How to Train Your PM180-8

[Pete Schumacher]

Re: How to Train Your PM180-8

This is a going to be a weak debate if we're just going to agree with each other. :D


So you're saying you play a 1kHz tone which generates a 4th harmonic at 4kHz, which excites more harmonics. I'd think the hamonics would have to be pretty significant to compound this way significantly. On the other hand, if we look at a driver like the Seas L18:
[ATTACH=CONFIG]58038[/ATTACH]
This driver rings like a bell at 7kHz, with the output a good 15dB above the normal passband for this driver. Here we can see that cone resonance is present as the lower frequencies excite it, but I think this distortion is present more because the output of the driver at 7kHz is that much higher, than the fact that the output at 7kHz also generates high harmonic content. The 3rd harmonic at 2.5kHz is high because the driver itself will amplify that output by 15dB, and there's nothing a crossover filter can do to fix that other than to drop the signal at 2.5kHz itself (crossover lower).

The real killer is the 5th harmonic. THAT kind of distortion is audible due to the fact that there is so much separation from the fundamental. That particular driver needs to be limited to woofer use only crossed 500Hz or lower.

The distortion of the PM180 is definitely rising above 2KHz, but the nice thing is that the cone doesn't seem to amplify higher order harmonics like the ringing bell example above. The usable band to 2KHz stays very clean despite the breakup modes of the cone.

But now that PE knows about the cone and pole piece mods, maybe they'll take it to heart, add it to the next iteration of the PM180, jack up the price a little bit, and they'd really have a top tier driver with few shortcomings at all. I'd still buy it either way. But I won't give up my Revelator mid either. :D

[dcibel]

Re: How to Train Your PM180-8

I agree completely. Harmonic distortion does matter, and I also find it hard to believe -40db is not audible. I think you are onto something with the Lx conversation and why so many Dayton RS designs sound like they are not optimized. That, and their cone is horrid.

The driver graph pictured is taken straight off of Zaph's website. My personal experience with that driver does not show 3rd being higher than 2nd. YYMV.

This is a going to be a weak debate if we're just going to agree with each other. :D

Let's say a driver had really bad distortion at 4kHz but you made a filter at 1kHz, - but since there is 2nd and 4th order distortion happening that could excite that 4kHz with "bad distortion" do you t hink that would be heard at all?

So you're saying you play a 1kHz tone which generates a 4th harmonic at 4kHz, which excites more harmonics. I'd think the hamonics would have to be pretty significant to compound this way significantly. On the other hand, if we look at a driver like the Seas L18:

This driver rings like a bell at 7kHz, with the output a good 15dB above the normal passband for this driver. Here we can see that cone resonance is present as the lower frequencies excite it, but I think this distortion is present more because the output of the driver at 7kHz is that much higher, than the fact that the output at 7kHz also generates high harmonic content. The 3rd harmonic at 2.5kHz is high because the driver itself will amplify that output by 15dB, and there's nothing a crossover filter can do to fix that other than to drop the signal at 2.5kHz itself (crossover lower).

[Pete Basel]

Re: How to Train Your PM180-8

MCM 55-3870. 5" aluminum cone, cast basket driver. Sadly NLA.

My point of view is easily understood - it is based on claims that harmonic distortion "basically doesn't matter" and that -40db represents the threshold of audibility. My point of view is that basically, harmonic distortion does matter and threshold of audibility is probably lower than -40db, hence the usage of drivers that offer extremely low distortion at all listening levels - not just at insane SPL where better drivers are going to hold together in a more linear fashion.

In short - other than brand recognition, why are we wasting money on $280 5" drivers? It isn't because there are things we "cannot measure", either - I think we have measured these things to no end. Pastries taste better out of a pink box etc, so lets assume there is already some of that involved in the driver choice but I don't believe wanting to use a Revelator is the sole justification for using a Revelator over any number of other drivers out there.

I agree, and I'd say that -40dB is just a guideline, certainly odd order distortion is going to be more audible, there are many factors to consider.

[jamikkim]

Re: How to Train Your PM180-8

Let's say a driver had really bad distortion at 4kHz but you made a filter at 1kHz, - but since there is 2nd and 4th order distortion happening that could excite that 4kHz with "bad distortion" do you t hink that would be heard at all?

[johnnyrichards]

Re: How to Train Your PM180-8

I agree completely. Harmonic distortion does matter, and I also find it hard to believe -40db is not audible. I think you are onto something with the Lx conversation and why so many Dayton RS designs sound like they are not optimized. That, and their cone is horrid.

The driver graph pictured is taken straight off of Zaph's website. My personal experience with that driver does not show 3rd being higher than 2nd. YYMV.

[dcibel]

Re: How to Train Your PM180-8

Let's do it

Why wasn't this driver more popular with high end commercial and DIY guys? As a midrange, it easily exceeds the -40db threshold of so-called "inaudibility". It offers superb linear response, excellent build quality etc... Remove the excursion demands of the bottom two octaves from the driver, what makes a Revelator a better choice?

[ATTACH=CONFIG]58029[/ATTACH]

Fun, while its difficult to apply any direct comparison of distortion graphs without knowing details of mic distance and SPL, or even the size of the driver in question, the immediate things to note in your graph is the F3 is higher than F2 right across the entire midrange, and has highish F4 as well. A Revelator will have 3rd and 4th order distortion a good 10dB lower across the entire range.

I find it hard to believe that -40dB distortion to be "inaudible", if that were the case pretty much every driver Zaph ever measured has inaudible distortion over the useful passband. Relative distributions of harmonics contributes greatly to the sound character IMO. One thing you'll also find with the higher end drivers is that they won't fall apart at high volume, due to better suspension components and better motor design.

Another interesting thing Zaph did was take some measurements of inductance over excursion. It was discovered in this testing that some cheaper motors had a significant change in inductance over excursion in the higher frequencies, which IMO would basically represent as amplitude modulation at high volume, which wouldn't be caught in a high level FR sweep. Perhaps a pink noise measurement at low and high volume overlaid could display how much of an effect this has in the real world. A change in inductance at high frequency on a metal cone driver could be very detrimental, as it would shift the frequency of a notch filter that is supposed to cut the cone resonance out of the picture. I found it interesting that Seas Excel drivers had performance here equal to an under-hung motor, so this isn't a concern with their drivers at all.

Regardless, the points I was trying to make is where to draw the line on the benefit of out of band distortion from cone breakup. If we cross over at a point where the fundamental is down 12dB at 3.5kHz, I am going to make a possibly incorrect assumption that the harmonics generated at 3.5kHz are also down 12dB, do we care about their contribution to the output since they are attenuated by the crossover.

Apart from driver performance, visual appearance also plays a lot into one's decision to use a driver. If the driver performs well, but is ugly, it probably won't be a popular choice. For example, in my mind B&G drivers perform very well, but I'd never use them because of their appearance.

[johnnyrichards]

Re: How to Train Your PM180-8

MCM 55-3870. 5" aluminum cone, cast basket driver. Sadly NLA.

My point of view is easily understood - it is based on claims that harmonic distortion "basically doesn't matter" and that -40db represents the threshold of audibility. My point of view is that basically, harmonic distortion does matter and threshold of audibility is probably lower than -40db, hence the usage of drivers that offer extremely low distortion at all listening levels - not just at insane SPL where better drivers are going to hold together in a more linear fashion.

In short - other than brand recognition, why are we wasting money on $280 5" drivers? It isn't because there are things we "cannot measure", either - I think we have measured these things to no end. Pastries taste better out of a pink box etc, so lets assume there is already some of that involved in the driver choice but I don't believe wanting to use a Revelator is the sole justification for using a Revelator over any number of other drivers out there.

[Pete Basel]

Re: How to Train Your PM180-8

True, but what I find most beneficial is the complete annihilation of the distortion peak at 3.5Khz, although its' audibility is VERY debatable.

Agree completely. It is just an all around better driver, FR and distortion.
And I agree that it is an excellent driver stock, just saying if you want to take it even further
this is good to know.
I don't know how the cones are manufactured but if it was easy to incorporate the slits rather
than as an after thought it would be a plus. There are other Dayton drivers that need help
even more.
I'd like to see some of the 8" drivers with serious breakup issues "fixed".

[Pete Basel]

Re: How to Train Your PM180-8

The distortion profile I posted wasn't a Dayton driver. I am aware of why Dayton isn't used in more commercial offerings. Such a pity, too.

Heard pretty hit or miss Accuton results. Some sound like cheap aluminum Dayton drivers, some sound legendary. Hard to argue the appeal of the price tag association when selling speakers, though.

Would you tell us what the driver is, so that we can better understand your point of view?

[johnnyrichards]

Re: How to Train Your PM180-8

The distortion profile I posted wasn't a Dayton driver. I am aware of why Dayton isn't used in more commercial offerings. Such a pity, too.

Heard pretty hit or miss Accuton results. Some sound like cheap aluminum Dayton drivers, some sound legendary. Hard to argue the appeal of the price tag association when selling speakers, though.

[DoubleTap]

Re: How to Train Your PM180-8

It isn't debatable, if it were - we would all be using cheaper drivers.

Totally untrue, of course it's very debatable. Otherwise nobody would dare use an Accuton for a midrange.

[Pete Schumacher]

Re: How to Train Your PM180-8

Let's do it

Why wasn't this driver more popular with high end commercial and DIY guys? As a midrange, it easily exceeds the -40db threshold of so-called "inaudibility". It offers superb linear response, excellent build quality etc... Remove the excursion demands of the bottom two octaves from the driver, what makes a Revelator a better choice?

[ATTACH=CONFIG]58029[/ATTACH]

1. For a Dayton driver, it's expensive. That's one strike against it with the DIY crowd.
2. For commercial guys, it's a Dayton driver which is associated with the DIY crowd. Strike 2.
3. Revelator has a pedigree. Most DIY builds don't need high sensitivity, so the pedigree wins. Strike 3.

It really is a nice sounding driver and there aren't many high sensitivity mids out there to choose from with performance like this for this price. That's one reason we went with it in our higher sensitivity design. Can't use a Revelator if you want better than 90dB sensitivity. And most commercial designs don't even try to leave the 80s/2.83V, so why use the PM180?

Just some observations as to why it doesn't show up in more builds.

If a DIY guy is going to do a high sensitivity 3-way, it should be high on the list as the mid over a number of other "pro" models.

[johnnyrichards]

Re: How to Train Your PM180-8

Let's do it

Why wasn't this driver more popular with high end commercial and DIY guys? As a midrange, it easily exceeds the -40db threshold of so-called "inaudibility". It offers superb linear response, excellent build quality etc... Remove the excursion demands of the bottom two octaves from the driver, what makes a Revelator a better choice?

[dcibel]

Re: How to Train Your PM180-8

It isn't debatable, if it were - we would all be using cheaper drivers.

Ya! Debate time! Let's argue over the merits and audibility of decreasing distortion at 3.5kHz on a driver that will in most cases be low-passed at 2kHz.

[johnnyrichards]

Re: How to Train Your PM180-8

True, but what I find most beneficial is the complete annihilation of the distortion peak at 3.5Khz, although its' audibility is VERY debatable.

It isn't debatable, if it were - we would all be using cheaper drivers.