Low Qts drivers : Adding mass

[316a]

Hi ,
I have a pair of vintage 12"drivers , Richard Allan CG12 super which were cheap so I'm thinking of 'design mule'project for the workshop . These have a published Qts of 0.105 , which seems ridiculously low , and Fs of 30Hz . Although I am unsure how accurate the published specs are , these immediately appear to be like a guitar speaker with low Fs . Originally the drivers were mounted in a 55l sealed enclosures which did not deliver much bass , even in the corner . Modelling in Hornresp these seem ok for basshorn use due to the high BL , low Qts , but I just don't have the space for a pair of whacking great basshorns The drivers seem to be sensitive (high 90's) so I'm thinking perhaps add some mass to the cone via some small neodymium magnets and mount in vented boxes , I can easily lose 6dB if it means getting some bass out of the things . I've consulted the Loudspeaker Cookbook and a few other sources and can see that by adding mass Qts will increase , Fs decrease as will BL , but what about Vas and other parameters ? I'm no expert , help appreciated

cheers

316a

[AndrewT]

Hi,
I had a pair of these as a 3way (Pavane) in 1971. When 10W to 20W was sufficient power.

They are more efficient than a Tannoy HPD385 in 100L vented cab.

But they only handle 25W (that maybe includes the other drivers and xover losses) and will not like added mass due to limited thermal capacity.

Try adding small series R instead to raise the Q.

ps I don't think the Qts is quite as low as you quoted, what was your source?

pps BL will not change.

[kelticwizard]

316a:

Do you have any other specs on that driver besides Qts? Dc resistance and Qes would helpful in figuring how much any given added resistor would both raise the Qts and lower the sensitivity.

DC resistance, (Re), acan be measured with any voltmeter, analog or digital. If we absolutely have to, we can probably estimate Qe as being just a touch above Qts.

Two formulas can help us.

A) Resistor = [(Qts required / Qts measured) X Re] - Re. We can, if we want, substitute the specified Qts for Qts measured.

B) Qes' = Qes(1+(resistor/Re))

If we can calculate the change in Qes, I can plug the numbers into BoxModel and give you the drop in sensitivity. That is if I have a value for Re. And depending, of course, if the value of Qts is indeed 0.105

[316a]

I don't think the Qts is quite as low as you quoted, what was your source?

...here !



link

[kelticwizard]

Surprisingly, the Thiele-Small parameter chart does not include Re. Any chance of taking a quick measurement in order to give us the DC resistance?

If not, could you tell us if it is a 4 ohm or 8 ohm? Most, but not all, 8 ohm speakers have a DC resistance around 5.8 ohms.

[kelticwizard]

Well, filling in the numbers from the chart into WinISD Pro, plus assuming a typical 84 sq in for a 12 inch driver, gave us the Thiele-Small paremters we need, including Re. Re is 9.0, (8.98) ohms, which makes this a 14 ohm nominal speaker.

[kelticwizard]

When using the two formulas above to select the value of the resistor, then filling in the resulting new Qes value into Bullock and White's Boxmodel freeware, we end up with a Qts reading slightly lower than what we aimed for. The discrepancy is not great, and a good box can be built using either the aimed for value even if the accurate value is Bullock and White's.

So, I'll just list some useful Qts values.

Aimed for Qts = 0.26 Resistor = 13 ohms. Qes = 0.268. B&W Qts value = 0.24. 96.2 dB sens.

Aimed for Qts = 0.32. Resistor = 19 ohms. Qes = 0.34. B&W Qts value = 0.30. 95 dB sens.

Aimed for Qts = 0.4. Resistor = 26 ohms. Qes = 0.427. B&W Qts value = 0.37. 94.2 sens.

I give the above numbers under advisement. Intuitively, the resistor values seem too high for the comparatively small decrease in sensitivity. Adding a 26 ohm resistor to a driver with an Re of 9 ohms and a sensitivity of 100 dB only brings the SPL rating down 6 dB?

Seems shaky, but that is where the numbers and programs seem to lead. So if anyone wants to jump in, please do so.

[AndrewT]

Hi,
so much for my idea of small series R!

No @ 6.3%
We sure seem to have forgotten how to make efficient speakers today.
That kind of efficiency will cost quite a lot nowdays whereas Richard Allen were a fairly cheap range in their day.

BTW. mine were CG12 not S.

[316a]

Originally posted by kelticwizard
Well, filling in the numbers from the chart into WinISD Pro, plus assuming a typical 84 sq in for a 12 inch driver, gave us the Thiele-Small paremters we need, including Re. Re is 9.0, (8.98) ohms, which makes this a 14 ohm nominal speaker.

Hi ,
Thanks for your help so far I measured Re of a driver earlier , it's 5.8 ohms . I've also used autocomplete in WinISD to generate a model for this driver , and it's pretty much similar to yours apart from Re

cheers

316a

[316a]

Hi ,
I've just noticed that WinISD has provision for adding different values of series resistance . I've added 15 ohms series resistance , and used a 100l cabinet with 36Hz port tuning . -6dB point is 37Hz , seem to getting somewhere now , otherwise I'm totally barking up the wrong tree

cheers

316a