Seas 25TV EW Theile Small parameters - Qts & VAS values way off

[4R5]

I was reading PeteB's posting about the Dynaco A25 Speakers

Specifically his measured T/S parameters of a Seas 25TV EW woofer that he had.

I recently picked up 4 of the 15 ohm version of the 25TV EW that was used in the Dynaco A-50, assuming the T/S parameters would be close to the 8 ohm version. What I found is the Qts and VAS values are way off from what PeteB posted.

I then pulled a 25TV EW woofer from one of my A25s and found those measurements are off as well.

I'm using WinISD with a sound card (no external power amp) with a 121 ohm series resistor.

Here's a table of my results along with PeteB's results.

	25 TV-EW Seas Spec	25 TV-EW Pete B data	25 TV EW 2674 date 8 ohm measured	25 TV EW 15 ohm measured
Sd	285	350	350	350
Fs	20 - 25	23.6	31.3	33.9
VAS		205	109.69	103.54
Re		5.7	6.96	12.19
Qe		0.48	0.973	0.928
Qm		4.2	7.321	7.552
Mms	30	37.5	41.06	36.99
no		0.36
SPL ref	88	87.6
BL	8?.8	6.6	7.597	10.175
Qts	0.35	0.43	0.859	0.827
Cms			0.631	0.596
Rms
Dd			211	211
Le			4.603	5.967
added mass			32.78	32.678

Am looking for suggestions as to why my values are so far off.

Thanks

[4R5]

I should add as you can see in the table, Fs for the 15 ohm is much higher than the Seas spec for the 8 ohm, and even my 8 ohm Fs is higher.

Also the Qe & Qm values are almost 2X what PeteB measured, which contributes to the higher Qts values.

[Pete B]

Qe and Qm scale linearly with Fs, so part of the error can be accounted for due to the Fs difference.

Consider that a base set of measurements do not require the delta mass or compliance second

measurement.  Fs and the Qs are included in that and Fs simply requires an accurate generator 

which should be easy with a digital source.  You could check for distortion in your gear, do you 

have an oscilloscope to have a look.  You could try to confirm Fs by the old way of using a high

impedance source and just measuring the voltage peak with a voltmeter.  I'd say that this is the

first step.

The Q values can have error due to resistance in your test leads or calibration error.  If your system

can measure impedance try an 8 ohm resistor and see if there is error across frequency.

[4R5]

Hi Pete, thanks for the reply.

First off I mentioned Win ISD which was a mistake, I'm using REW for analysis.

I did measure Fs with a stand alone generator and frequency counter, ironically I found Fs to to go down with drive voltage.

This is one of the 15 ohm Seas drivers:

Drive Voltage    Fs
100mV           37Hz
200mV           37Hz
300mV           36Hz
400mV           36Hz
500mV           35Hz
600mV           35Hz
700mV           34Hz
800mV           34Hz
900mV           33Hz
1V                   33Hz
1.5V                31Hz
2V                  30Hz
2.5V              29Hz

REW runs their Z at 100mV. Yet their value is what I measured at 1V.

I did verify accuracy, I used a measured 21.1 ohm resistor and REW reports back 21.1 across the frequency spectrum.

I do have a scope, will take a look at the waveform.

How did you measure your driver - manually or with another program?

Thanks!

[Pete B]

I've been measuring with Bill Waslow's very old program LAUD for over 20 years and I trust it. 

I try to measure tweeters around 1V and 2 to 3V for midranges and woofers.

I wonder if Fs depending on drive level has anything to do with the rubber aging.

I also wonder if REW might have an error in the calculations.  There are free programs that 

go along with ARTA for measuring T&S parameters, perhaps you could use that as a sanity check.

[4R5]

Thanks Pete, I'll try ARTA.

[4R5]

Pete, one more question, when you mention 2 to 3 volts for mids & woofers, is thatat the driver or at the source? What series resistor do you typically use?

[4R5]

Correction: what value of series resistor. Thanks!

[Diogenes]

Factors to consider: these are old drivers. Rubber surrounds are not immortal, they stiffen with age (plasticizers leach out and the flexible long chain polymers break into shorter molecules that results in a hardened material). This can drive Fs up as compliance goes down. In contrast, spiders get very soft with age, which can drive Qms way up. These two factors interact with one another, so it's hard to say which way things will end up. Most importantly, these have alnico magnets. Alnico V, the kind used in these drivers, loses ~3-5% of its flux strength per decade just sitting on the shelf and more if it's being exposed to temperature swings, shocks or a magnetic field such as is formed by music being played through a voice coil. Too much power, a clipping amplifier or even just touching the input lead on an amp and causing a ground hum can cause significant demagnetization of the drivers. When the Dynacos were new, Dynaco, Scan-Dyna and SEAS were seeing high rates of woofer returns due to loss of flux, upwards of 30% typically, from the speakers being used. It was a major impetus to SEAS to develop the ceramic magnet 25 F-EW, long before the civil war in Zaire led to the cobalt crisis that forced other manufacturers to change over, as ceramic doesn't have the demagnetization issues.

You're most likely measuring the after effects of significant loss of flux strength in the motors, softened spiders and hardened rubber surrounds. This makes it hard to get old  Dynacos to match in mismatched pairs, actually, if they haven't been exposed to the same environment and use all their lives. Your only options are to completely tear the drivers apart, have the motor remagnetized and recone them (not recommended as you can't get the right parts anymore), or buy a few more drivers and test them until you find a set that match 'close enough'.

[4R5]

Good points, thanks. My original quest was to try to replicate Pete B's values. Granted his tests were 16 years ago, but the drivers were then already 30+ years old. I did find that one of my Seas woofers, a 25TV-EW but with a round rim, not the more typically seen rim with the mounting "bump outs" had a different set of TS parameters than the other one with the "bump outs". Like wise the 15 ohm version TS parameters more closely matched the round rim 8 ohm driver. The 8 ohm one with the "bump outs" had TS parameters close to Pete B's data.

[Diogenes]

Yeah, you're definitely seeing issues from softening spiders, hardening surrounds and, the most significant impact, loss of flux strength in the alnico. Just to help, here's a more complete set of factory parameters for the 8ohm version of the 25 TV-EW:

The round frame SEAS drivers were introduced at the same time as the 25 F-EW. They're actually the F-EW's basket, but were used on the 25 TV-EWs that continued to be produced only on an OEM basis for Dynaco and others. That's if they're marked as SEAS. The Scan-Speak clones always had the round frames.

[4R5]

Thanks for the full specs, have been looking for them for a while.

[Pete B]

I use a 10 ohm resistor so it is about half the voltage at the speaker.

One thing I should have mentioned is that the rubber edge could have hardened, though mine

looked fine.  It depends on the type of rubber and I have many woofers with rubber edges 

that have Fs now that is 2X what it should be.  Wintergreen oil is talked about a lot on the Web

and Youtube and it works great for a day or 3 let's say.  Doesn't take long for them to be back

to 2X.  I will probably try brake fluid next time being careful not to get it on the glue.