Horn/Driver Comparisons

[Mr. Widget]

As most all of you know Zilch has been working on a variety of 2-way designs using a variety of JBL compression drivers and CD horns for his "Quick and Dirty 4430" thread. Today we spent a few hours comparing five CD horns and six different JBL compression drivers. We set them up and took quasi-anechoic measurements with Clio.

We used the following horns:

1) PT-F1010 - 100° X 100°, 6.5" X 12"
2) PT-H1010, 100° X 100°, 12" square
3) PT-F64, 60° X 40°, 6.5" X 12"
4) SF-95, (Sound Factor), 90° X 50°, 1" thread-on throat, 6.5" X 12"
5) P-Audio PH-230, 100° X 100°, 2342 Clone, 1" bolt-on throat, 9" Square

And we used these drivers:

1) 435Be 3" Beryllium Diaphragm 1.5" exit from JBL Consumer
2) 2435HPL 3" Beryllium Diaphragm 1.5" exit
3) 2431 3" Aluminum Diaphragm 1.5" exit
4) 2418 1.75" Titanium Diaphragm 1" exit
5) 2426 1.75" Titanium Diaphragm 1" exit
6) LE85 1.75" Aluminum Diaphragm 1" exit

[Mr. Widget]

Before we began our horn comparisons we wanted to explore the low end capabilities of the 2431 and 2435HPL compression drivers with their small back cap and compare them to the 435Be with it's large back cap. To do this we mounted the three drivers on a 500Hz 1.5" throat horn I have that is about the size of an Altec 511. It is not a CD horn so it's on-axis response is reasonably flat without compensation.

Here we see that the low end performance of the 435Be, 2435HPL, and 2431 are all quite good. It would seem that the larger back cap isn't necessary for performance below 1000Hz. We did not do any listening tests or power tests measuring distortion, but the initial measurements were surprisingly encouraging. I am not sure if it is due to the back cap or other tweaks, but the 435Be did exhibit, albeit by a small margin, the flattest overall response of the three.

[Mr. Widget]

Next we tried the 2431 on both the PT-F1010 and PT-H1010 horns. We have included the plot of the 2431 on the large horn to show how the LF limiting is due to these smaller horns. Also since these are CD horns they would need passive or active EQ to compensate for the HF fall off.

[Mr. Widget]

Next we tried the 2435HPL on the PT-F1010 horn. We left the 2431 plot up for comparison. We also plotted a second 2435HPL. This one (Shown in Green) seems to be defective... perhaps why it was available inexpensively on eBay?

[Mr. Widget]

Next we tried 4 different 2431 drivers on the PT-H1010. Zilch has one pair (Green and Purple) that were purchased about a year ago, and another pair (Red and Blue) that were purchased recently. Each pair seems closely matched... but the two pairs do seem a bit different from each other. Perhaps there was a slight design change between the two runs?

[Mr. Widget]

Next we tried the PT-F64. Here we have kept the plot of the 2431 on the PT-H1010 for comparison with the plots of the 2431 and 2435HPL both on the PT-F64. This horn is much deeper that the other in this test and while stile considered a CD horn, it obviously need far less compensation... at least below 9KHz.

[Mr. Widget]

Next we tried the remarkably inexpensive SF-95 horn. In the photo it looks a bit like the PT-F64, but it is a 1" horn instead of a 1.5" horn, not quite as deep, and made of an inexpensive plastic. Here we compare the 2431 on the PT-F1010 with the 2418 on the PT-F64. We measured two examples of this very economical driver and horn combination... they were surprisingly similar in performance. Surprising that such inexpensive drivers would be so precisely manufactured.

[Mr. Widget]

Next we mounted a 2426 driver onto the SF-95 horn to see if the 9500Hz suck-out is due to the horn or the driver. We see that there is no suck-out with the 2426.

[Mr. Widget]

Here we compared the 2431 on the PT-F1010, with a 2426 and adapter and LE85 and adapter all on the PT-F1010.

[Mr. Widget]

Finally we mounted the 2418 on a P-Audio PH-230. Here it is compared to the 2418 mounted to a SF-95.

[Mr. Widget]

Kind of hard to make any definitive conclusions. It would seem that the PT-F1010 was the best CD horn of this bunch in that it had the smoothest downward slope. It is too bad that it doesn't allow for a crossover below about 1.2KHz. I suppose we should have compared it to the H9800 just for grins. I hope this data helps others on their audio journeys.

Widget

[Guido]

Thank you so much Zilch and Widget

I've a question regarding post #9:

Does it mean that the 2431 is
5 dB more sensitive than the 2426 and
3 dB more sensitive thean the LE85 ?

[Titanium Dome]

Looks like you guys had a fun day.

Thanks for sharing.

I've often wondered how the 1" Bi-Radial horn on the larger SVAs would fare in such a comparison.

[Steve Gonzales]

This is a very enlightening thread, thank you both for your time and for sharing the info.

[Zilch]

Here's my commentary, while the work is still fresh. We haven't sorted it all out yet, but there were some surprises:

Post #2: The size of the back cap is not the big deal with respect to low-end response we thought it was, at least insofar as frequency response is concerned. Whether the smaller-can 2431 and 2435 sound any good below 1 kHz is another matter. On an appropriate horn, they'll at least play down there.

Post #3: Driven by 2431H, the performance of PT-F rectangular and PT-H square waveguides is virtually identical. PT-H plays a little lower, perhaps; it'd take a statistical study to distinguish them, most likely.

Post #4: We don't find a HF extension performance advantage with Be diaphragm (and $1000 premium cost) 2435HPL. As suspected, one of them is defective. They're both going back to JBL for warranty repair.

Post #5: Confirms my feeling that JBL is making design and/or manufacturing improvements on 2431H. Performance above 12 kHz appears "chaotic," but they all play similarly there on RTA. Newer ones are better in HF extension. About 15 dB of compensation is required to make them flat to 20 kHz on PT-F waveguides.

Post #6: For those desiring a narrow sound field, PT-F64 may be the ticket. Remarkably flat up to ~9 kHz, they play lower, too.

Post #7: The "economy" Ti-diaphragm 90° X 50° combo, it sounds good with proper compensation, actually. PT-H95 is available for 1.5" exit drivers, but the PT-F version of this "preferred" pattern is only prototypes thus far, alas.

Post #8: Hmmmm. Gonna have to see if it sounds better, too. Another economy alternative, though I'd want to support the much heavier 2426 driver independently. That is one cheap horn, for certain. There's also a SRX 90° X 50° horn for 1.5" drivers that might be worth looking at:

http://www.jblproservice.com/pdf/SRX...es/SRX712M.pdf.

Post #9: We think the adapter is responsible for the bumps in the FR. Actually, two adapters are required to mate the LE85 with PT waveguides. Still, as you may know, it's my favorite-sounding combo of "vintage" drivers. I play it with 3110A crossover for compensation.

Post #10: PH-230 is P-Audio's inexpensive 2342 clone. It'd take polar studies to determine if it's worth a whit for constant directivity. The real thing is still available, tho....