DIY M2 Questions

I built a pair of M2s lately and I have a few questions. I quickly checked the tuning and it seems to be 24 Hz, not 27. The ports are 8 3/4" or 22.2 cm long with a rounded outlet and the internal volume should be really close to 140l. I used a rigid type of damping material inside, but I know I didn't put in enough. It's probably only 3/4 thick. Would that cause the tuning to be too low? I'm going to play them a bit and check the tuning again.

Is anyone using an Xilica DSP? I am using an XP-2040. I adjusted everything so it was pretty flat according to REW, however it sounded a little strange. I removed some of the input filters and it sounds better to me. Right now they sound as good as my 4345 clones.

Here are some pictures of the flared ports. I don't have access to a CNC to make the lower part of the baffle round, so I just rounded the edges. I didn't want to try to do it by hand.

Beautiful looking cabinets… Very nice looking speakers!

Your tuning could be off because of the amount of fill, the type of fill, slight variations in the volume of the enclosures, and the exact profile of the flared ports. Another thing that can change your tuning is having minor air leaks. Looking at the beautiful construction, I assume your cabinets are airtight including the mounting of the horns. That said, a very small air leak will change the frequency of the tuning.

Regarding your settings on the DSP and the subjective sonic performance, I have found in both analog and digital systems adding too many filters to correct small anomalies has a negative effect on the sound quality.


Widget

I will add more fill and hopefully that will change the tuning. It probably won't happen for a few weeks.

Does your EQ response look something like this?:

http://www.audioheritage.org/vbullet...l=1#post393453

(assuming you also duplicated the passive bits).

Unless you were able to duplicate the port profiles, end to end, I would not be surprised if that was the cause of most of the tuning shift from the expected value.
There are also different methods to determine "port tuning"... impedance minima, minimum cone motion, ... they don't always line up. I don't know how JBL or others here who have made successful clones measured this... they might have more insight.

Cabs look very nice! I expect you're aware of all of this as the DSP settings and plans have been pretty thoroughly flogged.

Nice build!

I have replicas with the Crown 4x1200 cdi with JBL provided M2 presets included and have compared the manual settings with the built-in presets.
Interestingly enough the built-in preset gives the equivalent of a +8 dB low-shelf at 135 hz instead of the -8 dB high-shelf as stated in the document by POS.
Adjusting nothing else, that change is the only change needed if I enter manual values to match the frequency response of the M2 preset. Don't know why this is the case, but it is. The built-in preset is made in the FIR section of the Crown amp and I can't decipher those numbers, only measure the speaker output.

So I would probably start there and see how that sounds. Also, there's actually two presets for the M2; one full-range where there's a boost at 21 hz and one with the 27 hz tuning and probably a subsonic filter.

A -8dB higshef is the exact equivalent of a -8dB lowshelf at the same frequency and Q, with a 8dB gain difference.

Quote:

---

Originally Posted by pos

A -8dB higshef is the exact equivalent of a -8dB lowshelf at the same frequency and Q, with a 8dB gain difference.

---

This reminds me of the time when I thought I was being clever and interposed high shelf and low shelf on a DSP box. No problem with the gain, but the CF in one case was a 3 dB down point and in the other case the CF was 3dB up point. IOW, the CF had to be re-specified. It makes sense now, but at the time I was scratching my head for a couple of hours (I was slow that day ....). I have no idea how the Crown defines the CF, but ....... sometimes it pays not be too clever.

Oups I meant to write "A -8dB higshef is the exact equivalent of a +8dB lowshelf" in my response.
Gotta avoid typos with those things... :banghead:

Quote:

---

Originally Posted by withTarragon

This reminds me of the time when I thought I was being clever and interposed high shelf and low shelf on a DSP box. No problem with the gain, but the CF in one case was a 3 dB down point and in the other case the CF was 3dB up point. IOW, the CF had to be re-specified. It makes sense now, but at the time I was scratching my head for a couple of hours (I was slow that day ....). I have no idea how the Crown defines the CF, but ....... sometimes it pays not be too clever.

---

Yes good point, the "corner frequency" definition of shelves can vary from implementation to implementation, this is even worse than EQ conventions.
In rePhase the corner frequency is the middle of the shelf (eg -4dB for a -8dB shelf), thus symmetrical.

Quote:

---

Originally Posted by pos

Oups I meant to write "A -8dB higshef is the exact equivalent of a +8dB lowshelf" in my response.
Gotta avoid typos with those things... :banghead:

---

You did have me scratching my head on that one wondering where you were going with that...


Widget

I read that adding more damping material lowers your tuning, so I'm not going to do that. I'm going to buy DATS V3 and get a more accurate measurement. That will be later in the year, Summer is still here and there is better things to do. I hope I don't have a leak anywhere, if I do it would be around the horn or woofer, not the box. They sound really good, so it's not a priority.

Here are my filter settings from the Xilica.

They look great!

I use impedance and phase to find Fb. Can't you shorten the ports?? Can't tell if they are glued internally or not guess you have to pull the 2216 for access.

Rob:)

They are held in with epoxy, so they are pretty permanent. It might be difficult to cut them with all of the bracing in the cabinet. I might have to use a hack saw blade.

When I run a slow sweep there isn’t any cone movement at 24,25, and 26 Hz, it starts moving again at 27 Hz. I’m not even sure if this is a problem?

I also thought I could flare the ports more to shorten the tube length. I could run a 30 or 45 degree router bit in the hole and make the flare V shaped, not round. The two articles that I have read contradict each other.

I think it may be more trouble than it's worth. You don't want to do a hack job on them. Can you run Imp/Phase plot on them?? I use the 0 phase crossing point to determine Fb. If it's only a couple of Hz is it really an issue?? I doubt you could hear any difference especially that low.

Rob:)