Mark Audio CHR-70 gen1 self destruction.

[weltersys]

Originally Posted by markaudio

5th's last post (No. 34) is correct re my design work (possibly a fellow driver designer/maker in the making?). The is precisely the point about this aspect of my design work, that you can use the larger excursion with linear results, on non-regualar LF loads. 5th, description of my work is better than mine.

One thing Matsubara san and me working on is more advanced rear suspensions (spiders). We are now starting to use a spider with spines. We've done allot of work to control the resistivity at the outer excursive limits. Currently, all the latest batches of Alpair 10. Gen's 2's are fitted with this up-grade (see pic). The latest design of this suspension type has additional stroke resistance from 7.2-mm onwards, but that doesn't mean its "open season" to use the Alps with this upgrade as woofs!

I'll have to think some more on how to provide data or advice that is practical to use, meaningful to the majority of end-users. Just coming up with X-Mech, X-Max, X-Zing, X-Ping, X-Thingy........whatever we call it won't hep the typical Diyer trying to figure the next project. This measurement issue has been around the block so many times already.

Cheers
Mark

Mark,

As 5th Element points out Xmax being less than Xmech problems aren't typically encountered.

In fact, your speakers are so far the first ones I have heard of with this problem in the 40 odd years spent doing enclosure design, listening to a wide variety of speakers and reading thousands of specification sheets.
That said, I am very familiar with stress cracks in aluminum diaphragms.

Rather than try to come up with a new parameter name for the stress cracks your cones are subject to at around half of Xmax, and some smaller fraction of the Xmech figure as normally defined, you could simply include a statement like this below the specifications (which should be included with the speaker):

“Caution: Although the Alpair 10 Xmax specification of 8.5 mm is accurate, and insures low distortion on occasional peaks, the cone may develop stress cracks if allowed to operate in this range continuously.
We recommend a maximum of 4 mm one way excursion, 8 mm (5/16 inch) peak to peak for long term use.”

That statement would satisfy those that actually bother reading the specification sheet, and clarify the unusual nature of the driver's mechanical limitations.

Art Welter

[markaudio]

Originally Posted by weltersys

Mark,

“Caution: Although the Alpair 10 Xmax specification of 8.5 mm is accurate, and insures low distortion on occasional peaks, the cone may develop stress cracks if allowed to operate in this range continuously.
We recommend a maximum of 4 mm one way excursion, 8 mm (5/16 inch) peak to peak for long term use.”

That statement would satisfy those that actually bother reading the specification sheet, and clarify the unusual nature of the driver's mechanical limitations.

Art Welter

Hi Art,
Did similar to your suggested statement a few months back. Floated it on the forum, got panned for being un-measurable (impractical) by numerous Diyers. That's when me and Matsubara san started re-developing rear suspensions as a possible technical solution/alternative (putting the brakes on near X max).

Thx
Mark.

[tuxedocivic]

Strange to me that DIYers find something like what Art is saying impractical. Xmax is a commonly used figure. When I design a can I use the ts params and check power input with xmax. Simple enough. Now knowing what I know, for your drivers I'll do the same, but with only 4mm. No one is saying use a pair of calipers on an active woofer. That's just silly. The users who don't understand this are probably not designing the cab.

[weltersys]

Originally Posted by markaudio

Hi Art,
Did similar to your suggested statement a few months back. Floated it on the forum, got panned for being un-measurable (impractical) by numerous Diyers. That's when me and Matsubara san started re-developing rear suspensions as a possible technical solution/alternative (putting the brakes on near X max).

Thx
Mark.

Putting the brakes on before the driver breaks is a good idea ;^).

I find it laughable that people think high tech accelerometers or laser devices are needed to measure excursion.

I have explained in dozens of posts that persistence of vision actually makes it quite easy to measure cone excursion, even easier if a paint or ink dot is put on the cone.
Anybody not blind can easily see the difference between 8 mm and 17 mm excursion on a 100 mm driver.

Anybody that dismisses measuring excursion as being un-measurable (impractical) has not attempted it, you can dismiss their dismissal.

You could print a little ruler directly on the spec sheet included with the driver with a peak to peak "safe zone" of up to 8mm in blue, and a red "danger zone", and even a black Xmech black zone.

Your PDF says:

"If you see the cone of the speakers making large movements, immediately turn the volume down."

If I had not been reading this thread, I would have assumed "large movements" to be movements in excess of Xmax.

You need to be far more specific in your specifications or reasonable, educated purchasers of your product (who have not read your warnings here) will continue to break them, even if you do manage to put the brakes on the suspension.

Art

[5th element]

Originally Posted by markaudio

(Sorry 5th, I know you mean well but......)
Re 5th's running in 1,2,3 method, Please DON'T follow this pattern. Its most important to use a variable load (music) during the running in period. Regular/fixed inputs won't properly load stress the driver's power-train.

I was not recommending that anyone actually does that. I was making a point that you could easily detail a specific regime for breaking in your drivers that wouldn't be ambiguous, in the way that 'play x type of music at low levels for the first 100 hours' is.

You could go so far as providing a file that can be used for the first break in period of 24 hours, then another file for the 2nd etc. If you include a 50hz test tone along with these then you can say, play the 50hz test tone and make sure the voltage across the drivers terminals when the driver is in free air is = 0.5Vrms. Then you know and they know that they are going to be playing the break in file at the correct volume. And why 50hz? Because most cheap crappy multimeters should be able to measure AC reasonably accurately at 50hz.

Quite frankly I'm surprised that you seem to think that what I recommended would damage your drivers. The voltage figures were picked directly after simulating with one of your drivers in what would be close to a free air environment and were always quite a bit below Xmax.

If you're implying that your drivers are that fragile then I think it absolutely necessary that you provide files and details on how exactly how to break them in.

Simply playing a 100hz test tone @ 1Vrms = less then 1mm of excursion peak to peak and should not damage the driver. The 40hz test tone @ 1Vrms would give ~1.7mm peak to peak excursion and the final one - 40hz @ 2Vrms would give ~3.5mm peak to peak excursion. Are you trying to tell me that your driver would be harmed by doing such a thing? If so then I think you need to redesign them.


I think Art hit the nail on the head when he said

You need to be far more specific in your specifications or reasonable, educated purchasers of your product (who have not read your warnings here) will continue to break them, even if you do manage to put the brakes on the suspension.

No one has any problems with the way you're doing things, it's just that your comments regarding excursion in the datasheets or on line are always extremely vague and open to many different forms of interpretation.

Saying that you've tried providing more concrete figures and that it created more problems then it solved is plain nonsense. If people are asking you because they don't know what it means then I do not think you are in any way out of place by simply replying, "if you don't know what it means, then please go learn and find out".

It is not your job as a transducer designer to teach people loudspeaker basics. Just because you're easily contactable doesn't change this either.

I suppose you could add in at the bottom of every spec sheet, if you don't know how to calculate the excursion demands/requirements of a driver vs drive level when in a suitable box, then go and download for free winISD Pro and educate yourself.

Oh and as to the definition of Xmech. Whether or not you agree with what's written on wikipedia is besides the point. Of the manufactures that I've seen describing the parameter it is always implied that this is the maximum travel you can take the driver through before damage will occur, it is often called different things. Interestingly SEAS, Scanspeak, Beyma, 18Sound, Precision Devices, Eminence, etc all list a parameter that would be described as Xmech. Eminence even go so far as to explain what this means (what they call Xlim) in their section on Understanding Loudspeaker Data | Eminence Speaker. Now you're not going to tell me that all of them are wrong are you?

[vacuphile]

Bumped backplate, wider surround and a progressive spider would go a long way to get the Xdestruction more in line with the Xmax. But that would drive up the price, and I don't know what the market will bear. Plus, a wider surround might have an unwanted weight panelty, etc. And even if you did that, in the end, something else, like the VC, would give the ghost when people don't use drivers within their enveloppe. No loudspeaker manufacturer can protect himself from that.

Listing Xmax figures is tricky, I fully agree with Mark on that. Also, because who on this forum has a strobe setup to reliably measure the X on their driver? Providing a specification that no-one can put to any use is pointless.

Having followed a number of threads on applications for Mark's drivers, my guestimate is that most problems are not caused by the drivers, but by the enclosures. Folded horns etc. provide in their working region good acoustic coupling between the driver and the air, restraining cone movement, but below their cut off frequency, the driver becomes 'free floating'. Only two options here: 1) limit power, or 2) filter out the low frequencies that the loudspeaker cannot produce anyways.

Amplifiers used to have a 'rumble' filter that could be used to that end, but no longer so.

[markaudio]

Originally Posted by 5th element

1 - Quite frankly I'm surprised that you seem to think that what I recommended would damage your drivers. The voltage figures were picked directly after simulating with one of your drivers in what would be close to a free air environment and were always quite a bit below Xmax. If you're implying that your drivers are that fragile then I think it absolutely necessary that you provide files and details on how exactly how to break them in.

2 - Saying that you've tried providing more concrete figures and that it created more problems then it solved is plain nonsense. If people are asking you because they don't know what it means then I do not think you are in any way out of place by simply replying, "if you don't know what it means, then please go learn and find out".

3 - It is not your job as a transducer designer to teach people loudspeaker basics. Just because you're easily contactable doesn't change this either.

4 - Oh and as to the definition of Xmech. Whether or not you agree with what's written on wikipedia is besides the point. Of the manufactures that I've seen describing the parameter it is always implied that this is the maximum travel you can take the driver through before damage will occur, it is often called different things. Interestingly SEAS, Scanspeak, Beyma, 18Sound, Precision Devices, Eminence, etc all list a parameter that would be described as Xmech. Eminence even go so far as to explain what this means (what they call Xlim) in their section on Understanding Loudspeaker Data | Eminence Speaker. Now you're not going to tell me that all of them are wrong are you?

Hi 5th,
I'm surprised by some of these comments, wide of the mark. Dealing with your main comments in numbered order.

1 - I think you be have the makings of driver designer, but you'll need to focus on component manufacture and operation. Its taken me 12 years+ so far (ouch). The primary objective when running-in a driver is to use variable load to the prevent a fixed or semi-elastic mechanical memory formation within the suspensions. This is particularly important for most synthetic rear suspensions, cotton suspensions are slightly less critical. Partly why many makers use Klippel or similar variable load test signal generating equipment rather than solely relying on fixed input tests. Apologies for not explaining this in my last post, although I have covered this is other posts over the years.

Regarding failure rates generally, the dealers tell me I'm in the lower-middle of the basket of domestic DIY driver makers, that is a lower than average return on failed drivers. Thats more or less where I'd expect Markaudio drivers to be. The last two years, our net return rate is 2.1% of sale, that ain't bad for low load drivers as this includes things like drivers damaged in transit etc.


2 - Its not me you need to convince. Take a look at some of the previous posts on this thread. Not everyone thinks the way you and Art do. Quoting you "if you don't know what it means, then please go learn and find out". I fear this is a bit too brutal for many Diyers and customers.

3 - Now you're telling me job? Diyer's ask me to comment (as per thread starter), I do my best to contribute to the debate. But I'll decide what my job's about thanks all the same!

4 - Wiki, I'm a bit surprised their researcher didn't qualify their X-Mech reference. Its not representative of the bulk of the quoted driver specs. Quoting Eminence is a mis-leading reference. Pro Audio is possibly the only part of driver making industry that some might quote a mechanical limitation specifically for the safe operation of drivers. As far as I can see, even in that part of the industry, Eminence is primarily the exception, rather than the rule. I (and some other members) recognise the operational distinction. You choose to differ on this issue.

This is the "Full Range" forum. Audax, Dayton, Fostex, Eton, Fertin, Feasterex, Fountek, Hi-Vi, Morel, Peerles, Tangbang, Wavecor and more, are the likely typical makers that Diyers on this forum have familiarity. These makers don't quote the figs you want or offer direct/specific advice relating to X-Mech.

Most would agree that Fostex has been the predominant maker in Full Range for the 1/2 century. I couldn't find any mention of X-Mech. But I did find this reference on their international site, which appears to be the nearest to the issue:

INPUT POWER
Indication of input power may be for both 'maximum allowable input power' and 'rated input power', or for either one of them. The fact that they are given different definitions makes it necessary to confirm that such figures are adequate for your intended application.

(1) MAXIMUM ALLOWABLE INPUT POWER
This term defines the maximum input power which can be momentarily applied to the driver. However, note must be taken that such power is a measurement of certain individual frequencies, and thus, not allowed for every frequency point. To be safe, input power defined using the term 'Music' should be considered, in that it indicates peak input power encountered during reproduction of a general music source. In no sense do input values mean that no sound will be produced unless such power is input, nor may drivers be damaged unless the output power of an amplifier is lower than the input value. Under ordinary listening conditions in homes, the possibility of excessive power being input will be rare even with an amplifier having output capability of up to several hundred watts, except under extraordinary conditions. Apart from occasions for PA purposes or where drivers are driven for experimental purposes, input values may be regarded as certain guidelines for enjoyable reproduction of Hi-Fi sound in homes.

(2) RATED INPUT POWER
This value represents the upper limit of the power which may be continuously input into drivers. It must be noted, however, that, as with the maximum allowable input power, such values are effective only for certain selected frequencies, not for the entire audio spectrum. Nor, does it mean that such value can be input if the frequency of interest is constant, as for some special purposes such as measurement or experiment purposes. Continuous input of a single frequency should be interpreted as if the driver load is substantial and very burdensome.

Please feel free to lobby these makers, and the bulk of the rest of the industry to do it your way. I see no meeting of minds on this issue. You're using me as a soft target when the issue is much broader than 1 small maker.

Mark.

[5th element]

Originally Posted by vacuphile

Bumped backplate, wider surround and a progressive spider would go a long way to get the Xdestruction more in line with the Xmax.

From what I understand Marks drivers are being damaged by the edges of the cones deforming or tearing, as a direct result, presumably, of being 'pulled' in the wrong direction too strongly by the surround. I am not a transducer designer so I cannot say what all of the issues would be of increasing the surround size. I suppose one solution might be to thicken the very outside edge of the cone to reinforce it so it resists being bent/torn, but then again that might fly in the face of what people want these drivers to do, especially as people are asking for more sensitivity and that would act to lower it.

Originally Posted by vacuphile

Listing Xmax figures is tricky, I fully agree with Mark on that. Also, because who on this forum has a strobe setup to reliably measure the X on their driver? Providing a specification that no-one can put to any use is pointless.

Well none of us actually measure the excursion of our drivers when in use, we simulate how the driver will respond given the environment it is placed in and then calculate the maximum safe voltage that can be applied for a given frequency so that the excursion limitations of the driver aren't exceeded. With modern digital media, with 0dB representing the maximum signal level, it is quite easy to determine the maximum position you can set on the volume control so as not to exceed the drivers maximum excursion at any given frequency. Obviously a driver such as these should be high passed so as not to try and reproduce frequencies they cannot, as that would otherwise result in a very limited maximum volume if the driver were to try and reproduce something around 20hz.

Originally Posted by vacuphile

Having followed a number of threads on applications for Mark's drivers, my guestimate is that most problems are not caused by the drivers, but by the enclosures. Folded horns etc. provide in their working region good acoustic coupling between the driver and the air, restraining cone movement, but below their cut off frequency, the driver becomes 'free floating'. Only two options here: 1) limit power, or 2) filter out the low frequencies that the loudspeaker cannot produce anyways

It's funny you should say that because briefly I've been trying out different enclosures for the CHR70 and from my perspective the only 'normal' type of cabinet that you should use with one is sealed. They don't appear to fit that well into ported enclosures and they unload at pretty high frequencies too resulting in wasted excursion on frequencies the driver cannot produce - A sure fire way to damage the driver if you want to drive it with pop/rnb style stuff.

One configuration they appear to work very well in is a simple sealed box of 10 litres with an electrical high pass @ 100hz with a filter Q of 0.7. Like this they practically fall into an acoustic LWR 4th order roll off at 80hz, perfect for partnering with a small sub.

It also bears mentioning that even in that configuration, at certain frequencies the driver will most like experience damaging levels of excursion perhaps (7.5mm peak to peak @ 80hz, is that safe?) when driven with the maximum rated input power of 20 watts. Given all of that and considering most people probably don't high pass them, I am not surprised that people end up damaging them.

[5th element]

I am not using you as a soft target Mark, at least not intentionally.

Originally Posted by markaudio

Hi 5th,
I'm surprised by some of these comments, wide of the mark. Dealing with your main comments in numbered order.

If I have something wrong then I really don't mind being corrected, I am after all here to learn as much as the next guy is.

Originally Posted by markaudio

The primary objective when running-in a driver is to use variable load to the prevent a fixed or semi-elastic mechanical memory formation within the suspensions.

Presumably if 'memory' were to form in the suspensions then the driver would sit off centre after the memory had formed. A sine wave should send the suspension through one direction then repeat the process in completely the opposite direction, so you'd figure any memory created from the out stroke would be removed by the in stroke, I guess you're saying this is not the case or am I missing the point? Regardless of that though if sine waves aren't a good idea for running in your drivers, I did say that you could provide your own sound files for the effective running in of your drivers and only use a brief sine wave to set the level at which to run them in at.

Originally Posted by markaudio

I fear this is a bit too brutal for many Diyers and customers.

Why is that? I see DIY primarily as a learning experience with benefits. This is the part where doing it yourself comes in, you don't do it yourself if you're incompetent or ignorant. You might start out with the idea whilst being ignorant but because you want to do it yourself you teach yourself the required skills to accomplish the job successfully. In this case learning how to simulate drivers in boxes in free computer programs. From what I've read about your drivers, learning a little about excursion and how to predict/design around it would seem critical to maximising the performance of their designs.

Originally Posted by markaudio

3 - Now you're telling me job? Diyer's ask me to comment (as per thread starter), I do my best to contribute to the debate. But I'll decide what my job's about thanks all the same!

I'm saying that if you provide the specs and the DIYers don't know what to do with them, then it isn't your job to teach it to them. Removing a spec or an xmax recommendation from a datasheet, because people ask because they don't know what to do with them, doesn't solve the problem, it just ignores it. And then like Art said, if you're knowledgeable and see the datasheets you see an xmax figure and you then design a loudspeaker system around that figure and end up damaging your driver.

Originally Posted by markaudio

Quoting Eminence is a mis-leading reference. Eminence is primarily the exception, rather than the rule.

I'd say they are in fine company along with the other pro audio manufactures I mentioned, unless that wasn't your point.

Originally Posted by markaudio

Pro Audio is possibly the only part of driver making industry that some might quote a mechanical limitation specifically for the safe operation of drivers.

And this is where you've hit the nail on the head. An industry or situation or set of circumstances where it is necessary to define mechanical limitations so as to prevent damage where it is important. The pro arena is different in this regard vs the majority of typical consumer audio drive units manufactured. However your drivers aren't typical and this is the point. Of the manufactures you mentioned...

Originally Posted by markaudio

Audax, Dayton, Fostex, Eton, Fertin, Feasterex, Fountek, Hi-Vi, Morel, Peerles, Tangbang, Wavecor

I bet that almost all of them can be used at their rated xmax without damage occurring. Your drivers however cannot. This is where your drivers and the pro world meet. In the pro world it is necessary for the loudspeaker designers to know exactly where the limitations of the drive units lie so nothing becomes damaged. The same is true for your drivers, as the xmax parameter quoted in your datasheets is not representative of a realistic safe repeatable excursion, how are people supposed to know how hard they can drive them? It is necessary for you to define an Xmech type parameter so people stand the best chance of keeping their drivers undamaged.

If the majority of DIYers don't know what to do with an Xmech figure this isn't a good reason to omit it from a datasheet. And the fact that lots of other loudspeaker manufactures don't quote the figure isn't any defence of your position either. They provide an Xmax figure, a useful xmax figure and you know that you aren't going to damage their drivers if you use them within the rated xmax.

You however don't provide any solid figure for safe excursion, xmax is clearly isn't it, but what is?

Originally Posted by markaudio

Most would agree that Fostex has been the predominant maker in Full Range for the 1/2 century. I couldn't find any mention of X-Mech. But I did find this reference on their international site, which appears to be the nearest to the issue:

Yes, but you wont break any of their drivers running them at their rated xmax.

[18Hurts]

Very interesting,

Most drivers have the weak point in the suspension--not the cone when dealing with Xmax. Other drivers have issues dealing with the voice coil--MA drivers weak point is the CONE which makes sense to me.

Why? Because Aluminum breaks--the nature of the beast. It is lighter by volume than titanium so for the lightest things possible--aluminum is king. Sure, it has the durability of a potato chip but you know that going in. Look at racing bicycle frames--the steel or titanium ones have a lifetime warranty...aluminum is 3 to 7 years. Aluminum bicycle rims last 10 to 15 years--and crack but such is life. Want to really break an aluminum rim? Anodize it, scratch it up with the rim brakes (creates stress risers) and buy a new rim after 5 years or so.

Look at springs--steel or titanium since they can flex within specs for years--decades and as long as you don't exceed the stress specs--very durable. There is a reason they don't make aluminum springs--they will break since the material does not like to be flexed--it fails and it is not a question of if but of when.

So the MA cones get stress fractures running at Xmax--not hard to understand why--the cones are the lightest possible to get the frequency response and are made of aluminum. The generous Xmax allows them to do something FRs traditionally can't do--real bass response but you have to pay to play--aluminum breaks.

JBL figured that out in the 1970's--their compression drivers all failed after a period of time from stress fractures. They switched to titanium since they have the flex durability over long periods. True, the marketing people spewed the "titanium sound" but it was to eliminate aluminum stress failures.

I've beat the !@#$! out of my paper coned 15" Cerwin Vega woofers over the years--they now are my HT subwoofer at the age of 21 years old. No aluminum cones, no foam surrounds--paper and fabric.

Maybe Mark can make a "bass head" version of the Alpair with kevlar added to paper cones? It will change the signature sound of the Alpairs but would be marketed for the "Xmax" compressed music folks as an alternative. The aluminum version could be the "audiophile" version with the special paper cone for the folks that need Xmax with higher powered SS amplifiers?