FR125S post break-in T/S parameters

[maxro]

I have a box design that lowers the resonant frequency of the system.

Care to elaborate on that?

Max

 

[soongsc]

Care to elaborate on that?

Max

The design is still in patent review cycle, so I'm not able to to provide too much details. But the concept is to make the box to not create a constant load on the driver, but a load that varies with the time and displacement of the driver. Ultimately it is desireable to make the resonance out of the audio range. If possible, no resonance at all without creating adverse effects.

 

[maxro]

Okay, I don't know a lot about this stuff. But, how do you lower a driver's resonant frequency using a box, or any other means for that matter?

Max

 

[soongsc]

Well, if you notice that a box with one port has one resonance below the driver natural frequency and one resonance above. We like the one below, but not the one above, so you design the port so that the higher one goes away.

 

[maxro]

Soongsc,
I notice you said "one" port. From this I'll infer that your design involves a series of ported chambers. I won't pry any further as you've got your patent to sort out. Though, please do post some specifics once you've got the patent protection.

Max

 

[soongsc]

Soongsc,
I notice you said "one" port. From this I'll infer that your design involves a series of ported chambers. I won't pry any further as you've got your patent to sort out. Though, please do post some specifics once you've got the patent protection.

Max

Sure thing. It was filed as a "new invention" the first attempt to challenge this category, so I have my fingers crossed as well.

 

[DanWiggins]

There really isn't a standard, but there should be!

We try to measure at 1W, since - with the drivers we make - that tends to be an SPL level in the 85-90 dB range, which I believe represents a realistic high average output level for most users (90 dB @ 1m is about 80 dB at 10 feet, a typical SPL level for listening).

And it also seems to me to be somewhat useful, especially combined with the efficiency of the driver (which is defined at 1W, 1m).

I guess as we move into higher and higher efficiency designs, perhaps a measurement of T/S parameters at 90 dB SPL would be a good thing. Maybe 94 dB SPL, since that corresponds to 1 Pascal of pressure?

The more linearity of the driver, in terms of BL and Cms over stroke - the more validity the T/S parameters have over a wider input power range. If BL and Cms vary, then the derived T/S parameters will also vary.

Remember, there are really only 5 physical parameters of a driver, from a T/S parameter viewpoint:

- Mms: moving mass
- BL: motor force factor
- Re: DC resistance of the voice coil
- Sd: surface area of the cone
- Cms: compliance of the suspension

All other T/S parameters are derived from these basic 5. If any one of these changes over your operational area, then the derived T/S parameters (Fs, Qms, Qes, Qts, Vas, SPL) will also change.

Dan Wiggins
Adire Audio®

 

[Vikash]

The more linearity of the driver, in terms of BL and Cms over stroke - the more validity the T/S parameters have over a wider input power range. If BL and Cms vary, then the derived T/S parameters will also vary.

I think I've just had an OB epiphany.

 

[GringoAudio]

I like 'em

Well... I had a problem with my NAD pre develop at the same time I hooked up these FR125 drivers. It is doing something nasty to the sound. I apologize to all who have a vested interest in these drivers..Bob et al. I like them. I will now have to start all over listening to them. Thanks all for the info and sorry again. I NO TWEETERS hooked up now.

 

[Geek]

Since I got my pair of FR125S drivers yesterday, I thought I might toss in some params I measured as well

Diver 1:
Fs: 68 Hz
Re: 7.1 ohms
Le: 0.297 mH
Z @ Res: 44.1 ohms
Qms: 4.35
Qes: 0.835
Qts: 0.7


Diver 2:
Fs: 61 Hz
Re: 7.2 ohms
Le: 0.307 mH
Z @ Res: 53.8 ohms
Qms: 4.9
Qes: 0.757
Qts: 0.656

I measured these after a 3 hour break-in (if you can call it that) with 15Hz @ full Xmax. The setup was as per the MHSoft page.

 

[Vikash]

Looks good. Thanks for the data.

Any chance of a VAS measurement

 

[Vikash]

Updated:

Ordered by qts:

	fs	vas	qts	qms	qes	#avg_over
geek	64.5	-	0.678	4.625	0.769	2
vikash	65.598	5.938	0.705	5.271	0.815	4
npdang	64.015	4.699	0.714	4.240	0.860	-
dave	75.370	3.370	0.780	5.220	0.930	8

 

[flaevor]

just a beginner question

The varying measurements led to the following question in my mind: When taking the measurements of a speaker does the ambient air pressure of the room as well as the humidity or temperature affect the measurements (signifcantly/audibly)?


I have never measured a speaker myself and have absolutely no idea how much variation is common or even tolerable so I thought i'd ask.

 

[soongsc]

Signal type and amplitude have more significant impact.

 

[Vikash]

I have also noticed changes in how the measured speaker is placed/oriented and how sturdily its clamped, and in room temperature variation. I can't quantify how significant this is overall atm, but it certainly plays a part.

 

[DanWiggins]

I have also noticed changes in how the measured speaker is placed/oriented and how sturdily its clamped, and in room temperature variation. I can't quantify how significant this is overall atm, but it certainly plays a part.

Excellent points! Orientation - and for small drivers, baffle loading - will affect the T/S parameters as well. Different orientation, or even different baffle loadings will change the measured parameters pretty significantly.

Temperature affects the suspension stiffness, Cms, which is reflected in Fs, Qms, Qes, Qts, and Vas (it affects all parameters). And temperature also affects the DCR of the driver, too, when in the extreme. As will power dissipation in the driver.

Atmospheric pressure isn't a concern, unless the driver is mounted in a non-infinite baffle (or close to it).

Then add in that it is rare to find spiders with tighter than 20% tolerance in terms of stiffness, and we expect to see quite a bit of variation.

For the table above, the summary of drivers, note that as Fs moves up, Qts moves up, and Vas moves down. That's strictly a function of the compliance of the spider. And is about what I would expect to see as a typical range of drivers. Of course, if you plot the responses of those different T/S alignments, you'll find VERY little difference for a given box/tuning combination...

And then there's the estimate of Sd - how was it measured? How was the lack of a dustcap accounted for? Change the Sd, and the Vas is changed. We measure from 1/3rd of surround across (from inside diameter of the surround), and for the lack of a dustcap, it's simply the loss of the ID of the former.

Likewise for DCR; was it measured with a multimeter? How accurate is it for single-digit impedances? Did you remove the series DCR of the test leads? I use a meter that's been calibrated against a 4-wire impedance bridge.

I guess to sum up, there are enough issues with measurement of T/S parameters, and with driver variability in the first place, that expecting close grouping is pretty extreme. I guess one could go through and hand-screen drivers or spiders for a specific value, but that is quite expensive, and when you look at the actual difference you see nearly zero effect:

An externally hosted image should be here but it was not working when we last tested it.

I raised the traces from dave and npdang by 1 dB, so they were grouped at the high end with the others. And as we see, there's extremely LITTLE variance in the actual results of the measurement changes. We're talking a grouping with a deviation from 30 Hz and up of less than 1 dB. I think that's quite tight...

This is also indicative of simple suspension changes. If the suspension is tighter, Fs and the Qs go up, but Vas drops in a complementary way, and the net alignment is pretty much the same. That is because the box dominates the compliance of the system, not the speaker's suspension.

Hope this clarifies a few things...

Oh, and Vikash? I hope that OB epiphany was a good one!

Dan Wiggins
Adire Audio®

 

[GringoAudio]

So the best box size for this driver is....?

How did the volumes for the boxes recommended for this driver come about? 7L and the larger one? Neither correspond to volumes generated by plugging any version of the T/S parmeters hav seen derived for this driver. I am a very prescriptive builder and don't know much about varying from what the formuli give me so I probably need a lot more educating in this matter... Thanks. I have gone back to using a tweeter crossed in 2nd order at 20Khz ...and yes I can hear the difference.

 

[flaevor]

I appreciate the answers. I didn't want to infer that atmospheric conditions could have lead to different measurements or that the differences were significant, but rather it just made me wonder if the results would vary with drastic differences in atmospheric conditions.

For example I grew up in a desert area nearly a mile above sea level. Things often work a little differently at high altitudes since they are generally engineered in places at or around sea level. And since the majority of the civilisation seems to live in coastal areas they never notice that anything is wrong.

Anyway thanks for the insight.

 

[GM]

This is also indicative of simple suspension changes. If the suspension is tighter, Fs and the Qs go up, but Vas drops in a complementary way, and the net alignment is pretty much the same. That is because the box dominates the compliance of the system, not the speaker's suspension.

Greets!

For sure! I have a design routine based on conical horn theory from well before T/S that only requires an approximate Sd and Fs spec to achieve excellent results on the same order/FR as an optimized MJK MathCad one, but the trade-off is a longer pipe length and much greater Vb.

GM

 

[GM]

Re: So the best box size for this driver is....?

How did the volumes for the boxes recommended for this driver come about? 7L and the larger one? Neither correspond to volumes generated by plugging any version of the T/S parmeters hav seen derived for this driver.

Greets!

When I plug the published specs into my calculator I get Qtc = 0.701, or close enough to a 0.707 T/S max flat sealed alignment. Not sure what other size you're referring to, but no need to get too concerned about much in the way of variance, especially sealed, as long as it errs on the large side since once any real power is applied to it Qts will rise.

GM