Project M2 DIY Thread

[sguttag]

In the case of the 110, the cinema version only has 4 AEC components BUT we get 128 I/O network streams instead of 64. I'm told that the differences between the "C" versions of the cores will continue to diverge as they optimize each for their respective markets.

But back on topic...So they just released designer 7.1.0 for Q-SYS. I'm reading the Release Notes (because I do such things and I happen upon:

Loudspeakers: Custom Voicing FIR filtering supportThe Custom Voicing component now supports FIR filtering, adding highly accurate and complex EQ processing when using CXD-Q and DataPort-enabled amplifiers with 3rd party manufacturer loudspeakers.

So stay tuned there. I see more experimenting in my future, this weekend!

Then again...I may have spoken too soon. The next item on the list shows that the CXDQ 4-channel amps now support 1024 tap FIR filters which would come up short. I suspect that the DSP processing power in the amplifiers are significantly less than the cores.

[lesliew]

When I bumped into the ceiling, I was doing other processing and had several Pro Logic blocks in the design (they are, currently, just in the "cinema" version of designer so don't go looking for them or the DPA-Q amps in the inventory.

Steve

That command line switch worked perfectly and I was able to use the Pro Logic block on my core.

BTW 1 Pro Logic block added 1% to the utilization.

[lesliew]

Loudspeakers: Custom Voicing FIR filtering supportThe Custom Voicing component now supports FIR filtering, adding highly accurate and complex EQ processing when using CXD-Q and DataPort-enabled amplifiers with 3rd party manufacturer loudspeakers.

Opportunistic that they should implement that days after we were talking about it. Although 1024 taps wont be anywhere near enough for this particular application.

Its certainly possible that the hardware is different but more than likely its just a re-allocation of resources, it is a Linux kernel after all.

[sguttag]

So, as I transition from "prototype" to "production" on the M2(ish) clone, we made some improvements to the rigidity of the cabinet (it is thicker now) and some internal reinforcements. There are very few and significantly smaller vibrations on the cabinet, regardless of frequency/volume.

However, while doing some final port tuning and vibration testing, I'm noting that I'm getting air leaks from the compression driver itself when playing bass tones. It isn't coming from the horn to cabinet junction (and that is well gasketed) but right through the bug screen of the D2 compression driver. If one put's their finger up to the output of the D2, you can "tune" the noise coming from it. Has anyone else come across this?

Another "feature" we put into our version of the cabinet is a rear door for the HF section so it is much more serviceable. This is not the source of the leak and the air leak from the D2 was verified on the prototype, which didn't have the rear access panel.



Incidentally, for the LF driver (2216nd), I did use the JBL 360-70002-01 and it worked well. I didn't detect any leaks around the LF section (or anywhere else that shouldn't be leaking). Just the D2.

[Robh3606]

Have you tried putting a light inside the cabinet in a darkened room? If air is getting through light should as well. We backlight hermetic enclosures to look for light leaks. That seems really odd the only thing I can think of is a gasket leak between the driver and the horn.


Rob

[DES-1]

the only thing I can think of is a gasket leak between the driver and the horn.

This. Did you torque down the horn sufficiently? It's a pretty thin vinyl gasket, at least on my D2s.

[sguttag]

Yes, it is a thin gasket but typical of HF drivers and yes, I think it is well torqued

I like the light idea and will see if that yields anything. I'm also thinking about "bagging" the D2 to make sure nothing is seeping in via the many, many layers to it.

Note, the adapter flange is what mounts onto the waveguide and there isn't any gasket there. That is just metal to plastic and depends entirely on the manufacturing tolerances of the two. I'm using the factory flange too (and yes, it is mounted properly). I'm surprised there isn't even a paper gasket there. Then again, even the LF driver doesn't show a gasket in the exploded view either. For that one, I did add a gasket. That is the only area I can think of that air could leak out.

At least on my version of the M2, getting to that flange for removal is easy to get to. I might have an Altec thin rubber gasket I could get to fit on that flange to waveguide junction.

-Steve

[lesliew]

^^^

I like the idea of the open back, the horn is a real pain to mount in its standard form.

Here are some photos of my build, nearly ready for paint. We are using the original grills on this build so I had to modify the CNC'ed curved baffle slightly. In my original build since I was mounting my M2's behind an AT screen I wasn't bothered about sourcing the grill so used 0.75" thick MDF to make the curved baffle. As it turns out the grills profile actually required 1" thick mdf to be used so that the highest point on the curve runs in line with the grill. Small modification but details matter.

Also used 1" thick birch plywood to make these cabinets which was a lot nicer to work with only problem with plywood is you occasionally hit a void, hence the bondo.

[sguttag]

Those look sweet!

I see you solved the air leak from the HF section too! I never thought to use the packing bag as an air gasket! Genius!

[lesliew]

Yea I should patent that idea.

[srm51555]

Nice work!

[sguttag]

I was correct, as it turns out, air was leaking through the adapter plate to waveguide junction. If you all don't have a gasket there (and are using the factory adapter plate) then you most certainly have a leak in your system too.

Below, you can see there is a gap between the end of the waveguide and the height of the adapter plate so the gasket that is part of the D2430K will never seal that junction.



I decided to make a gasket by cutting a thin strip of gasketing material (sourced from Parts Express that I had already) that is about 1/8" thick but I slit its width to about 3/16" wide and then formed it around the waveguide.



You need to have it no wider than 3/16" to allow it to squish out a bit when the driver presses in on it. Honestly, I'd prefer to have a shallower gasket as 1/8" is probably a bit tall.

The results were quite satisfactory, however:



I detect no leaks from the cabinet now (from unauthorized holes, that is) and I feel that acoustically, it got a bit cleaner with this too (in the HF department). Possibly because the driver is better coupled with the waveguide more than any detrimental bass effects getting to the HF, but other people would know better than I.

Looking on the M2 exploded diagram, they DO show what appears to be a gasket but don't list any part number. In fact, the line that comes down to what seems to be a gasket calls out the #10 flat washers, not a gasket.



Thus far, I have not come up with a JBL part number for this gasket.

The 2509A horn mounting bracket came with a gasket so it could be used with the 1.5" throat drivers/horns its dimensions were 2" OD, 1.5" ID and 1/8" thick but it was a rather stiff rubber that was used to fill in where the metal was in the horn bracket. I don't think it would work as well in this application because it wouldn't give like the gasketing material I used. Furthermore, I don't have a part number for that gasket since it was not listed by part number with the bracket. I suspect that 1/16" gasketing material would be ideal and cut it to just shy of 1/4" but what I did works effectively.

[pos]

Great finding!
Would not a thick paper do the job?

I have some 2450SL gaskets here (#74113) but these are self adhesive and are larger than the D2430K front plate.
These are 1mm thick uncompressed with the adhesive protection layer in place.

[sguttag]

Great finding!
Would not a thick paper do the job?

I have some 2450SL gaskets here (#74113) but these are self adhesive and are larger than the D2430K front plate.
These are 1mm thick uncompressed with the adhesive protection layer in place.

I don't know. Paper gaskets generally do best when sealing two surfaces that have very minor imperfections such as two machined surfaces. They don't have much "give" to them so they also act like a spacer as well. As such, you will almost always find paper type gaskets spanning the fastener area so that the force of the fastener is compressing the paper and thereby creating the seal.

If the paper were inside of the mounting area, as in this case, you'd have to get the thickness just about exact. That is, it would have to be just above adapter plate but such that it would be within the paper's compression range, otherwise, the paper will create more of a spacer that will prevent the waveguide from properly mating with the adapter plate and thereby putting stress on the waveguide/adapter plate junction. Depending on what gives and by how much, you could set yourself up for vibrations between the plate and waveguide.

It could work but I wouldn't recommend it. Getting a foam type gasket or, ideally, the right thickness rubber gasket is going to be the best way. That rubber gasket would only need to be as thick as a typical flat washer for a #10 screw (that is the only "measuring" device I had when I was putting this stuff together over the weekend). So, if your paper could compress enough, then yes, it should work but I'd feel better about rubber.

[Robh3606]

There is no gasket on the face of the D2 driver?? Reason I am asking is there usually is something paper on the 2425/6 as an example.

Rob