Dayton RS52AN-08

[mgrabow]

can anyone recomend crossover points for the Dayton RS52AN-08? Also does anyone have the response file to import?

[Pete Schumacher]

Re: Dayton RS52AN-08

can anyone recomend crossover points for the Dayton RS52AN-08? Also does anyone have the response file to import?

With 2nd or 3rd order electrical filter on the low end, you can cross as low as 500Hz, if you don't intend to push it real hard. Shallower slopes require a higher XO point, 800Hz or so.

On the high end, the large flange sets the upper limit to around 4000Hz, preferably a bit lower to get good integration with the tweeter.

As far as response files for import to a crossover design program, I've got Zaph's plots converted to FRD for use in Jeff B's PCD program.

What are your intended baffle dimensions?

[benchtester]

Re: Dayton RS52AN-08

Let me start the feedback (oops Pete filed first). I am sure others will have more to say.

I have mocked-up Zaph's ZDT3.5 with the earlier aura tweeter; it sounds fine. I suspect the current Dayton tweeter is good too.

http://www.zaphaudio.com/ZDT3.5.html

(He uses LR2 - 850 Hz. and LR4 - 3500 Hz.)

It is important to filter out the break-up notch, it is a horrible screech if you don't.

Some designers may be tempted to go higher with the RS52, but the off-axis starts falling off above 4 khz. I consider that to be the upper frequency limiting factor.

[mgrabow]

Re: Dayton RS52AN-08

Let me start the feedback (oops Pete filed first). I am sure others will have more to say.

I have mocked-up Zaph's ZDT3.5 with the earlier aura tweeter; it sounds fine. I suspect the current Dayton tweeter is good too.

http://www.zaphaudio.com/ZDT3.5.html

(He uses LR2 - 850 Hz. and LR4 - 3500 Hz.)

It is important to filter out the break-up notch, it is a horrible screech if you don't.

Some designers may be tempted to go higher with the RS52, but the off-axis starts falling off above 4 khz. I consider that to be the upper frequency limiting factor.

Thanks, where do you suggest placing the notch?

[mgrabow]

Re: Dayton RS52AN-08

With 2nd or 3rd order electrical filter on the low end, you can cross as low as 500Hz, if you don't intend to push it real hard. Shallower slopes require a higher XO point, 800Hz or so.

On the high end, the large flange sets the upper limit to around 4000Hz, preferably a bit lower to get good integration with the tweeter.

As far as response files for import to a crossover design program, I've got Zaph's plots converted to FRD for use in Jeff B's PCD program.

What are your intended baffle dimensions?

It is not for my project so I am not sure on the baffle size...

[Pete Schumacher]

Re: Dayton RS52AN-08

It is not for my project so I am not sure on the baffle size...

The reason I ask about the baffle is that it will affect response, and consequently the XO part values.

[benchtester]

Re: Dayton RS52AN-08

Thanks, where do you suggest placing the notch?

Zaph describes it better than I can, in the above project page. The breakup is at 13 kHz. He uses a 12 uF cap across the driver terminals; but it does double duty, so it may not be correct with a different crossover. A good rule of thumb (again per Zaph) is out-of-band notches should be parallel to the driver and in-band notches should be in series with the driver. (This keeps the impedance more uniform, and to me seems to make intuitive sense.)

In my messing around, I just copied the whole tweeter and midrange circuit. I adjusted the padding resistors (R4 and R6) to fit. Like I said, it sounded pretty good.

[Pete Schumacher]

Re: Dayton RS52AN-08

Zaph describes it better than I can, in the above project page. The breakup is at 13 kHz. He uses a 12 uF cap across the driver terminals; but it does double duty, so it may not be correct with a different crossover. A good rule of thumb (again per Zaph) is out-of-band notches should be parallel to the driver and in-band notches should be in series with the driver. (This keeps the impedance more uniform, and to me seems to make intuitive sense.)

In my messing around, I just copied the whole tweeter and midrange circuit. I adjusted the padding resistors (R4 and R6) to fit. Like I said, it sounded pretty good.

Well . . .

Zaph's "notch" circuit is in series with the driver. It is formed with the small cap in parallel with the series filter inductor, L8 and C10.



"Be aware that 3 caps are paralleled for C10, to specifically reach .30 uF. Common single cap values such as .22 and .33 will not be a precision strike on the mid dome's breakup node. For resistors, nothing special except that R14 is modified to compensate for the inductor DCR in that circuit, while R20 is doubled up for power handling. The large Erse inductor was selected for it's low DCR to keep the overall system sensitivity as high as possible. They are well built, and I like the way they can be screwed into a board. When mounting screws are used, the screw is far enough out of the flux field to not affect the inductor value. "

[benchtester]

Re: Dayton RS52AN-08

I stand corrected. Before I didn't understand the function of the .3 uF cap since the impedance is not significant by itself. It is now clear to me that it works with the inductor as a tank circuit. Could someone explain to me how a tank circuit has infinite impedance at the tuned frequency? Right now I have to take it on faith.

I also misstated his "general guideline" (found under Designing this Crossover in the ZD5 project page). It should be: series notches should be used outside the driver's bandwidth and parallel notches should be used inside the driver's bandwidth. In particular, because it keeps the impedance from getting too low. Note that here that "series" and "parallel" apply to the notch configuration. It can get confusing because series notches are often in parallel with the driver and parallel notches are often in series with the driver. Another point I could use clarification on: is the "driver bandwidth" the range of the driver or the passband range of the driver in the application?

However, I was correct in stating that Zaph describes it better than I can.

[DannyT]

Re: Dayton RS52AN-08

Pete used similar filter on Presstina but with a .15uf parallel to the 1mh series inductor. I would like to know how to calculate the uf value to strike the break up notch of RS52AN?

[mgrabow]

Re: Dayton RS52AN-08

Thanks for the info on the "out of band" notch filter. I would be curious to see that on a spectrum analyzer.. being that far out of band and all... More info than I asked and that is helpful Thanks mutch...

[Pete Schumacher]

Re: Dayton RS52AN-08

Pete used similar filter on Presstina but with a .15uf parallel to the 1mh series inductor. I would like to know how to calculate the uf value to strike the break up notch of RS52AN?

Since the inductor in the Prisstina is double that of the ZDT3.5, the cap would be halved to keep the same resonance.

The equation to calculate the tank resonance is:

f = 1/(2*pi*sqrt(L*C))

[benchtester]

Re: Dayton RS52AN-08

Thanks for the info on the "out of band" notch filter. I would be curious to see that on a spectrum analyzer.. being that far out of band and all... More info than I asked and that is helpful Thanks mutch...

I am unclear if the breakup point is considered out of band since nobody uses it there or if it is in band because the speaker can generate a lot of sound there.

But to the spectrum analyzer point, you can see the theoretical design with Zaph's modeled transfer function (curve that dives down at 13 kHz).

http://www.zaphaudio.com/ZDT3.5-transferfunction.gif

And you can see the resulting measured driver response (green curve):

http://www.zaphaudio.com/ZDT3.5-measured-farfield-tweeter-mid.gif

That's why I like ZaphAudio, there is a lot of info already there.

[Wolf]

Re: Dayton RS52AN-08

It is now clear to me that it works with the inductor as a tank circuit. Could someone explain to me how a tank circuit has infinite impedance at the tuned frequency? Right now I have to take it on faith.

The reason it is infinite impedance at this particular frequency, is that The capacitor passes freqs above the tank, and the coil passes freqs below the tank. Due to the components respective properties, the impedance will be very high between these 2 part's passband. The term 'infinite' could be a little arbitrary due to fact of it being "very high in terms of speaker impedance". It might not be "infinite', but "very high".

Typically, out-of-band notches are done in a shunt manner. Sometimes this is not as easy, due to the impedance issues at hand, and sometimes opting for a series notch is a better modus operandi.

FWIW, I had to use a tank in the top-end of the Mn-14's tweeter filter to suppress it's breakup, but this was not an out-of-bandwidth resonance.

Later,
Wolf

[benchtester]

Re: Dayton RS52AN-08

The reason it is infinite impedance at this particular frequency, is that The capacitor passes freqs above the tank, and the coil passes freqs below the tank. Due to the components respective properties, the impedance will be very high between these 2 part's passband. The term 'infinite' could be a little arbitrary due to fact of it being "very high in terms of speaker impedance". It might not be "infinite', but "very high".

Typically, out-of-band notches are done in a shunt manner. Sometimes this is not as easy, due to the impedance issues at hand, and sometimes opting for a series notch is a better modus operandi.

FWIW, I had to use a tank in the top-end of the Mn-14's tweeter filter to suppress it's breakup, but this was not an out-of-bandwidth resonance.

Later,
Wolf

I did get the "infinite impedance" quote from Wikipedia so it must be true. ;) (under the parallel circuit, impedance)

http://en.wikipedia.org/wiki/LC_circuit

I guess it is spelled out there. But I am very visual and I have trouble with imaginary numbers (the j in the equation).

If I just look at the real components of the impedance, the cap and inductor have a real impedance of 40 ohms each at 13,000 Hz. If you treat them like resistors, the parallel combination is 20 ohms. You would do better with just the inductor at 40 ohms; in which case, I think the transfer function would look like a steadily decreasing line. By adding the cap, you get a sharp dive at 13,000 hz and a recovery up to -70 dB, though I doubt this rise is of benefit.

I understand that the inductor passes current below the notch and the cap passes current above the notch. What I don't get is that they interact in a way that less current is passed at the notch frequency than if either of them were used alone.

Is it like an English roundabout, currents at tuning frequency just sloshes around the cap and inductor and never gets out the other side? (If this is too far OT, I can just deal with it.)