Gpa 416-8c

[alancohen]

That's interesting. There's a 10dB slide starting at 900Hz and ending around 2300Hz. That's understandable. I wonder what accounts for the 9dB bump from 2300 to 2700?

Was this measured inside?

[Old Guy]

That's interesting. There's a 10dB slide starting at 900Hz and ending around 2300Hz. That's understandable. I wonder what accounts for the 9dB bump from 2300 to 2700?

A little phenomenon called beaming, and perhaps a bit of cone breakup overtones thrown in.

[Alien_Shore]

That's interesting. There's a 10dB slide starting at 900Hz and ending around 2300Hz. That's understandable. I wonder what accounts for the 9dB bump from 2300 to 2700?

Was this measured inside?

Yes, measured in my garage. This weekend, I'll take them outside, refresh the imepedance and T/S plots, and take another FR measurement.
That crazy FR rise at the end of the range is pretty typical woofer cone breakup; from what I've read*, paper cones and dustcaps (particularly lightweight ones) have exaggerated peaks like that.

*Some interesting references:
The Art of Speaker Design Part II
Cone Breakup Simulation
Woofer Design Tutorial

[cradeldorf]

shouldn't you measure them where they're gonna sit and then tune them to a flat response?

[Alien_Shore]

shouldn't you measure them where they're gonna sit and then tune them to a flat response?

Yeah, but when you're going to design an enclosure, how should you measure for that? I don't have anechoic chamber or anything so I was thinking just do it outside to get the best "raw" FR measure I can to import into my design tools.

So I did re-measure these after they've broken in for a while and the numbers did change a bit:

Here are the revised T/S:
Re = 7.46
Fs = 24.22
Qts = 0.303
Qes = 0.319
Qms = 6.003
Le = 0.582 (*)
Mms = 85.31
Vas = 17.03

(*)I think my initial value for Le was improperly inflated. I believe this is the accurate number.

I still don't trust the Vas number or the Mms. I still plan to build a 1.5CF sealed box to test Vas that way, but have not yet gotten to that.
But if you believe this new Vas, WinISD is saying to use a 6.82 CF vented box tuned to 24Hz.

I took the 816 cabinet outdoors this weekend and measured. Here's the plot comparing the first (brand new driver, measured indoors) with the new measurement:

[cradeldorf]

But if you believe this new Vas, WinISD is saying to use a 6.82 CF vented box tuned to 24Hz.

I took the 816 cabinet outdoors this weekend and measured. Here's the plot comparing the first (brand new driver, measured indoors) with the new measurement:

I'm sorry I didn't realize you were trying to design cabinets, I thought you were just trying to tune them. I think 6.82 is a good size cabinet. Don't you?

[GM]

Hmm, typically Fs drops, Vas increases, Qts remains ~ the same, ditto measured efficiency, so something seems amiss. FWIW, the most accurate way to measure Vas [and the most tedious] is to find the sealed cab net Vb that raises Fs 1.56x, then divide it by 1.44.

Don?t currently have access to my collection of measured specs, but Le @ the 1 kHz reference is ~1.2-1.5 mH IIRC. A traditional 15? with 0.582 mH would need a shorting ring, which Altec never used AFAIK on any of its woofers, relying on its inductance rise to extend the driver?s HF response, especially when driven by a high output impedance.

You can check it though by working off an impedance plot:

Le = 1000*SQRT[[Zxo^2-Re^2]/[2pi*1000]] where Zxo is the resistance @ 1 kHz = 1.4954 mH Vs DATS?s 1.478 mH just ?eyeballing? Zxo = 12 ohms off your original plot, so it was accurate enough for me originally since it falls in line with my faded memory.

GM

[Alien_Shore]

Hmm, typically Fs drops, Vas increases, Qts remains ~ the same, ditto measured efficiency, so something seems amiss. FWIW, the most accurate way to measure Vas [and the most tedious] is to find the sealed cab net Vb that raises Fs 1.56x, then divide it by 1.44.

Don?t currently have access to my collection of measured specs, but Le @ the 1 kHz reference is ~1.2-1.5 mH IIRC. A traditional 15? with 0.582 mH would need a shorting ring, which Altec never used AFAIK on any of its woofers, relying on its inductance rise to extend the driver?s HF response, especially when driven by a high output impedance.

You can check it though by working off an impedance plot:

Le = 1000*SQRT[[Zxo^2-Re^2]/[2pi*1000]] where Zxo is the resistance @ 1 kHz = 1.4954 mH Vs DATS?s 1.478 mH just ?eyeballing? Zxo = 12 ohms off your original plot, so it was accurate enough for me originally since it falls in line with my faded memory.

GM

Yep, you're right on, GM. John Murphy (DATS designer) let me know that DATS has selectable ranges and units for several of the t/s parms. My first Le readings for this driver (~1.5mH) were at 1kHz, while the second readings (~0.5mH) were at 10kHz. For others using DATS, you simply double-click on the units box to toggle (you can see which parms have selectable units because of the little box drawn around them).

I'll try to get that test box done this weekend. I'm not trusting my delta mass technique

FWIW, here are the T/S params in free-air, cone vertical, before and after break-in (sans Vas added mass that was skewing the data):

GPA 416-8C	new	after break-in
Re	7.77	7.59
Fs	23.55	22.88
Qts	0.294	0.293
Qes	0.308	0.309
Qms	6.638	5.931
Le (@1kHz)	1.51	1.47

[GM]

I'm not trusting my delta mass technique

Why not?

Not familiar with these types of programs, so can you post all specs it spits out when no attempt is made to measure Vas other than inputting the 12.75" piston diameter?

GM