Project M2 DIY Thread

[tommytoa]

I asked one of the engineers behind the M2 project directly and he confirmed the different versions of the D2430k is the same, but those for M2 is just cherrypicked for frequency response. I suppose I have non-M2 drivers because the two samples differs some in frequency response. A measurement microphone and some EQ solves that problem.
Unless one of them is wired incorrectly, of course. I feel the need to check..

One other thing that I found strange was the frequency response with Pos' Minidsp/Rephase settings. Here is the near field of the tweeter with his settings compared to tweaked settings I did with Hypex DLCP. Pos in blue;

[withTarragon]

Given the scale on the plot, it looks like those differences are +/- 1dB. That is actually pretty good, perhaps very good. In my experiences small differences measured at the high end can seem to come out of nowhere (sneeze and a minute later there is a dB difference).

[tommytoa]

Yeah, multiple measurements are needed to confirm when your aiming for accuracy. I'm not sure Umik-1 is the right tool either, but I find the original M2 settings to be too bright. They elevate all ssss-sounds to a degree I just can't live with.

Loading the rephase settings into minidsp 2x4hd made the speakers incredibly anemic with what seemed to be far too little bass. I'm not sure where I went wrong, but I made it sound better with manually entering the values in Hypex DLCP. It seems like the shelving is different though, so I'm still working on that.

I'm way too stupid to figure out how to measure the output of the Hypex dlcp to confirm anything. It involves a card with sound according to POS, and I gave up at that

But I have a question for anyone who is smart; For some reason I need to lower the right tweeter with 6 dB more than the left to match them (and reverse the phase). Both passive filters look ok and wired correctly. How could one tweeter be 6 dB louder than the other if the output of the dlcp and the amps (I've tried swapping them)? Could it be anything other than the passive filter gone wrong?

[grumpy]

Not sure what you swapped and what the result was.
If the problem follows the tweeter... perhaps one of the two internal diaphraghms has an issue (?)

If the problems stays with the DIY M2, you can continue backward in the chain, swapping speaker inputs, amp outputs, DSP inputs, until something changes.

Caveat emptor regarding the 'smart' bit.

[withTarragon]

.................
..........
But I have a question for anyone who is smart; For some reason I need to lower the right tweeter with 6 dB more than the left to match them (and reverse the phase). Both passive filters look ok and wired correctly. How could one tweeter be 6 dB louder than the other if the output of the dlcp and the amps (I've tried swapping them)? Could it be anything other than the passive filter gone wrong?

I don't have a Hypex DLCP ( I am not even sure what that is).

However, do you have an RMS Voltmeter (it does not need to be a fancy one)?
Are you or a friend comfortable measuring the voltage (VAC) leading to the tweeter? Do this on each tweeter to verify a difference (6 dB is a doubling (or halving) of the voltage). Then work your back in the chain.

Good luck,
-Tom

Edit: I see Grumpy beat me to it ...

[tommytoa]

Thanks guys!

Update;
Hm, something strange is in the works here. I tilted the right speaker and placed it back down and now the right tweeter is only 2 dB hotter than the left tweeter. I can't be sure it's the wires or something fishy with the pre because I was calibrating and uploading tons of different settings to the left speaker through the Hypex DLCP dsp/pre for measurements and testing.

Well well, it seems like I have to open up and check wires and stuff. I'm a total noob when it comes to electrical stuff, so not finding myself very motivated to do surgery but I guess I have to bite the bullet and check.

Question;
How does the speaker measure in near-field with the standard M2-presets within BSS/Crown with the mic aimed at the JBL logo? (design axis according to Sprinkle).
Does anyone have REW measurements of that that they can send me? I'm interested in very near-field measurements of the tweeter and the woofer with the original settings as well so I can compare especially the slope implementation of Hypex.

I'm getting very different results from the listening window measurements than what JBL shows in the marketing material if I use the original settings as given by POS.

[pos]

The listening axis is indeed right between the horn and woofer, but you cannot use that for nearfield measurements as you would get a much higher angle that way that what you would get at typical/recommended listening distances, which would drastically affect HF response.

As for the listening window measurement, having it really flat in the UHF involves having a rise in the on-axis response, as illustrated by the official spinorama.
Chances are that this particular spinorama is not representative of the final product, as no one has ever measured a rising UHF response on an M2... (and this is probably a good thing )

[Robh3606]

Chances are that this particular spinorama is not representative of the final product, as no one has ever measured a rising UHF response on an M2... (and this is probably a good thing )

Hello POS

No kidding! So when you put in the factory presets the stock system is flat on axis? I didn't know that good to know! Are you measuring of the logo axis or on the waveguide throat. I found the it doesn't take much going below the throat to see a drop 10-20K. Have you seen the same??

Rob

[tommytoa]

Good point! I'll measure straight in front of the tweeter @ 1 meter and see how that looks.

[pos]

Hello POS

No kidding! So when you put in the factory presets the stock system is flat on axis? I didn't know that good to know! Are you measuring of the logo axis or on the waveguide throat. I found the it doesn't take much going below the throat to see a drop 10-20K. Have you seen the same??

Rob

Hi Rob

That spinorama might come from an early prototype, but might not have passed the listening test...

The UHF response drops off axis, but that is already pretty good for a 1.5" throat. Clearly not as good as a 2344, but also better behaved.


Here is what I measured a while bask, 1m on each driver axis: http://www.audioheritage.org/vbullet...l=1#post392775
This uses a wide dB scale as what I intedend to measure here was the crossover.

The elevated response around 10kHz is there, similarly to what tommytoa measured, but given the date I think this measurement was made with the old rephase preset, which was a direct replicate of the crown settings as exposed in the document.
A few days later thanks to grumpy and sebackman and was able to compare it to the actual BSS response: http://www.audioheritage.org/vbullet...l=1#post393277
and adjust rephase settings accordingly: http://www.audioheritage.org/vbullet...l=1#post393453

EQ settings interpretation and biquad calculation do vary quite a bit from one unit/brand to another, so that is why the best thing to do is to measure the actual output of the device.

[tommytoa]

Ok, testing now. Measurements @1 meter on-axis confirmed with laser.
First is the close-field measurement of Rephase settings vs 1 meter converted settings according to POS' brilliant crossover guide;

1/12 smoothing


Here is the converted settings @ 1 meter vs tweaked manual settings to get a linear response. Changed the slope from 1,35 to 1,135 to avoid banana-shaped response;


I have to experiment a little with the response around 1500 hz because that region may suffer from cancellations between stereo pairs depending on seating distance from both speakers.

[tommytoa]

Here is the slope difference between 1,35 and 1,135;

[tommytoa]

And here I tested close-field (inside throat) at 110 dB to calibrate driver spl and check for distortion differences.

[pos]

You really need to measure the output of your DSP and adjust from there, as it is both easier and (much) more accurate than measuring the drivers + room.
By the way, do you have an accurate calibration file for your mic?

[tommytoa]

You really need to measure the output of your DSP and adjust from there, as it is both easier and (much) more accurate than measuring the drivers + room.
By the way, do you have an accurate calibration file for your mic?

Yeah, about that. Too difficult for me to understand how to do that. I tried searching the Googlenet, but too much noise on this interweb stuff!

I'm contemplating buying some Crown CDi amps to get the standard tuning experience, but am unsure about the fan noise in the CDi 2/600 amps..