Cut-off frequency for JBL STX825 horn (p/n 5006815)?

[dubkarma]

I've been experimenting, using the JBL D2430K driver with the STX825 horn and am beginning to wonder if a 1000 Hz crossover point is too low—especially if I change the slope from 24 dB/oct. Linkwitz-Riley to 12 dB Butterworth. I know it's not too low for the driver, but perhaps it is for that small horn. All I've been able to determine from JBL's published information is that in the STX-825 speaker system a 1.3 kHz crossover point is used, though the slope is not specified.

[pos]

Do you have measurement tools?
The crossover JBL specifies in their loudspeakers datasheet is typically an acoustical crossover, and a Linkwitz-Riley one 99% of the time.
The electrical filters and EQ points used to obtain it are at different frequencies and slopes.

[ivica]

I've been experimenting, using the JBL D2430K driver with the STX825 horn and am beginning to wonder if a 1000 Hz crossover point is too low—especially if I change the slope from 24 dB/oct. Linkwitz-Riley to 12 dB Butterworth. I know it's not too low for the driver, but perhaps it is for that small horn. All I've been able to determine from JBL's published information is that in the STX-825 speaker system a 1.3 kHz crossover point is used, though the slope is not specified.

Hi dubkarma,

Looking at the JBL STX825 service manual

http://www.jblproservice.com/pdf/STX%20Series/STX825.pdf

it seems that it is used 3rd-order hi-pass network for the UHF driver (slope= -18dB/octave) and the crossover point is mentioned to be round 1.3kHz. Applying 2.7uF, 1mH, 12 uF elements, the other elements in the UHF section is used to 'flatten' driver-horn combo FR response.

regards
ivica
So using some

[sebackman]

With a little eq it is usable down to 1kHz with excellent results. XO 24LR

[Horn Fanatic]

I've been experimenting, using the JBL D2430K driver with the STX825 horn and am beginning to wonder if a 1000 Hz crossover point is too low—especially if I change the slope from 24 dB/oct. Linkwitz-Riley to 12 dB Butterworth. I know it's not too low for the driver, but perhaps it is for that small horn. All I've been able to determine from JBL's published information is that in the STX-825 speaker system a 1.3 kHz crossover point is used, though the slope is not specified.

Greetings -

The rule of thumb is that a high frequency horn, or horn tweeter, will be crossed over approximately one octave above the horn cut-off frequency. Consider the horn mouth perimeter dimension as the wavelength of the lowest frequency it can reproduce before reaching a frequency below that which will cause distortion due to standing waves.

When a wavelength is equal to or greater than the circumference of an orifice, a portion of that energy is reflected back to the source causing a standing wave. The same problem will hold true for low frequency horns. All commercial and domestic folded bass horns have distortion problems due to internal standing waves because the wavelength of the lowest frequency they will reproduce is always greater than the horn mouth perimeter. Corner horns can reduce that problem.

This is why you can watch the food cook in a microwave oven, without being cooked yourself. The wavelength of the microwave is greater than the circumference of the individual holes in the door screen. The energy is reflected back into the oven.

For example; the crossover point for the ALTEC 811 horn is 800 cycles, although the horn mouth cut-off frequency is approximately 400 cycles based on the perimeter dimensions. Those dimensions being the height of the horn mouth by the length of the arc, not the horn mouth width.

Perhaps you could provide the actual horn mouth dimensions without the flange. If the horn perimeter is equal to or less than 13 inches, and you're crossing over at 1K, you may have a problem.


HF

[Ruediger]

These two pages are an extract from the book "Acoustics" from Leo L. Beranek.

Ruediger

[catsnakebirdman]

Not sure if this helps, but here's an impedance measurement of a 2447 (2452SL diaphragm) on an STX825 horn.

In series with the driver were a 15 ohms sensing resistor and a 56 uF capacitor for blocking out lower frequencies. This capacitor and the fact that I forgot to turn of the highpass filter in the amp while doing the measurement (24dB/oct at 1100 Hz) lead to a terrible SNR at the lower frequencies. Also, the measurement rig is not calibrated, so the measurement is only qualitative.

I'm not adept at reading this image - just thought it might possibly be useful for you to identify the resonances mentioned above. If you want, I can repeat the measurement without all the highpass stuff to get a cleaner and more extended result.

The mouth perimeter of the STX825 horn is about 74 cm/29 in, which is the wavelength of a 465 Hz-wave.

[Ian Mackenzie]

I've been experimenting, using the JBL D2430K driver with the STX825 horn and am beginning to wonder if a 1000 Hz crossover point is too low—especially if I change the slope from 24 dB/oct. Linkwitz-Riley to 12 dB Butterworth. I know it's not too low for the driver, but perhaps it is for that small horn. All I've been able to determine from JBL's published information is that in the STX-825 speaker system a 1.3 kHz crossover point is used, though the slope is not specified.

It’s really a case of evaluation of your woofer and compression driver/horn with pink noise varying the crossover point to get the best blend. It’s nearly always best to err on the high side for horn woofer crossover points rather than attempt the lowest crossover point like many audio diy amateurs (American) who think a horn always sounds better crossed over as low as possible. Distortion, pattern control and power handling always go to hell when attempting the lowest crossover point.

I recommend you acquire a measurement kit and consider an active crossover.

I hope this helps

[Dave M]

In my experience, those newer JBL very short horns (or Waveguides) + snout-less drivers can be crossed at lower frequency than the older conventional horns with the same mouth size for much less side effect, probably due to extremely short overall horn length. Only +-2dB resonance and no rising distortion seen even at @800Hz when I measured 2450SL + STX/2332 (STX825 and 2332 are the same size, but a bit different shape). In fact, old Electro Voice 800Hz horn with larger mouth size had much higher distortion and resonance at the same frequency.

I guess the reason why JBL crosses STX @ 1.3K is just for max power handling. JBL DMS-1 Studio Monitor with 2450SL + 2332 is officially crossed at 1K with 14" woofer, not 15", and I confirmed that I could go even lower than that without any drawback.

[Dave M]

BTW, for home usage, I found that it would be preferred to apply -2dB BBC dip EQ (google it) for those JBL horns. A bit more enjoyable to listen to compared to flat frequency response setting. No harshness, no brutally honestness of typical JBL horn. :)

[Ian Mackenzie]

https://www.google.com/amp/s/www.avsforum.com/forum/155-diy-speakers-subs/2709033-2453h-sl-stx825-waveguide-over-2265g-1-center-channel.html%3Famp%3D1u

[Dave M]

Ian, thank you for the interesting link.
I'm also in Long Island, and it's good to know someone in my neighborhood has the same preference as mine. :bouncy:

PS: I rechecked the REW measurement data that I have. I said +-2dB resonance, but it is actually +-1dB 800Hz - 2K after applying 1/12 smoothing and some correction EQ. Not bad at all.

[sebackman]

Greetings,

The STX #5006815 is a completely different animal than the 2332. Not comparable at all. 25 years apart. I like 2332 and have had a bunch of them but nowhere near the STX. Same WG as used in VTX F12 & VTX F15.

If ego permits having a plastic WG I recommend everyone to upgrade, huge difference.

The STX works very good with the 1,5 drivers (ie 2447, 2450SL, 2451X, 476X and also 2452 & 2453) and less so with 3” drivers (243X), they get a dip in the curve.

SL are perfect dias for this WG. The best combination I have tested is a 2450SL or a 2451X motor with the 475Nd dia. In my opinion that is superior to D2s on this WG (and others).

This would give you a bastard 476Nd, which JBK never made as far as I know.

STX can be used down to 800Hz without any problems in home environment and using a DSP. I typically cross at 1kHz as I use 12” beneath.

I have used them with 15”s and the cross at 800Hz. You cannot do that with 2332.

This WG is very good IMHO. I have sold all other horns and currently only run this WG and a bunch of M2’s in various configurations. These are really very close to the M2 in sound. It’s cheap and crappy plastic but it works. It looks like the older PF things but perform complete different, Eyes can be deceiving.

I have a calibrated measuring rig and that confirms the ears 😊

Kind regards
//RoB

[Dave M]

Hi Rob, I have a big respect to you, but I have to say I have a different opinion about 2332 vs STX825 after 2 years of comparison. :)

They are the same size and are interchangeable on my JBL DMS-1 (2450SL-A) speakers, and share almost the same frequency curve, except STX825 has a slightly extended frequency with +-1db less diffraction above 5K when the same EQ applied. It is easy to compensate extended frequency just applying additional a few dB EQ over 10k on 2332. They both can be crossed around 800Hz in my situation.

As Rob says, the difference can be easily heard even after compensation EQ applied. Newer JBL driver like 2450SL does not have a long snout as classic JBL divers, but 2332 still sounds like a classic horn, while STX825 waveguide has almost no horn signature. Yes, 2332 has the sound of "horn" speaker that we are familiar with. It does not sound like a STX waveguide, for sure.

I find Optimized Aperture (2332/2352) horn has their own very charm. 2332 is a very interesting postmodern design, an extremely short conical horn with an exponential horn like curved mouth and a tiny bi-radial horn like vertical slot at the throat. It would probably be one of the most neutral sounding "Horn" that JBL has ever developed. They are the "last JBL Horn" designed with the accumulation of their wisdom right before the new designer brought "Waveguide" to JBL with his completely new mathematical method.

JBL waveguides are NOT improved versions of their horns. Those JBL waveguides have very little relationship to their previously designed horns, culturally, philosophically, nor technically. I believe that those new JBL waveguides were designed by a person who dislikes the sound of horn, for people who dislike the sound of horn, and they do not sound like a horn.

Well, I think I have to confess I usually prefer horn to the dome tweeter, and this is my personal opinion, but I finally concluded that STX850 sounds somewhat flat and a bit forward for my taste. Overall, 2332 horn sounds more neutral, 3D, musical and relaxing to my ears in my studio. It seems like OP has a similar impression about them interestingly.

If someone wants to try them, I would recommend to test carefully before you decide which to use, conventional horn or waveguide. They are different animals. Enjoy!


83207

[DES-1]

...This WG is very good IMHO. I have sold all other horns and currently only run this WG and a bunch of M2’s in various configurations. These are really very close to the M2 in sound. It’s cheap and crappy plastic but it works. It looks like the older PF things but perform complete different, Eyes can be deceiving.

I have a calibrated measuring rig and that confirms the ears 😊

Kind regards
//RoB


Good to hear your continued work with the STX confirms your earlier research! I decided to go this way rather than an M2 WG for my build, partly due to aesthetics & simplicity but mostly based on your observed performance. I had a few delays but am nearing the time to fire them up.

[Ian Mackenzie]

Hi Dave M,

Each to his own.

The DMS 1 was hyped at the time as an improved version of the 4435.
However it was a flop commercially.

The consumer M9500 K2 with the more traditional bi radial horn profile is still highly regard by audio critics and was successful.

Both the DMS 1 and the M9500 share the same woofer as l recall.

Btw the manufacturing quality of the 2332 was absolutely terrible. I have a pair.

They are as rough as a badgers arshole.

We are all entitled to our own opinions about what we prefer and that is what makes “audio reproduction” a diverse industry in terms of product offerings.

[Dave M]

HI Ian,

I had 4425 and still have UREI 813C. As a "real" large studio monitor, I have no hesitation to say DMS is definitely better than that generation of JBL monitors, but it was released when Genelec and TAD Kinoshita was dominating the large monitor market, and JBL was considered to be no longer in fashion. I don't think Kinoshita style configuration with Genelec style time alignment was a bad concept, but it's also hard to deny DMS/9500 are nothing but a Kinoshita vertical twin knockoff.

BTW, I do not really see any manufacturing quality issue of 2332, and in fact, it looks much more expensive than STX to my eyes, but my 2332 looks like it is painted by JBL to match the color / texture with the DMS enclosure.

[Ian Mackenzie]

Maybe your 2332 was a different batch. Others here have commented on the finish.

I don’t wish to comment on the comparisons other than the M9500 was unique in terms of the drivers, the horn and the system integration by Greg Timbers.

http://www.allegrosound.com/JBL_K2_M9500.pdf

[Dave M]

Hi Ian,

hahaha, I guess JBL painted 2332 to make it look appropriate for expensive studio monitor. :)

"The Horns are made of aluminum WITH the exception of some versions of the 2332. The version in the later array cabs is shiny Fiberglas where the version in the DMS-1 monitor is spray-coated on the face like the rest of the baffle. "
http://www.audioheritage.org/vbulletin/showthread.php?35019-Array-horn-history-revisited&p=354757&viewfull=1#post354757

I have read a very interesting Greg’s comment about 9500 before (probably here in this forum). What I said about Kinoshita/DMS/9500 is the vertical twin woofer configuration. Yes, enclosure design, horn and driver on them are very different. Interesting thing for me is, 475Nd driver on M9500 has a classic long snout unlike 2450SL. 2450SL was the world first snout-less driver (SL probably means Snout-Less) developed with 2332/2352 Optimized Aperture horns. JBL could have used M9500 horn for DMS-1, but the reason why they didn’t should be they wanted to promote their new generation of lower distortion snout-less horn design to pro market, like they do the same with M2 now.