeters, though more so tweeters. At equal measured volume, high frequencies will naturally sound louder. Also consider that tweeters only have to make a tiny movement and move a tiny mass relative to woofer to create the same sound level. Consequently, you can have a much lower powered tweeter in your system and still have it hold up to your woofers.
Although pretty fast and loose, a guideline might be 100:50:25. For a 100 watt system use a 50 watt Mid and a 25 watt tweeter, and functionally, you will have a system fully capable of handling 100watts.
Consider the SPL (sound level) rating of tweeters vs woofers. If you use horns, tweeters can be well above 100db, keeping in mind that every 3db means doubling the power. And many non-horn tweeters approach 100db. Woofers are lucky if they break 90db.
ALso, for a given signal level, you need to dial-back (reduce the volume of or attenuate) the mids and tweeters to balance them with the woofer. So, despite the signal level being fed to the tweeter network, a much smaller signal actually reaches the tweeter. Sometimes 1/4th the power or less.
Next, let's do a reality check on your amp. What you really have is a massively stable 100 watt amp (to 8ohms) that has (pardon the expression) one hell-of-a power supply.
Again, remember, that simply dropping the impedance of the speaker system only consumes more power. There is no guarantee that it is actually delivering more acoustical power.
To go back to the automobile analogy, lower the impedance for no good purpose is like driving your car around all day in LOW gear. You have tons of power available and tons of power being consumed, but what purpose does it serve?
Also, consider this. Build a Sub with two 12" subwoofers in one cabinet. Then build a separate 3-way cabinet in some configuration that pleases you. A straight forward 3-speaker 3-way with a 10" woofer would kick some serious butt when placed along side the Subwoofers (in this case, I'm thinking two subwoofers, one for each front stereo speaker).
So, the 3-way could be WMT, or WWMT, or WWMTM.
I think both these design could be found on the internet or reasonably conceived. That would virtually be a wall of sound, and a lot easier to move around.
As to combining them, that is, connecting them to your amp. If at all possible, keep the impedance as high as possible. If you can get Pro Subs in 16 ohm and parallel them to make 8 ohms, that would be preferred. Then I think about the best you can do for a complex front tower would be 4 ohms (possibly 6 ohms, but doubtful) unless you went with a straight WMT configuration which could be at 8 ohms.
I actually think there is someone here who has a configuration similar to this, I've seen the photo in his signature.
So, now we have 8 ohms (Sub) in parallel with likely 4 ohms (tower) for a resulting 2.67 ohms. That would be bad for a normal amp, but if your amp is as good as you say, it should tolerate it.
Dividing the audio spectrum in a conventional 4-way IS more difficult than you think, and needs sharp drop offs, 12db or higher at each crossover. Building a standard 3-way, even if it is a multi-woofer tower design, is a far easier. Then just add a subwoofer to the 3-way and the windows will be rattling, plaster will be falling, and the neighbors will be calling the cops.
As a side note, most conventional 4-way systems are not Sub, bass, Mid, high; they are bass, mid, high, and super-high. See the JBL 4-ways.
So, again, expand your design goals. Yeah, I get it, you want huge 4-ways that will eat tons of power. But to what end, for what purpose? Home theater? Stereo Music? Public Address? DJ music for hire?
So, a 3-way with a separate Sub, or the 3.5-way with a Sub equivalent built-in, is going to be easier, and give better results than a full spectrum standard 4-way system.
Really, the difference is not much more than semantics, but from a practical perspective 3.5 trumps 4.0.
EDITED:
Also, you still haven't mentioned budget, unless I missed it, even if it is only fantasy ballpark budget. Budget matters. I can pull this off for $500 to $1000 or it can be done for $1000 to $5000; or if you prefer, $5000 to $10,000. But those are very different approaches with very different results.
That's enough for now.
Steve/bluewizard
I agree. Maybe use some of that big stash of O/P devices in a few smaller amps and tri/quad amp.sreten said:
CD's, Heil type drivers and a couple of ribbons are all that are 100dB or better. Domes typically get nowhere near it.BlueWizard said:
Pro LF drivers on the other hand, regularly exceed 90dB and are typically in the 94-98dB range> I'm not including dedicated sub drivers in this comment.
BlueWizard said:
That will be the day of miracles 🙂 but actually I have never tried a 4way, so I really dont know how hard it is, but even a 3way is close to impossible to get flawless
Well, a standard 3-way (or 3.5-way) build is easy because you can copy an existing design, and have all the detail worked out for you.
Building the Sub would be moderately easy (more or less) especially if it is not a Super_Sub intended to go down to extremely low levels.
Personally, I can't hear anything below 30hz, so a Sub or an 0.5 way that just re-enforce the sound in the 20hz to 100hz range would be MORE THAN enough for me.
Steve/bluewizard
Could you imagine the Values you would need to make a passive Xover @ .5 Ohms
I think your Z would go up from the DCR of all of the Giant inductors that would be needed.
I think your Z would go up from the DCR of all of the Giant inductors that would be needed.
All good intentions I am sure, but since when does a 3way with sub count as a 3.5way ?
Or is it just a new term I have missed
I would think that a 3.5way would consist of at least 4 drivers with double alike woofers crossed differently, one low and the other higher
unless the mid is a DOME, it really doesnt make sense...Zaph knows that, its pure logic
Tweeter would be more like supertweeter, which would make design a fair bit easier
In my book, a 3way with subs is to be considered a 4way
In my book the midbass would be a closed design 8-10", and the natural rolloff would be designed into the complete system...if only you could get plateamps with 12db slopes...ahh, I think Rythmik does do that one
Or is it just a new term I have missed
I would think that a 3.5way would consist of at least 4 drivers with double alike woofers crossed differently, one low and the other higher
unless the mid is a DOME, it really doesnt make sense...Zaph knows that, its pure logic
Tweeter would be more like supertweeter, which would make design a fair bit easier
In my book, a 3way with subs is to be considered a 4way
In my book the midbass would be a closed design 8-10", and the natural rolloff would be designed into the complete system...if only you could get plateamps with 12db slopes...ahh, I think Rythmik does do that one
tinitus said:
I think its an incorrect term.
the .5 refers to an LFE channel for surround sound.
Some times the .5 is used for bass management, so the LCR LS RS signal is crossed over in your Pre amp and sent to the LFE.
3.5 way is not a real speaker configuration.
- Status
- Not open for further replies.