Scaling up the SS15 for an 18" driver and other compact 18" designs

[col]

Starting this thread after a need arose from the "SS15 build x8" thread (instead of hijacking it). The initial idea being explored is whether the SS15 could be scaled up to take a 18" driver.

Xoc1 has input some time into hornresp and come up with a concept that can be seen in this post:

http://www.diyaudio.com/forums/subwoofers/202797-ss15-build-x8.html#post2858583

$$Maan is very interested and is preparing to make some sawdust. I had been thinking about the concept but opted to build a Cubo 18 extended as a prototype. I'm not sure if it is going to play low enough into the 40hz range yet though so I am still very interested.

My work so far:

rebuilding the 18"s

regards,
col.

 

[salzburgsoundsystem]

Of course, this does raise another question...

Great idea!

but this does raise a question:

Which 18 will be the all-rounder / benchmark that the 3015lf once was, and which price category should we be considering?

Regards, Ben

 

[m R g S r]

A slightly scaled up ss15 would be the TH-18...very similar folding just sacled up.

Unless you mean just widening the enclosure to fit an 18" driver? Many 18's are much more powerful than a smaller 15, and would probably flex the thin 1/2" walls very easily. So then you get into re-designing the enclosure for 3/4" wood...I think that's what pyro did for the pyroboxes...

 

[m R g S r]

What's the compression ratio on the Cubo? It doesnt look to be like much at all? And path length looks pretty short for 40hz..

 

[Xoc1]

The Cubo is a hybrid tapped horn with an added bandpass chamber.
The chamber makes a big difference to the path length needed,
A similar chamber can be used with a front loaded horn, and is used by Turbosound and Funktion One in their horn loaded bins.

 

[col]

Great idea!

but this does raise a question:

Which 18 will be the all-rounder / benchmark that the 3015lf once was, and which price category should we be considering?

Regards, Ben

Good question, I've already got 4 x Peavey low rider 18"s so I'm kinda of stuck with them. It would be good to use a reasonably priced driver but depending on where you are and shipping costs that differs. At the moment in Australia the best value for money is the Chinese Ande drivers. I guess in the US Eminence would be the first choice and in the UK Precision Devices or B&C.

Maybe that is why the Cubo designs are so popular as they work with a wide range of drivers.

col.

 

[m R g S r]

The Cubo is a hybrid tapped horn with an added bandpass chamber.
The chamber makes a big difference to the path length needed,
A similar chamber can be used with a front loaded horn, and is used by Turbosound and Funktion One in their horn loaded bins.

Okay gotchya. I was thinking it didn't look like a traditional tapped horn.

 

[col]

I guess the trick here would be working out a design where the bandpass chamber can be tuned easily to suit the chosen driver. Also, how does one determine the chamber size and tuning port?

 

[col]

Peavey Low Rider in Cubo18ext SPL and FR tested on Grass

I did some testing today. results were a bit disappointing. Basically, the Cubo18ext is not suited to the Peavey Lowrider. It can't go anywhere near low enough. Also, measuring the original 130ltr BR cab showed why the cabs need changing! Oh well, back to the drawing board.

First up I did a impedance test with WT3, magnet in the horn:

This gave a good indication of whats to come

Second up a holmimpulse frequency response:

I compared this with the original 130ltr BR cab (in red):

Then I did some 2.83v/1M SPL tests:


Cubo18ext 130litreBR
40hz 69.2 68.7
45hz 81.9 75.3
50hz 70.4 67.8
55hz 79.5 76.7
60hz 84.8 80.0
65hz 88.6 84.3
70hz 89.0 85.6
75hz 89.2 86.6
80hz 89.0 86.9
85hz 89.0 86.4

 

[col]

The measurements did indeed show at least a 3db increase in volume over the badly tuned 130ltr BR cab though. Also, with all that data and the actual drivers TS parameters it should be possible to work out how to model the cabs in Hornresp?

All the data is in the photo gallery:
rebuilding the 18"s

cheers,
col.

 

[Xoc1]

Hi Col
Your sensitivity figures seem to be low.
It is still OK for comparison purposes though.
The Peavey Lowrider 18 is quite efficient in a BR box - although the one you have seems to be tuned quite high! If I try to sim the BR box with the port tuned for a 80hz hump, the peak at 80 hz is 105db @ 2.83V - 2pi
I tried a sim with the Lowrider in the Xoc1 TH 18 and its fairly poor with a big dip at 80Hz.
I simmed the Lowrider in a 200 litre reflex tuned for a flat response, and compared that with the Cubo Sub (not the Cubo 18 as you built), and the JBell style 18 TH that I drew up. This shows the differences in efficiency. To me, with your driver, the Cubo Sub wins!
Regards Xoc1

 

[weltersys]

The measurements did indeed show at least a 3db increase in volume over the badly tuned 130ltr BR cab though. Also, with all that data and the actual drivers TS parameters it should be possible to work out how to model the cabs in Hornresp?

All the data is in the photo gallery:
rebuilding the 18"s

cheers,
col.

Both cabs can be modeled in Hornresp.
Glad to see someone actually measuring :
Cubo18ext 130litreBR
40hz 69.2 68.7
45hz 81.9 75.3
50hz 70.4 67.8
55hz 79.5 76.7
60hz 84.8 80.0
65hz 88.6 84.3
70hz 89.0 85.6
75hz 89.2 86.6
80hz 89.0 86.9
85hz 89.0 86.4

The measurements show a pronounced 10 dB dip around 50 Hz, probably from a reflection off a nearby wall, possibly a mic problem.
Your test mic should be placed on the ground to avoid path length errors, with close measurement distance, nulls can be large.

Both the Cubo and the BR cabinet appear tuned around 45 Hz.
The Cubo would require more folds to tune lower, but will loose sensitivity down low if that is done. The 6.6 dB advantage at 45 Hz over the BR would probably decrease to around +3 (or less) if the Fb is lowered to 40 Hz.

Same holds true with the BR, longer ports will tune lower, but with less output.

Hoffman's Iron Law applies to all types of enclosures- loud, low, small, pick two.

TH can provide about a 6 dB gain over a BR cabinet, but require a larger cabinet to achieve that much gain.

If you want to keep box size small and LF output below 40Hz is your main goal, you are better off with a speaker with TS parameters like the Eminence Lab 12 , a pair can go low and loud in a smaller BR box, but won't have the upper bass output of the 18" Low Rider.

You can see in the chart below that almost doubling cabinet size is required for a 3 dB 40Hz increase, and that comes with a loss in the 60 Hz range.

Art

 

[col]

Thanks for your comments. My MiniDSP balanced kits arrived yesterday. Once Iv'e got them boxed up I will do some more tests. I'm wondering what can be achieved with a bit of EQ'ing on the Cubo18ext. Another thing that I'm contemplating is extending the shelf that forms the port in the Cubo chamber, maybe I can lower the tuning. If all this fails will build a Cubo Sub and do more comparisons.

Xoc1, I noticed that the SPL reading were low for 2.83v too. Could that be because of the Re being 5.5ohms? THe actual TS parameters of this driver are along way off of the factory spec:

cheers,
col.

 

[weltersys]

. I'm wondering what can be achieved with a bit of EQ'ing on the Cubo18ext.

cheers,
col.

The Cubo you built appears to be "off the pipe" below 45 Hz, excursion probably goes up like a skyscraper below that point, only the suspension controls the excursion below Fb.

Paint a white dot ( White-Out, Liquid Paper, silver Sharpie pen all work well) on the cone, use sine waves at around 40 volts and measure excursion at various frequencies. You may find the cone runs out of Xmax even at that level (around 200 watts) by 40 Hz.
Once you determine excursion, you can set the HP filter accordingly, then boost without "flapping" the cone.


As far as the "too low" spl readings, when you place the mic on the ground plane you will see some increase in level, that coupled with the SPL meter perhaps reading a bit low may account for the difference.

Art

 

[col]

It seems that the people having most success with Cubo 18 are using the PD1850, which looks like a very serious beast

Precision Devices International Limited

comparing the TS I notice that the VAS is almost half. Would this be the main parameter determining how the driver tunes in the chamber? I'm wondering if I get another 18" driver with much lower VAS if it will go lower in Cubo 18?

Art, will try the "white-spot" technique!

cheers,
col.

 

[weltersys]

Hi Col
The Peavey Lowrider 18 is quite efficient in a BR box - although the one you have seems to be tuned quite high!
I simmed the Lowrider in a 200 litre reflex tuned for a flat response, and compared that with the Cubo Sub This shows the differences in efficiency.
Regards Xoc1

If you raise the 200 liter BR tuning to match the F3 of the Cubo Sub,("apples to apples" if cabinet size is also matched) then compare impedance curves, (the Cubo sub is probably still lower impedance, so draws more power at the same voltage than the BR) you may find the Cubo Sub is not so different in “efficiency” to the BR.

 

[Xoc1]

If you raise the 200 liter BR tuning to match the F3 of the Cubo Sub,("apples to apples" if cabinet size is also matched) then compare impedance curves, (the Cubo sub is probably still lower impedance, so draws more power at the same voltage than the BR) you may find the Cubo Sub is not so different in “efficiency” to the BR.

Hi Art
I measured the volume of the Cubo Sub and it is about 300 Litres.
I have simmed it against a 300 Litre BR cab , "apples to apples"
Hornresp has an efficiency graph so no math required.
As you say there is no free lunch here, just choices to be made!
BR is grey line, Cubo Sub is black.
Regards
Xoc1

 

[weltersys]

Hi Art
I measured the volume of the Cubo Sub and it is about 300 Litres.
I have simmed it against a 300 Litre BR cab , "apples to apples"
Hornresp has an efficiency graph so no math required.
As you say there is no free lunch here, just choices to be made!
BR is grey line, Cubo Sub is black.
Regards
Xoc1

The drooping upper response of the BR Hornresp predicts for the BR will be about half in reality.
From a smooth FR standpoint, I'd take the BR, assuming the excursion looks similar.

At 300 liters (about 10 cubic feet), the cabinets are not so small, might as well "man up" to a 15.5 cubic foot Keystone or your TH-18 and get down to 35 Hz.

It was interesting to hear PASC's comparisons of the two designs with the same speaker.

 

[col]

The drooping upper response of the BR Hornresp predicts for the BR will be about half in reality.
From a smooth FR standpoint, I'd take the BR, assuming the excursion looks similar.

I am not so interested in the upper response as I have 4 x 15" kick-bins that take over around 90hz. The Cubo18ext I built is roughly 250ltr. If I compare a 200ltr BR the BR wins hands down, well on FR anyway. It's looking like I might have to pull my head in and build the humbler and more bland BR cabs.....

At 300 liters (about 10 cubic feet), the cabinets are not so small, might as well "man up" to a 15.5 cubic foot Keystone or your TH-18 and get down to 35 Hz.

If I "man up" to TH-18 it won't fit in my SUV, and the Keystone will probably do my back in, then it will be "man down!".

Would it have been possible to scale up the SS15 with the Low Rider or is it just not a very horn capable driver?

cheers,
col.

 

[Xoc1]

I am not so interested in the upper response as I have 4 x 15" kick-bins that take over around 90hz. The Cubo18ext I built is roughly 250ltr. If I compare a 200ltr BR the BR wins hands down, well on FR anyway. It's looking like I might have to pull my head in and build the humbler and more bland BR cabs.....
If I "man up" to TH-18 it won't fit in my SUV, and the Keystone will probably do my back in, then it will be "man down!".
Would it have been possible to scale up the SS15 with the Low Rider or is it just not a very horn capable driver?
cheers,
col.

Hi Col as you note the Peavey 18 Low Rider is good in a BR box, and not so good in a Horn.
The initial scaled up SS15 design was not brilliant with the Peavey - The TH18 sim was even worse! The Cubo sub sim showed some promise, the throat chamber seems to help produce a flatter response.
I have updated the drawing with a bandpass chamber, and run a few sims.
This one seems forgiving of increasing the Le value which is likely to be bigger than the quoted spec.
Internal volume is 340 Litres.
BTW with a driver more suited to to a horn, the response with this design is not flat at all.
There is no optimisation for sheet sizes. External size is 95 cm high 57 wide and 72 deep. Hopefully SUV sized.
If its not worth it, then building a BR box tuned to the response you would like is your best option.
If you want to explore this option I could produce a dimensioned drawing, and maybe a Akabak sim. Grey line in SPL chart is first attempt with the Peavey driver.
Regards Martin (Xoc1)