Greg - I wasn't aware of that concern - would it perhaps have referred to higher compression that could be expected a front loaded design? I do know that Alpairs have been used in several BLH /manifold enclosure designs without any reported damage - at least when driven at moderate SP levels.
I would say that that comment is slightly uncalled for. Any competent designer will simulate the excursion demands placed upon a driver to make sure that its capabilities aren't exceeded at a target SPL. A lot of the time simulations are done before any driver has been purchased because sometimes you can remove a selection of drivers from the potential 'list' because they wont do what you want them to do.
You are quite right in saying that perhaps the minority group of people who damage the drivers have a different definition of what reasonable is and hence they damage the drivers. But to say that the designers wanting an Xmax/Xmech figure so as to be able to design a loudspeaker without some compromises and also to ensure that they aren't going to run into problems on large music transients in another thing entirely. Linkwitz especially pays careful attention to xmax in his designs to make sure that the drivers aren't pushed beyond their limits at peak levels - as far as I remember his opinion of this level is 105dB which he's arrived at, knowing him, from measured peak transients of orchestral concerts.
I do not listen at high SPLs myself, nominally probably around 80-85dB or less and the system will happily reproduce 105dB and that is down to around 20hz. I pay very careful attention to the xmax capabilities of the drivers that I use. In every design I play around and change the loading depending on what the end system target is going to be. For example in some systems I know they are not going to be played loud so I maximise the bass extension potential by applying a small degree of bass lift via a LT. In some other systems I again change the box design to maximise the low frequency extension at the cost of maximum SPL capability. I can only do this when I know an accurate Xmax or safe excursion parameter for a driver.
In my main system I've over designed it on purpose and in this I pay very careful attention to the power amplifier requirements, the driver excursion, the size of the box and the maximum peak SPL capabilities I want it to be able to have. As I want smaller rather then larger sealed boxes for some of the drivers I have to balance everything very carefully as efficiency drops off as the box is made smaller. Once again to do this and to maximise the systems potential around my requirements, the xmax parameter is absolutely critical.
The P10 Castle cabinets appear to have two drivers in, one on the top one on the front, the internal volume is about 25 litres and the recommended port is tuned to around 52Hz. This seems a rather high tuning, you could of course have lowered the tuning yourself.
Either way are attached simulations for the EL70 in the castle cabinet. C1 is how the P10 PDF details the box. C2 is with the port length increased to 22cm. This lowers the tuning down to 40hz, gives you significantly more output slightly lower down in frequency, lowers the overall excursion down low at the cost of slightly higher excursion a bit higher up and slightly higher excursion at sub frequencies. As you're applying EQ down low anyway, this could theoretically reduce the necessary EQ and hence lower excursion, but also lower the excursion simply because of the lower box tuning. The lower box tuning also comes with slightly lower group delay and a better characteristic in its impulse response (it 'rings' slightly less) so could possibly give you better bass.
I noticed you built removable bottoms into your design, this would theoretically make it very easy to try out different port tunings. You may have already done this however. But if you haven't it might be worth trying it out. Also the gentle roll off below 100hz towards the lower port tuning is sometimes smoothed out by room gain and can help with integration into some rooms (normally smaller ones where room gain is greater).
If your cabinets aren't that heavily stuffed then you may need to increase the port length to 24cm for the 40hz tuning.
Attachments
5th Element,
I can't argue with your position regarding Xmax being a critical parameter for you as a designer of loudspeakers. I know very little about this subject and you obviously know a considerable amount. Perhaps your goal is to have all manufacturers agree on exact definitions of "Xmax" and "Xmech" and also to agree to provide those specifications for all of their drivers, based on standardized measuring criteria. I can see how this would benefit a designer such as yourself, however it might be a bit optimistic to think that Mark can set these definitions and standards himself and the rest will follow. From what I understand in this discussion so far, technically, the stated Xmax on MA drivers is accurate; in other words, not intentionally misleading. It seems that the issue is that the Xmech is not defined, and that the Xmech < Xmax, which results in mechanical damage before Xmax is reached. However, if a not-so-experienced designer (someone less experienced than yourself) mistakenly uses the Xmax parameter to anticipate/estimate the physical (mechanical) excursion, then problems arise. Is this Mark's fault, or is it due to a combination of inexperience and/or variable definitions and/or interpretations of "Xmax"?
I cannot say either way, as I have already stated that I know very little about the subject. I would, however, like to sincerely thank you for the bulk of your last post. You have really gone the extra mile there to show me the alternative port tuning, prepare the graphs, and explain it all to me. I really, really appreciate that and I will definitely be trying that alternative tuning. I am pretty sure I have enough Baltic Birch to make 2 new bottom panels and sourcing some pipe should be pretty easy.
Cheers!
I can't argue with your position regarding Xmax being a critical parameter for you as a designer of loudspeakers. I know very little about this subject and you obviously know a considerable amount. Perhaps your goal is to have all manufacturers agree on exact definitions of "Xmax" and "Xmech" and also to agree to provide those specifications for all of their drivers, based on standardized measuring criteria. I can see how this would benefit a designer such as yourself, however it might be a bit optimistic to think that Mark can set these definitions and standards himself and the rest will follow. From what I understand in this discussion so far, technically, the stated Xmax on MA drivers is accurate; in other words, not intentionally misleading. It seems that the issue is that the Xmech is not defined, and that the Xmech < Xmax, which results in mechanical damage before Xmax is reached. However, if a not-so-experienced designer (someone less experienced than yourself) mistakenly uses the Xmax parameter to anticipate/estimate the physical (mechanical) excursion, then problems arise. Is this Mark's fault, or is it due to a combination of inexperience and/or variable definitions and/or interpretations of "Xmax"?
I cannot say either way, as I have already stated that I know very little about the subject. I would, however, like to sincerely thank you for the bulk of your last post. You have really gone the extra mile there to show me the alternative port tuning, prepare the graphs, and explain it all to me. I really, really appreciate that and I will definitely be trying that alternative tuning. I am pretty sure I have enough Baltic Birch to make 2 new bottom panels and sourcing some pipe should be pretty easy.
Cheers!
Last edited:
On the whole I'd say that they do tend to agree. Xmax is almost unanimously agreed on I think, and any time I see Xmech quoted it always tends to imply the maximum excursion that a driver can perform without causing damage. Maybe this differs a little from manufacturer to manufacturer, but the end result is often the same. Stay within Xmax and if you intend on leaving it sometimes make sure you steer clear Xmech.
I think most designers design entirely to the Xmax figure and manage their systems SPL requirements around remaining inside it. Sometimes with subs you can get away with exceeding it on explosions and stuff because sub drivers typically have a lot more excursion capability beyond Xmax (unless they are using some fancy motor technology) and the increased distortion because of exceeding xmax during an explosion is likely to go unnoticed, in fact it might be welcomed.
I've been using the same 4 Peerless XLS drivers for years now and I am sure when they first came out Peerless had specified that the Xmech was somewhere around twice Xmax. After their split from Tymphany, scanspeak decided to make their own version of it where, like the 830452 XLS, is quoted as having +-12.5mm of Xmax and Xmech is +-28mm. So there's plenty of extra room available for SPL peaks if you're willing to accept the greater distortion that occurs as you exceed xmax.
It would be nice if this happened, but its also not quite what I am asking for.
Exactly the motor designs do have the required amount of coil overhang as to equal the stated Xmax within the datasheets.
This is also correct and quite unusual, it is certainly the first time I have encountered it. Usually one can easily run to Xmax and use it as a safe operating area as you can usually exceed it by a little without problems. If Xmech isn't defined in these situations it doesn't matter because I'd say that most people typically design their systems to remain within Xmax 99% of the time. They know the occasional venture out of xmax is unlikely to cause any damage, providing it isn't too extreme, but nontheless Xmax is usually enough to be used as a common guide.
Marks drivers cannot apparently operate at Xmax without breaking though. Now lets say Mark had designed his drivers with half the xmax. Say they had around 2mm for the small diameter drivers. This would be more in line with other full range drivers, they'd have a larger xmax then some and also less then some but a decent average figure nontheless.
This would mean perhaps that Marks drivers could be operated at Xmax without any issues nor would there be any confusion. This would also mean that on the occasional dynamic peak that the driver would exceed xmax momentarily and the sound quality would badly degrade when doing so. Now as a design decision (and a good one at that I might add) Mark has decided to increase the xmax on all of his drivers so that on the occasional dynamic peak, xmax wont be exceeded. This means that the drivers remain linear and the sound quality doesn't noticeably degrade.
This isn't quite right. What I am saying is that even the most experienced loudspeaker designers wont know that that the Xmax parameter, as defined in Marks datasheets, is dangerous to the drivers health if operated at on a regular basis. It is in that respect ,unusual. Because this is unusual I think it requires clarification in some way and beyond what Mark writes in his datasheets that would be...
To an experienced designer this statement would automatically mean, as Art also mentioned earlier on, that if you are within Xmax you are okay. If you're exceeding Xmax, turn it down.
The trouble that Mark has encountered in the past, is that if he lists Xmech parameters, or say perhaps alters his statement to
Now something like this would be perfect for me because I could then use +-2mm as my guide when designing them into a complete system. But why Mark is against it is because many of the less experienced DIYers don't know what it means.
As this is a somewhat cautionary statement I would guess that a lot of DIYers would perhaps be worried/alarmed by it and have enquired as to what it means and this seems to be more trouble then it's worth.
I am saying that it would be better for Mark to include a statement such as that above and find another way to handle difficulties encountered with people asking what this +-2mm actually means. I can well understand his reluctance to provide such a figure in such a way, but due to the peculiar nature of his full range drivers with Xmech really being less then Xmax, I feel it requires some sort of clarification.
The simplest answer might just be, to leave the statement
In the datasheet for the less informed DIYers and simply add after the Xmax parameter, Xmech = +-2mm. To those not informed they will most likely just ignore it as a part of the parameters they don't need to look at or understand, because they are building a proven box design. But it would be perfect for those who actually know what it means. Having Xmech there would also serve as a way for Mark to say - Xmech is defined in the datasheets, so if you've damaged your drivers due to over driving them then it is your own fault. Technically it's their own fault even if Xmech isn't defined, but currently someone could turn around and say, I didn't exceed xmax though and they still broke. Now Mark would be flying in turbulent air because he doesn't actually state anywhere that his drivers can't be used continuously at their rated Xmax.
You are most welcome
I use LspCAD for my simulations, but you might be interested to know that WinISD pro is entirely free and will allow you do much the same.
LinearTeam
It is fairly intuitive to use too, if you click on 'new project' it provides prompts that should be enough to get you started.
Thank you GM,
Basically, a large amount of sims will be required to get the right tuning etc for durability. I'm pondering a 3 to 4" FR with a 6.5" woofer running a first order crossover at around 200Hz. My wife has Enya Watermark on her iPod and it's bass line drops to around 23-24Hz
Since the HT system has more controls than some aircraft, she prefers to use my desktop computer speakers for working around the house. Even foo-foo music has bass drops--have to design that reality in.
You are most welcome
I use LspCAD for my simulations, but you might be interested to know that WinISD pro is entirely free and will allow you do much the same.
LinearTeam
It is fairly intuitive to use too, if you click on 'new project' it provides prompts that should be enough to get you started.
I hate to go off topic, but I have exhausted my own knowledge and opinion on the thread topic, so let me just say that I have actually played with WinISD a bit before, and oddly enough I launched it just after reading your post and plugged in the t/s parameters of the EL70 and started playing. I just picked up some 3" ABS at Home Depot and I plan to do some more playing in WinISD and some experimentation. I realized that I can simply extend the existing port with a section of ABS and some duct tape for testing purposes, rather than making a new bottom panel, etc. If/when I have results to share, I'll start a new thread.
Thanks again.
I had an hour of free time and a pair of Pensil 10.2's. I loaded up U2's "Auctung Baby" track 1 "Zoo Station". I turned up the volume to 90dB @ 8 ft on a Radio Shack SPL meter and watched the excursion. This track has some good, solid bass in the low 40's, maybe high 30's if it is a 5-string. I estimate that the max excursion P-P was ~3mm! Track 3 "One" has some hammering kick drum and the max execution was ~5-6mm. What the hell are you guys doing to these drivers to cause mechanical failure???
I don't listen to much rock as many of you know. 90dB is too loud for me particularly with the lack of dynamics that is a signature of heavy rock. This is also into the permanent ear damage range. If I read Mark's comments above correctly, this is still within the acceptable range. Those are P-P readings, not one-way.
If you have to run higher SPL, then 1) you don't understand what single-driver is all about and 2) you don't need to spend your money on quality drivers because you will soon be too deaf to know the difference. I just acquired Fleetwood Mac's "Rumours" (I said I don't listen to much rock, didn't I?) Listen to track 3 "Never Going Back Again". THAT's what single-driver is all about.
Bob
I don't listen to much rock as many of you know. 90dB is too loud for me particularly with the lack of dynamics that is a signature of heavy rock. This is also into the permanent ear damage range. If I read Mark's comments above correctly, this is still within the acceptable range. Those are P-P readings, not one-way.
If you have to run higher SPL, then 1) you don't understand what single-driver is all about and 2) you don't need to spend your money on quality drivers because you will soon be too deaf to know the difference. I just acquired Fleetwood Mac's "Rumours" (I said I don't listen to much rock, didn't I?) Listen to track 3 "Never Going Back Again". THAT's what single-driver is all about.
Bob
Hi Bob,
Rumours is one of my favorites, bar none! I shredded my CHR-70 gen 1's on Dire Strait's second album (a very mellow album!) over a period of about a week, and it was at loud but very reasonable levels (my wife and I could still carry on a conversation).
But no regrets -- I loved them, and enjoyed them. They sounded fantastic, and they cost maybe $70 for the pair. Frankly, I've paid more and gotten less out of certain Fostexes.
In the end, they made a "knocking" or "clattering" which sounded like the VC was bottoming out (many times per second). As if you took a ruler, put it flat on the table, with half of it overhanging the table, then snapped it and let it vibrate.
I examined the cones and it seemed as if the dust caps (if that's what they really are) had pulled away, leaving a slight gap or seam, held on by several spot-welds. But maybe that seam was there in the beginning -- I hadn't checked originally.
But I am happy to pay for "R&D" and have absolutely no regrets whatsoever. Sure, I'll be more careful next time, but also, the driver will have evolved too. I like having more widebander options and I'm willing to pay! So no complaints here. They had a great (if short) life and they died doing what they loved.
EDIT: Actually it was Dire Straits' first album.
Rumours is one of my favorites, bar none! I shredded my CHR-70 gen 1's on Dire Strait's second album (a very mellow album!) over a period of about a week, and it was at loud but very reasonable levels (my wife and I could still carry on a conversation).
But no regrets -- I loved them, and enjoyed them. They sounded fantastic, and they cost maybe $70 for the pair. Frankly, I've paid more and gotten less out of certain Fostexes.
In the end, they made a "knocking" or "clattering" which sounded like the VC was bottoming out (many times per second). As if you took a ruler, put it flat on the table, with half of it overhanging the table, then snapped it and let it vibrate.
I examined the cones and it seemed as if the dust caps (if that's what they really are) had pulled away, leaving a slight gap or seam, held on by several spot-welds. But maybe that seam was there in the beginning -- I hadn't checked originally.
But I am happy to pay for "R&D" and have absolutely no regrets whatsoever. Sure, I'll be more careful next time, but also, the driver will have evolved too. I like having more widebander options and I'm willing to pay! So no complaints here. They had a great (if short) life and they died doing what they loved.
EDIT: Actually it was Dire Straits' first album.
Last edited:
Hi zman01,
The amp was a 2A3, so not many watts. I believe the problem is just what others have said, that (all drivers must be) excursion-limited. That would mainly be a function of the cab, the SPL, the source material.
If I had to guess, I would say that I'm used to a certain volume level (for what little rock I ever listen to, which is almost none). And my usual drivers are much more efficient, which is why I haven't blown any other drivers But I would absolutely get another pair! I'd just be more careful, e.g., with the cab and as Bob said, with filtering below whatever point the driver (in the particular cab) unloads or exceeds whatever excursion limit.
A friend has the A7's in OB and has no problem but listens to classical at moderate levels (I believe). So it's definitely a function of the user, the cab, the source, the SPL. Not really damping or amp damping factor in my not-so-experienced view.
The amp was a 2A3, so not many watts. I believe the problem is just what others have said, that (all drivers must be) excursion-limited. That would mainly be a function of the cab, the SPL, the source material.
If I had to guess, I would say that I'm used to a certain volume level (for what little rock I ever listen to, which is almost none). And my usual drivers are much more efficient, which is why I haven't blown any other drivers
But I would absolutely get another pair! I'd just be more careful, e.g., with the cab and as Bob said, with filtering below whatever point the driver (in the particular cab) unloads or exceeds whatever excursion limit.A friend has the A7's in OB and has no problem but listens to classical at moderate levels (I believe). So it's definitely a function of the user, the cab, the source, the SPL. Not really damping or amp damping factor in my not-so-experienced view.
I don't recall, but no modern day 'FR' driver I'm aware of except durable doped paper PA designs can stand up to a compression loaded FL horn without introducing too much distortion. Better to FLH load a true co-ax or tri-ax or only use over a narrow BW.
At the time, I'd yet to see any MA drivers 'in the flesh', so concluded that since Mark had worked at Jordan [another brand that's not recommended for any type of horn loading AFAIK] that the design was similar enough to not be loaded down more than with a TL.
Anyway, in retrospect, it must have only referred to the more dainty, Jordan-like diaphragm models since the EL70s I abused right out of the box finding their usable excursion limits [including inadvertently bottoming them out several times due to not being able to discern the difference between Xmax or even Xsus before reaching Xmech] with no formal break-in are apparently none the worse for wear from this and their ~ 8 hrs/day/1 year of playing jazz, movie soundtrack, live recorded CDs and DVDs at max daytime neighbor acceptable apt. SPL levels [~81 dB avg./~101 dB peaks/m], so wouldn't think twice about BLH loading them.
Still haven't been exposed to the 'metal' drivers, but we could 'kill two birds.....' by doubling Vas to both reduce box loading pressure and get the increased efficiency [3 dB] that some folks want, though in existing small cabs the bass would of course be rolled off more and larger cabs will still run out of cone travel at the same SPL, so the extra efficiency would only be for the mid-bass and up BW.
Early on, I pointed out on one of the early threads how much better MLTL sims looked with doubled Vas [in the form of a requested dual driver alignment] and the Pencil designs probably wouldn't need as much stuffing either, but it fell on ‘deaf ears’ for whatever reason IIRC.
GM
You're welcome!
Not really, any program that graphs excursion/power limited SPL response [I use BoxPlot 3.07] can dial it in quickly, which WinISD Pro alpha does with its scrolling fields feature. Just highlight the field and use the arrow keys, though best to start with a very large vent area and fine tune it once an alignment is chosen otherwise the response will flat-line and the tuning default to zero if vent length falls to zero during scrolling.
Don't know of any 6.5" I trust to do even 32 Hz except in a huge BLH, so recommend some form of high pass EQ to protect them. That, or use a much larger woofer.................
FWIW, awhile back a forum member beat me to posting about bolting a large 'sub' woofer to an OB attached to the back legs of his computer table after the '50s era EV ~ dipole console designs, so another option to consider and probably no more expensive and maybe even cheaper than dual 6.5"/whatever drivers. Just use either a dual VC to feed both channels into it or a L-R summing ckt. if single VC. Unfortunately, I lost the link in a computer/back-up crash, but maybe someone else saved it.
Frankly, I'm surprised this idea didn't 'catch fire' considering all the computer/min-monitor systems this forum generates, but AFAIK his is the only one done to date.
GM
Without doing a sim, I'm not sure if it could be reduced to just lining the top, back and at least one side down to below the driver as I always 'shoot' for in my designs, but doubling Vas [or drivers] doubles the cab's net Vb requirements to have the same response, so the stuffing density shouldn't need to be as high in such an acoustically oversize cab it currently is.
GM
These must be some seriously compressed recordings like I recently got from Wal-Mart's $5 bin [~5 dB max of dynamic headroom]. Fine for car, computer monitors, DJ, but worthless to my ears in a HIFI playback system for anything other than Muzak and/or talk radio duty. For sure, if I'd turned up the EL70s to this average SPL with Buzz's CDs, DVDs, non optimum [oversize] MLTL I'd of needed a catcher's mitt to grab the flying VCs the first time Eric Bromberg hit a bass string on 'Wood'.
GM
GM,
What you are stating is pretty much what I've been experiencing over the past few weeks. I've had little success stuffing TLs or TL variants (if a Pensil can be called as such). After building the Half Towers I tried with lining top, back wall, one side only and bingo - had a great sounding speaker straightaway.
I am trying to optimize my Super Pensil 12s (even bigger than the Pensil 12), and finding that with only lining the top, area near drivers (15" from top, both sides and back), the sound is not too different vs when the cab is well stuffed. I feel there is a hump or two in the LF only; please note lining is limited only to the top section.
Any suggestions where I can go with this?
-Zia
It may be slightly off topic but it does show how you can affect and reduce cone excursion by a little in the lower frequencies usually found in music, so is relevant to a point in this thread.
Adding in an extra section of pipe with duct tape sounds about right, certainly a method I would agree with an encourage
Bob I think it very much depends on your cabinet and the music you play. Some stuff has wicked bass content in it and even mastering issues where infrasonic stuff ends up coming out of the compressor/ processing. The Alpair 10 though is also nominally a 6" drive unit and the extra cone area really helps with keeping things under control. I'd imagine that most of the failures come from the smaller drive units. The pencils must use some sort of special alignment that I can't simulate easily. I tried looking up the 'MUK worksheets' but didn't find anything. In a standard large ported alignment though the alpair 10 will have its excursion kept in check down to a bit below 30hz, so your brick wall filter will essentially offer 'bullet proof' protection from really low stuff.
Any high aspect ratio cab where there's a physically wide enough gap between driver/vent to impact [shorten] the vent's length WRT a golden or acoustic ratio alignment of the same tuning [Fb], then it qualifies as a ML-TL. As such, I imagine all the different 'Pensil' designs qualify.
Again, without simming, I'm not all that familiar with these designs, just that low Vas, Qms drivers need considerably more net Vb in a MLTL alignment to 'breathe' than the historical norm due to their inherent stiffness and mechanical resistance to low power input, i.e. properties ideal for horn loading if they had the diaphragm construction/rigidity to handle it [assuming it's true that they aren't designed for it] and why the acoustically tiny, ~aperiodic [to mimic horn damping] Fonkens work so well.
Such alignments typically have an [under-damped] audible dip in its LF response that would indeed cause a double hump sounding LF response and short of a second driver or EQ to fill it in, the only other option that comes to mind is what I do, critically damp [~transient perfect] the vent’s peaking [‘ringing’] same as the pioneers of audio championed and accept its obvious deep LF roll off [which is what a properly fully stuffed alignment does in a brute force way, but also taking away way too much of the speaker's 'life' in the process for me].
GM
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.