2x8" PPSL midtops made from junk

Gents,

here is my detailed report about the 2x 8" PPSL top

Materials:

Horns

I was given a pair of nondescript Community audio horns with a 1" eminence driver and a piezo. The piezos will not be used, because one is broken, and I can't be arsed to find a matching spare. Dimensions: 40 x 20 cm

Woofers

I was doing a restaurant install job in which I had to reuse a lot of old components that the chinese entrepreneur had left over from previous ventures. Among them was a trolley full of "Factor8" install cabs from DAS audio. A few of the cabs were in very bad shape, with the injection molded plastic shell completely cracked. Before throwing them in the trash, I pulled the woofers and crossovers and the wiring from these units, and ended up with a set of four "G8" woofers and the crossovers

http://www.soundaudiosystems.com/pdf/das-audio/tech-sheets/factor-8t.pdf

Wood

This spring I started to check the big skip containers any time they popped up in the neighbourhood for anything salvageable. Last week I got lucky with the construction site next door. They discarded a temporary door (190 by 90 cm), which had been reinforced with 6-layer 9mm ply. I had to remove a dozen or so screws, but except for the screw holes the wood was ok. I also had two scruffy, unused picture frames (probably from IKEA) in my storage unit in the basement, which I decided to slaughter to be used as battens in my build. I highly recommend this method, if you can find picture frames made from an L-shaped profile

Design:

Cabinet shape and loading

I started to juggle first ideas after being given the horns. I thought it would be nice to get some woofers that are similarly sized, ideally eight inches, which would allow me to make the enclosure slim. I can't remember at which point I first had the idea to try a PPSL arrangement. When I got the four woofers I realized that putting them on a normal baffle under the horn would either mean making the cab too tall (about 90 cm) or too wide for my design goal (slim!). The solution was to load the woofers PPSL style. This would have two benefits: a) meet my design goals regarding size and form factor b) decrease the distance between the woofers for better coupling -> higher crossover points are possible.

Woofer T/S parameters

When I got the woofers, I sent a mail to DAS audio, but they didn't have any TS parameters for this old unit. I left the components on the shelf for a while. After I scored the wood recently I decided to try my luck with impedance measurements in RoomEQWizard to see if I can calculate TS parameters myself.

First big surprise: the woofers' calculated sensitivity is 86 dB.... OUCH! Thats low. I knew that the Factor8 box is not very sensitive, but this seemed a bit off. But checking the datasheet again more carefully, it turns out that the sensitivity figure of 91 dB applies to the "3kHz octave", which leads me to believe that in this box the tweeter has the higher sensitivity rating..... I don't have a xmax figure for these G8 woofers, but the flexible rubber surround suggests that it should have some displacement capability.

Cabinet simulation

Entering the parameters into WinISD gave a bit strange results. I wasn't able to come up with a sensible reflex alignment. WinISD suggested that a closed box would be the optimal solution. Doing some more sims it became apparent that I could make the box fairly small without sacrificing the bass response. I was planning with cabinet dimensions of 25 x 25 x 60 cm, which is 32 litres external, 28 litres estimated net volume. Below is a screenshot from winisd with two traces, one for a 99 l cab (red), the other for 28 l (yellow). I would have to make the cabinet three times bigger to gain 10 Hz of extension (-3dB at 62 Hz instead of 72 Hz). Funnily enough, the 28 l trace looks very similar to the Factor 8's plot in the datasheet, with a 3 dB hump at 100 Hz. I estimate that my net internal enclosure size should be very close to double that of the Factor8.

Using parametric EQ to flatten the hump, the result is a 12 dB/octave cutoff starting at 80 Hz. Putting 49 watts into the cab should give me 106 dB SPL with 3,5 mm excursion, If it takes 100 W without bottoming out, I can eek out 110 dB SPL with 5 mm excursion.

If I ever use the cabs with a sub, I might be able to get 3 - 6 dB more output by limiting excursion with a highpass filter.

I haven't searched for information on the Eminence HF drivers yet. I suspect they should outrun the two 8" woofers by far.

Wiring

A long time before actually building the cab I had already experimented with the naked components connected to the Factor8 cabinets crossovers. I simply connected two crossovers in parallel, each woofers to it's own crossover, and the two tweeter outputs in parallel to the HF driver. It worked reasonably well. It was hard to assess the woofer/horn balance since the woofers were just sitting on the ground without enclosures.

My plan was to include the crossovers for passive operation, but also include the option of bi-amping for increased performance.

Result (temporary!)

At the moment I have connected the components as described in the previous section about wiring. The HF horn dominates the woofers by about 6 - 10 dB. I had to stuff a piece of foam into the horn to make it balanced. Next time I open the box I will remove the connection to the second crossover and see if that will drop the level. Looking at some old measurements I did of this horn, it has a 10 dB hump between 1,5 - 4 kHz, to which I attribute the shouty sound.

As stated in the opening post, this project is about using random components I had, and is basically a (more or less) sophisticated exercise in turd polishing. Let's see where I can get with this.....

Still to do:

- Assembling number two

- ordering a pair of fancy adjustable angle pole hats

- install speakon connector panels and passive/bi-amp switch

- router/sand/polish/paint cabs

Processing

Once I get the inuke, I will start to work on some presets for both passive and bi-amping, utilizing EQ & dynamic EQ. The catch is of course that if I want stereo, I have to run in passive mode, while for bi-amping, I will have to go mono (or buy a second amp).

Edited by bob4 - 18 June 2017 at 8:26pm

Thank you for the compliment mini-mad

gen0me wrote:

mini-mad wrote:

This is the most fun I've seen anyone on here have in a long time!

+1

Tell us the ts parameters. Ill sim in hornresp. Rear chamber is 28dm3?

Ok, here are the Thiele/Small parameters calculated by REW:

RE 7.18 ohm
fS 56.1 Hz
MMS 22.99g
Zmin 7.58 ohm
QMS 7.065
CMS 0.350 mm/N
fmin 177 Hz
QES 1.246
RMS 1.146 kg/s
f3 1,037 Hz
QTS 1.059
VAS 17.95 litres
LE (f3) 0.821 mH
FTS 52.9 Hz
Bl 6.833 Tm
Dd 15.55 cm
LP 86.10 dB (1W/1m)
Eta 0.25 %
Sd 190.0 cm2

Added mass 3.90 g

I recalculated the box volume, looks like I was overestimating the internal net volume a bit.

My new calculation yields 30,4 litres (gross internal volume, including plenum, drivers and horn). Plenum size is 8 cm x 23 x 23, so about 4 litres, minus half a litre worth of corner cutouts, so 3,5 litres of net internal plenum volume. Assuming 3 litres of displacement by the other components, we arrive at 30,4 - 6,5= ~ 24 -25 litres of net internal volume.

Which one is real?

Does it play closer to hornresp simulation or winisd simulation? Does it have this high mountain around 400hz?
Could you make some fft measurements?

Edited by gen0me - 20 June 2017 at 3:42pm

Right but changing s1, s2, s3(they are ~) will just move the mountain. Even changing s2 for serious value due to the magnet. So is hornresp right or doesnt it totally fit for such enclosures?