New Natalie P build RS28A replacement

[ChadC]

I’m considering building a pair of Natalie P’s, but of course the RS28A is no longer available. Would the RST28A make a suitable replacement or should look for something else?

[Evil Twin]

As the original designer of the Natalie P, I unfortunately cannot recommend the RST28a.

The original RS28a was made by Usher for Parts Express. The RST28a is made by a subcontractor in China. I have tested and evaluated them.

Issues:

• Many specifications are missing, including Xmax, which is critical for determining usage at lower crossover frequencies such as the Natlie P employes
• The Fs is at a higher frequency, 800 Hz, which will interact with the crossover differently- the original RS28a is 600Hz.
• The RST28a does not measure as smoothly in my experience, and it may be that the impedance bump/anomaly at 2500Hz (can be seen in the data sheet) has a bearing on mechanical resonances influencing the response.
• The sensitivity is about 2.5 dB higher, which will put the balance off using the original crossover, and hints at a reduced Xmax.

All is not lost, however... as Yoda has said at times, "There is another..."

Consider the SB Acoustics SB26CDC-C0004.

• The sensitivity is the same as the RS28a- 90 dB.
• The impedance curve and resonance are very similar.
• The SPL response is quite smooth, and polar behavior is good, having a phase shield with similar effect as the RS28a.

SB26CDC-C000-4.pdf

[ChadC]

Thank you! Would anything need to be changed on the driver?

[Evil Twin]

The mounting dimensions for the driver are slightly different- the mounting plate outer diameter is 100mm for the SB26CDC versus 104mm for the RS28a. The magnet assembly is about 3 mm smaller, 70mm versus 73.

Note, my own RMS averaged forward response over a +45 deg to - 45 degree range shows very good overall response behavior. This measurement is with very little smoothing, only 1/48th octave.

[ChadC]

Thank you for the replies. I can’t wait to get started I build a pair of Natalie’s in 07 or 08, now my daughter wants a set. I better get going on the new ones or she’ll steal mine.

[Scareurpasenger]

An update to this affordable design would be quite useful to convert relatives to higher end audio. How substantial is the update to the crossover?

[Evil Twin]

My intent was to advocate a tweeter requiring no updates to the crossover. I believe the SB26CDC-C0004 meets that requirement. I have six on hand, and have had several system design ideas for it, and have done measurements with the included phase shield and with it removed, for potential use in a waveguide. Doing a POC single cabinet build to confirm is a reasonable idea, but the work queue is currently deep with higher priority efforts...

In fact, if ChadC is planning to undertake this build, I can send him a pair to test for his build, but if he does not have measurement equipment, he will only be able to subjectively evaluate. I probably even have a pair of RS180 in storage, but what I am lacking is spare time.

[Evil Twin]

The more important question to ask, is which cabinet configuration and size is built, as the "basic" Natalie P in the original PE enclosure was 22L and ported to that size, but larger cabinet configuration up to about 48L have been built (there are many Natalie P threads) with lower Fb tuning and a more extended bottom end, in a tower configuration. It really depends on the desired application and installation space.

[ChadC]

I don’t have measuring equipment, otherwise I’d be happy to help out.
The build I’m planning will have a cabinet in the 50L range. I’ll be ordering components in the next day or so, and have already started planning the build.
It will be interesting to see how the new ones compare to my existing pair. The cabinets will be similar.

[Evil Twin]

We would welcome seeing the results of your efforts...

[Evil Twin]

I have investigated this based on half space driver measurements and modeling in VituixCAD using baffle simulation to create full space behavior for the drivers mounted on the original Natalie P front baffle.

I will post both the result with the original crossover network and with a version with a few suggested modifications.

Using the unaltered original network:







Note that the LF response modeled for the RS180 is not ported, but the half space factory sealed IB data. The porting impacts response below about 150 Hz.


Suggested update for crossover:









Note that component designators are changed in VituixCAD, but component positions correspond with the original schematic drawing in LSPCAD.

Two LCR networks are added to the tweeter

• A parallel shunt Zobel network to cancel the Fs impedance rise and optimize the tweeter roll off; L4, C7, and R3
• A series ultrasonic network to null the peak at the diaphragm resonance of 26kHz. Note that the phase shield creates some other effects on axis; with this ultrasonic network, removing the phase shield will reduce HF irregularity on axis, but somewhat reduce off axis top octave response.

The woofer crossover network values are adjusted slightly, for the primary inductance L3, and the resonance notch network C6. Also, the series/parallel Capacitor C2 is adjusted and a 1 ohm resistor added in series.

[ChadC]

For C6, is that a 360 nanofarad or 360 microfarad cap?
For the inductors, is 18 to 20 AWG sufficient or do they need to be 14 to 16?

[Evil Twin]

For C6, 360nF. I.E., 0.36 uF.


The inductor values include DCR resistance in milliohms; select the wire AWG needed to give that DCR or a little less. Normally an inductor spec will include the DCR rating.

[Evil Twin]

Note, we are looking in to the possibility of building a POC measurement verification sample with the new drivers and crossover in the near future, but this could still be a few weeks away.

[ChadC]

Ok, waiting isn’t a problem. In the mean time, I’m having some problems sourcing some of the inductors. For L5, I can find one that’s 10uH but has a DCR of .07 Ohms.
For L1, I can find .40mH @ .15 Ohms, .47mH @ .20 Ohms, or .44mH @ .31 Ohms.
For L3, there’s a 1.1mH @ .57 Ohms, it’s kinda pricey, or a 1.2mH @ .29 Ohms, or a cheaper 1.2mH @ .33 Ohms.
I’m not sure how to choose between these.

[tktran]

I like the SB26ADC/CDC. But why o why do these use a 100mm faceplate? The SB26ADC would be a virtual drop in the the RS28a if they stuck to the original 104mm...

[Evil Twin]

Quite true... there are many corresponding cases like this.

[Evil Twin]

I have reviewed inductor options with regards to one popular US vendor for speaker building, Parts Express. Regrettably, they carry a relatively limited selection from some vendors such as Jantzen, and it seems that many parts from their "in house" Dayton brand have dropped off the list.

Often for inductors I prefer HiFi Collective in the UK- they cannot stock all the popular inductor types, but they get supply within a few days from Jantzen and others. And they make available essentially the entire Jantzen line, as one example, while Parts Express has a rather limited selection in comparison for many product families.


Here is an inductor summary with some alternate parts- DCR is a factor effecting crossover losses, filter damping, and the overall LF system damping- it does little good to have a low impedance amplifier connected to an inductor with close to an ohm of resistance in series with your 4 ohm woofer array. Recommended parts are checked. Using an LCR meter to unwind the inductor to the exact value is recommended if feasible.

[DeanP]

I noticed Solen.ca has a better selection of inductors, closer dcr specs too.

[ChadC]

Thank you. That helps a lot.
You mention using 4 ohm woofers. The ones I used in my first Natalie’s are 8 ohm shielded. Are those still the ones to use or should I use a 4 ohm unshielded version?

[Evil Twin]

The total load from two 8 ohm woofers wired in parallel is 4 ohms- this is what the crossover sees.

In there current series, this is the driver you should use...

https://www.parts-express.com/Dayton-Audio-RS180-8-7-Reference-Woofer-295-355?quantity=1

I do not believe the shielded one is available any longer- CRT systems are a thing of the past, for the most part.

Note also with regards to crossover inductor selection, the lower DCR part above is a Litz wire version, because this is all Parts Express carries in a lower DCR version at this time. HiFi Collective in the UK can source all of Jantzen's inductors with a short lead time- including low AWG parts with conventional solid core wire, which would be preferred for this application. Litz wire would be over kill unless working with a higher crossover point, and with 1.1 mH, that does not make sense...

[Evil Twin]

- a small update-

• The "vintage" RS180's in storage were located, but were found to be 4 ohm versions, so a new pair of 8 ohm parts were ordered for a POC build.
• Cabinet grade MDF has been set aside for the test build... along with some bamboo board for the front panel.
• A brief technical review of the RS180P-8 was conducted, and it should NOT be considered for this application, due to the differences in T/S parameters and a substantially higher Fs.

[ChadC]

This is what I currently have in my PE shopping cart.

https://www.parts-express.com/Dayton-Audio-RS180-8-7-Reference-Woofer-295-355?quantity=1

[JonMarsh]

That's just what you need!

[JonMarsh]

OK, I'm going to stir the pot a little today- if you want to set the Wayback Machine to 2005, I rounded up my inventory of original. vintage NIB RS28a, 9 pieces, and could sell some of those at my original cost plus shipping. Then you could use the original crossover. Obviously, only a useful idea if you haven't purchased the SB parts already.

But it's an option.