Tuning the Natalie P's (RS180-RS28) after building them

[mslatter]

This is a re-post from a PE TechTalk forum post. The guys there suggested I post here (duh!)

"I built Jon Marsh's Natalie P's (RS180-RS28 MTMs in a 9x22" baffle sealed cab) a few years ago after looking for a good mid-price project. I've been really pleased with them. I like the clarity and detail of the combination, and it was a pretty simple build with a series cross-over. They've been my workout speakers in my bedroom for a while now.

I've moved them into a different room and set up a bit more of a traditional listening position for them about 8 feet away, and now that I'm spending a little more detailed time with them, I hear a little harshness that I'd like to tweak out. I hear it primarily in electric guitars and saxophones. Voices sounds great, and other instruments are fine but with a little bit of volume behind it, the guitar and sax grates on me a hair, and I turn the volume down. In the past, when I've tuned car stereos, I've attenuated around 6kHz, if I recall, based on personal preference, but don't hold me to that.

So, the question is whether there's something I can do, short of getting an equalizer, to try to reduce the little bit of harshness I hear? Here's Roman B's comparison of the predicted frequency response curves for various cross-over designs with that speaker combination, by the way. http://www.rjbaudio.com/RS180MTM/rs180-rs28-mtm.html

Thanks in advance for the help!"

Suggestions from TT were to build another cross-over and to increase the dampening inside the cabinet to absorb some of the offending frequencies. I'm open to all suggestions!

[JonMarsh]

What's your system configuration look like, and what are the room acoustics like?

The NatalieP were designed to be as neutral as possible with a well behaved crossover transition and good attenuation outside the transition band.

Roman's simulation is pretty close to mine, and representative of how they measure on the listening axis.

Natalie MTM design

Please forgive me for not remembering who designed the Natalie version for these drivers but I would have to say that it is one of the most interesting crossover designs I've seen and it has a very nice looking response. The crossover is a sort of mutated series crossover that is like nothing I've ever seen before but the response looks great and the parts count isn't too bad. The details are shown below.

Ahem... :B :W OK, now that I've got that out of my system, what can we do for you?

In my experience, the things that could be causing your reaction could be impacted by several things:

• Room acoustics- a bright reflective acoustic OR positioning causing dips in the LF or midrange response will make the HF seem out of balance
• Program material- there are some classes of recordings which are mastered assuming playback systems that are far from flat (older car radios, small home systems) and use agressive EQ in the top and bottom end to boost presence. This can be very irritating- one could even compare early rock CD's against more recent re-master releases, with quite different tonal balance
• Playback gear- some DAC's and CD players have a brighter, harder sound than others- HT gear used for music playback usually have more issues in this regard. Without knowing your system, hard to comment

The logical thing to do if you want to change the speakers to "fix" your problem is to modify the crossover to introduce a downslope in the tweeter response. Actually, I'd usually recommend inserting an equalizer or tone controls into the system, adjust until you're happy, then measure the electronics EQ and duplicate with crossover changes. This is an "engineering" response to go as fast as possible to what you may prefer.

Practically speaking, what you can do that is fairly simple is ramp up the value of the resistor in series with the tweeter- you may find that scaling it up to 2-3 ohms may turn the trick for you.

What that means, referencing the original crossover, is adjusting R3.



If you have access to a test system and mic, I'd recommend measuring the system first with a short gate, to minimize room reflections, and just looking at the response from 200 Hz and up, see if it reflects the simulated response. Just to be sure you don't have something like a tweeter out of phase, for example.

I'd also think about those other factors; obviously you may not want to do a major system reconfiguration just because your speakers are a bit too analytical and clear for the way you've got them setup now, but in the long run, that may be the more beneficial approach. They do respond well to good equipment and setup, and folks have used them to A/B DACs in the $1K to $5K range with clear benefit to the more expensive models...

Good luck and let us know a bit more about your setup.

[JonMarsh]

One other thing- how much power do you have, and are you sure you're not clipping? Your description sounds like that could be a possibility, too.

[mslatter]

I can't thank you enough for responding! I spent the evening hard-wiring the power supply to my broken (thanks, kids!) mixer so I can hook up the measurement mic and do some captures tomorrow. I'd hate to think it was something as simple as an out-of-phase tweeter, but I've been stupid before, so we'll see. The room is sub-optimal, with hardwood floors and not alot of soft stuff around. The current source is a very old solid state receiver of consumer quality and unknown power, but probably not more than 60W. I'll bring down my 15W t-amp and try that, too. Sound source has been my iPhone 5C into the aux of the receiver, so the DAC isn't audiophile quality, and it's possible I'm overdriving the receiver, but it does seem to be a pretty narrow band that's annoying me, not a broad clipping issue. And I think I've been sensitive to this same range on completely unrelated equipment, too, so it might just the shape of my ears or something.

Would there be any possible benefit to increasing the dampening as suggested at TechTalk? I know I didn't surround the tweeter in a mini-box of its own as I've seen others do. And my box is probably stuffed with polyfill, but I couldn't swear it was high quality stuff or particularly dense.

It's really a small issue, because I just love these speakers otherwise. I've actually had the magic experience of re-listening to stuff I've heard hundreds of times before and hearing new things. That's been fun, and renewed my excitement for listening to music. I've never regretted a dime or a minute I put into these!

[fbov]

...Practically speaking, what you can do that is fairly simple is ramp up the value of the resistor in series with the tweeter- you may find that scaling it up to 2-3 ohms may turn the trick for you.

What that means, referencing the original crossover, is adjusting R3. ...

Had I decided to fight with the authentication (keep getting dropped so can't post) you might have received this same advice on TT... albeit from someone far less knowledgable!

The reason, in my case is that I've documented variations in RS28 sensitivity of about 1dB, and have planned to adjust R3 if I ever needed to. Haven't yet in 2 sets of NatPs, (shielded Al towers and unshielded silk bookshelves). Attached is some Fr data for my towers, close mic'ed and listening position.

Have fun,
Frank

[JonMarsh]

Personally, I really don't believe that altering the stuffing of the enclosure is going to affect the tweeter response- that could be the perceived case IF the woofers were operating up to a relatively high frequency and IF somehow reflections were getting back out the front through the cone- not gonna happen with the RS180.

Also, note and compare the roll off profiles on the woofers in Roman's simulations of the different crossovers- the NatalieP has a very aggressive trap filter in order to suppress high frequency ringing/trash from the cone (something other designers overlook at time, once they see the top level line looks nice), and the HF output of the RS180 is -50 dB from the nominal level.

Look at your setup, maybe try swapping in and out some friend's gear just for the heck of it, and if needed, try bumping up R3, start with 1.5 ohms (remember, this is in series with a 4 ohm nominal tweeter).