Exploring high performance ferrite driver options

[Wylandright206]

Hey y'all, I'm back. So as I mentioned in a previous post I've been simulating various different drivers in the t48 and I've found one in particular that stuck out to me. The PRV 15sw2000 is fairly new, within the last 5 years, it's 190usd and it's available all over the internet in the US. First party recone kits are readily available and shipping is surprisingly timely. I ordered one and it showed up about 3 days early so that was exciting. We dropped it in a freshly built t48 (no need for chamber reducers) and sent it to hell for about 20 minutes just to see if it could really stand up to the pressures in the throat before we go doing anything in the field. I am happy to report that after literally blowing the driver up, voice coil on fire and everything, this driver is a monster. The failure mode was entirely coil related, it bottomed out jumped the gap and lit on fire. We were DEFINITELY exceeding power rating significantly and the soft parts of the woofer were in great shape! Spider was totally fine, cone was intact and the surround looked great.

I used a high pass filter at 36hz 4th order butterworth because hornresp says it will take around 900wrms out of its 1000w rating with said filter. I might move that filter up 1hz but it's probably unnecessary. However we were producing some pretty insane spl before noting the voice coil bottoming out. Much beyond where we would set a limiter. It seriously sounded like there was more than one of them I was EXTREMELY impressed. Not only was it significantly louder than the 3015lf it performed better down below 50hz and the pronounced peak at 60hz after crossover filters with the kappalites did not exist with the 15sw2000. They sounded great without any sort of EQ where the 3015lf would have needed a fair bit of sculpting. I think this woofer is a great bargain! For the price I'm absolutely blown away. It digs just as deep, offers much more power handling and excursion than the 290$ 3015lf, and even if this cost significantly more I would be equally as impressed. I would argue that I actually think this woofer actually sounds better than the kappalite. I'm contemplating swapping some mine out.

We really didn't take any measurements or use a voltage meter like we should have, so this is far from any sort of scientific test. However I would be happy to recommend this to someone who wants to save a couple bucks on drivers and doesn't mind the 24ish lbs weight. Or even someone looking for an affordable upgrade path. It's significantly cheaper than any of the other high performance options Ive seen and offers a little bit more wiggle room if you abuse your t48s with dubstep like I do. I really hadn't seen many people post about PRV online it's a lot of cringy car audio stuff. People don't like the stuff because of the logos on the dust cap. But who's going to see that in the throat of a 10ft bass horn? I don't mind taking one for the team and reconing. The cost of the driver and a recone kit is still cheaper than a 3015lf. Anyways thanks for listening y'all.

PS. As a final note I will say that hornresp seems to be confused about this driver and thinks it's heavy enough to warrant "lossy LE" and applies this by default. I do not believe this to be an accurate model. The Lossy LE unchecked model correlates much more with what I was hearing when sweeping the sine wave generator back and forth. If lossy LE is checked it appears this driver sims pretty much exactly the same as the 3015lf. But that is not the case. With lossy LE unchecked the ripple at 60hz is gone.

[Seth]

Wyland, your posts are always so meaty. Thick, juicy, and full of valuable and/or interesting information. Thanks for that and thanks for taking one for the team in this instance

Manufacturers web page for the 15SW2000
PRV 15SW2000 on Parts Express

We really didn't take any measurements or use a voltage meter like we should have, so this is far from any sort of scientific test.

I love number crunching, data analysis, and would have loved to see relatively accurate response charts. But, more times than not, it's just not practical to do what it takes to produce those results, especially if you're not also measuring a known driver for comparison. In this instance, your firsthand experience and opinion are definitely sufficient in assigning value to the PRV 12SW2000.

It's great to know the driver is widely and readily available and that recone kits are available too. One thing you might consider before switching your entire herd over is; I bought a case of 8 PRV neo drivers for some planned ultralight DR200 builds. PRV has since discontinued the specific model and there's no longer a comparable driver available from them. Just something to be aware of, being a newer driver on the market, there's very little market history to suggest it'll be available in another 5 years. Although, with the recent merger/sale of Eminence, who's to say any of the Eminence lineup is likely to be available for replacements from them either? Anyway, just something to take into account.

So, Hornresp indicates 85v with 24dB/octave filter set to 36Hz keeps the linear excursion below 12.5mm throughout the whole pass band? Or, mostly so?

If I were building T48's and wasn't worried about additional driver weight, I think I'd be compelled to save the money and load the cabs with the PRV drivers. But, then I might also be compelled to purchase amps capable of producing the voltage limit of the drivers, which would probably negate overall savings, but result in increased maximum output potential/headroom.

Just based on the numbers themselves and ignoring other relevant variables, the additional Xmax equates to an increase in maximum output potential of 2.3dB and the difference in 60/85 voltage limit is 3dB.

[Wylandright206]

So according to hornresp, at 36hz 4th order butterworth the prv will take around 78v with a DC resistance of 5.5ohms which turns out to be around 850 watts, a little under that. This takes advantage of the full excursion of the driver. In my mind, even if you didn't take advantage of the full power handling, it exhibits much flatter response than the kappalites and just offers more energy at 40hz at a given voltage. When messing with the filter wizard and adding the usual filters the prv does not need EQ to be usable it's all well within 5db. The kappalite is about 9db down at 40hz from peak at 60hz. Also according to the simulations the group delay caused by mouth reflections is SIGNIFICANTLY reduced by using this driver. I'm not sure exactly why that is, but I do believe this contributed to my opinion that this driver sounds better than the kappalites. At least cab for cab. Ive attached some hornresp charts using 4th order Butterworth at 36hz and 100hz on the prv and 38hz/100hz on the kappalite. The difference in low frequency output is not insignificant. It's not just louder at 40hz it's also more sensitive there. I think it's worth noting that both of these drivers have a DC resistance of 5.5 ohms.

[Seth]

So according to hornresp, at 36hz 4th order butterworth the prv will take around 78v with a DC resistance of 5.5ohms which turns out to be around 850 watts, a little under that.

78v at 5.5Ω DCR is 1106 Wats. But, 78v and 850 Watts equates to roughly 7.2Ω impedance, which is what I assume your numbers are coming from. However, power isn't really important as long as it doesn't exceed the drivers thermal rating. Other than choosing an appropriate amplifier, Watts/power has little to no importance within this particular conversation... unless I'm overlooking something(?), which is entirely possible.

...This (voltage) takes advantage of the full excursion of the driver. In my mind, even if you didn't take advantage of the full power handling, it exhibits much flatter response than the kappalites and just offers more energy at 40hz at a given voltage.

It took me years to let go of thinking the power rating of a driver meant it was louder, somehow better than a driver with a lower power rating, or that if I didn't power it with at least that much "power" I was leaving performance on the table. Fact is, I'd personally be happy if a driver reached it's peak potential at a voltage that equates to any amount of power based on nominal impedance that's less than the driver's RMS rating... even half. Power and power ratings have nearly zero importance once it's verified the thermal rating of the driver isn't the limiting factor for setting the limiter. There are instances, but it's outside the scope of how we do things here.

When messing with the filter wizard and adding the usual filters the prv does not need EQ to be usable it's all well within 5db. The kappalite is about 9db down at 40hz from peak at 60hz. Also according to the simulations the group delay caused by mouth reflections is SIGNIFICANTLY reduced by using this driver. I'm not sure exactly why that is, but I do believe this contributed to my opinion that this driver sounds better than the kappalites. At least cab for cab. Ive attached some hornresp charts using 4th order Butterworth at 36hz and 100hz on the prv and 38hz/100hz on the kappalite. The difference in low frequency output is not insignificant. It's not just louder at 40hz it's also more sensitive there. I think it's worth noting that both of these drivers have a DC resistance of 5.5 ohms.

Looking at the charts, if the actual performance holds mostly true to the model, I'd say the increase in sensitivity in the 35-40Hz region is a huge benefit to the overall peak output level of the whole system. Like, considerably so. So much so, I'd personally choose this driver as a default recommendation for the T48 and only recommend the 3015LF for people who need the weight savings. Although, the weight savings of a 3015LF would come at a huge hit in output potential. Would be tough to steer anyone that direction... if the differences indicated in the model hold true and accurate.

Do you have the time, space, place, and gear necessary to produce 2.83v/1 meter response plot's for one of your cabs loaded with a 3015LF and another one loaded with the PRV 15SW2000? I would absolutely love to see the two plots tested in the same setting with the same test equipment. Willing to chip in financially if you need it.

This is really exciting to me.

I'm not personally well informed in the subject of Group Delay, however here's what Bill had to say about it in another thread:

...For group delay to be a factor the design would have to be so flawed that the group delay would be the least of it's problems. Audiophiles tend to be concerned about group delay because the value looks like it should be serious. If that was the case then a 47ms maximum and average of 20ms throughout the pass band should give a very bad result. Owners of a speaker with that spec, the Tuba HT, would disagree.

[Seth]

Six hours later and my mind is still churning on this deal. If the model is reliable in comparing the two drivers against each other, this PRV driver could improve sensitivity in the lower region and provide 36" cab sensitivity with a 24" cab



As far as I know, the forum chart-tool charts reflect actual measured results. In the chart above, I'm seeing roughly 5dB difference at 40Hz between the 36" and 24" cab measurements. And I'm also seeing about the same 5dB at 40HZ with the PRV driver in your modeling results in the chart below. Except, the 3015LF version of your modeled response is about 5dB shy in the 40Hz region compared to the measured response chart... the 24" cab models lower than it measures.



So, is the difference due to the values used in the reverse engineered Hornresp model? Is it the measuring technique used in creating the measured response chart? Is it just inherent differences to be expected between modeling and real life? Some other reason the modeled curve isn't closer to the same as the measured curve? Who knows?

The 36" T48 is a monster in more than one way. Most notable is the low frequency sensitivity. It really is among the best out of all the portable subwoofers ever. My mind really wants to believe a simple driver choice/change can get 36" response out of a 24" cab. My imagination then asks, if that's a true, accurate, and achievable result... what sort of unbelievable low end sensitivity would that driver in a 36" wide T48 produce?

We really do need measured response charts to tell the real story here. Both options measured same day, same set-up, same identical settings. Measure cab A, don't do anything but mute the source, wheel the other cab in place, Measure cab B. Hoping the results are glorious!

[Bill Fitzmaurice]

HornResp models aren't 100% accurate, as there are factors that it doesn't take into consideration. Predicted peaks and valleys are never as high or deep as the model predicts. Also, the modeled results shown here are off a bit. My model shows the PRV 2dB more sensitive at 38Hz, above that they're essentially the same.Below 38Hz it's 2dB lower in sensitivity. You won't get 36 inch wide results from a 24 inch wide box.

[Wylandright206]

From my past measurements of the t48 I can see the peaks are not as pronounced as the hornresp models but the model is fairly close. Having the hornresp model to compare driver performance is very handy. While they might not be 100% accurate to real world it gives me a good starting point that I can then measure whenever I recone this driver this week. Don't you worry Seth I will get you those sweet sweet controlled tests. According to hornresp the 24" prv single cab (lossy LE off) is less sensitive than the 36" 3015lf single cab by around 2 or so db, but in general has a very comparable slope. And when loading them with two drivers the prv becomes the clear winner. And the Max output provided by a dual prv loaded t48s is impressive. Louder than most double 18s regardless of power handling. However when using 1 driver the kappalite does look a bit flatter in the 36 inch titan.

Either way Ive used these subs every weekend for years and I know what I heard. These drivers are fabulous in this horn. The kappalites in 24 inchers want a lot of friends to give me the sound I like. Seems to require significantly less to get where I want with these heavy ferrite drivers. WILL UPDATE ASAP. I would also like to mention the recone kits they sell have the former already epoxied to the cone very easy to repair and they will actually sell you the glue they use, unlike eminence. I am very hopeful, this seems like a viable upgrade path for a cheapskate like me

[Rich4349]

There's something odd to the two graphs around 40-45hz.

With ONE 3015, the 36" T-48 makes about 99dB.

With TWO 3015s, it makes... about 99dB.

Should we check on the public works dept in Denmark? #somethingisrotten

[Seth]

There's something odd to the two graphs around 40-45hz..

Link, add to post, or describe which graph you're talking about please. It's probably staring me in the face, but for some reason it's not obvious to me which graph you're talking about.

However, I will say the cab (width) does the work. A dual loaded cab acts like two cabs of half the width and the natural occurring benefits of having a wider cab are greatly diminished.

[Rich4349]

The post immediately preceding mine.

[Seth]

Oh, I see what you're talking about... the single loaded trace from the top chart and the dual loaded trace from the bottom chart. You're right, that little dip right there is very similar on both traces.

Nice to know I'm not the only one looking so deeply into what Wyland has posted. Intriguing, isn't it?

[Seth]

There's something odd to the two graphs around 40-45hz.

There aren't too many dual loaded SPL charts in the forum chart tool, but this one with single and dual loaded T30's shows the same thing you're talking about. It seems to be due to a 2x cab having the response of two cabs of half the width, which don't have as much cab working for them, per driver. I know it's confusing and stretches' the mind to make sense, I had the exact same inquiry a couple years ago.

[Wylandright206]

Here are some comparison sims of single versus dual loaded 36 inchers with the two drivers. The single prv loaded one looks very similar to the lavoce waf154.00 and b&c 15tbw100 (significantly more expensive drivers I might add) in sims now that it has such a large rear chamber. Once you add a second driver you gain extension and output. But not at 40hz, same with the kappalites. Im not sure what's going on here since I just noticed this phenomenon as well, but my thought was that the two drivers sharing a chamber acting on each other was having a significant effect on their total compliance, which is what would lead the response to change pretty significantly. This effect can be seen when simulating all enclosure types with multiple drivers. Some drivers are not meant to share a chamber. You see lots of commercial bass bins that are just two single 18s stuck together. I would be happy to be proven wrong I just thought I would share my theory as to why the response is so different. Now the fact that 40hz doesn't gain benefit from either driver could be a coincidence or it could be something beyond my current scope of understanding. It sounds like Seth has had these thoughts before.

[Seth]

It sounds like Seth has had these thoughts before.

I have. And here's the conclusion I came to, which is the same as Bill has said forever... the only time dual loading a cab is desirable and appropriate is when you need to manipulate the final impedance of the cab. Otherwise, single loaded cabs, built as wide as you can transport are, hands down and by far, the optimum way to go.

I'm sure I'm not the only one who's mind keeps trying to figure out the best configuration and how to get the most output from a cab. Adding a second driver was my first impulse. And it makes sense. More is better, right? Arguably, yes. But, doubling the driver count is not the same as doubling the cab count.

If a guy has the ability to transport a 36" dual loaded T48, 99 times out of 100 he would also have the ability to transport 2 single loaded 24" T48's. Two single loaded cabs of half the width of a double loaded cab will match the performance potential. And two single loaded cabs of any width that is wider than half that of a dual loaded cab will outperform the dual loaded cabs performance potential.

So, unless we need to manipulate the impedance of a cab, which is nearly never the case, there is absolutely no reason whatsoever to even consider or speculate dual loaded cabs. That said, it seems to be one of those things that's like trying to not look at a naked lady or an automobile accident on the highway.

I might add, I'm no guru in this shit. I'm on the path with most of you guys. Always intrigued with it all. I love the data and I love the collaborative conversation.

[Wylandright206]

That said, it seems to be one of those things that's like trying to not look at a naked lady or an automobile accident on the highway.
[/quote]
You got that right that is exactly what's going on here. In mind it seems that replacing the driver with a higher performer is the best way to get more out of 1 box. We are very lucky that there are drivers coming out like the prv 15sw2000 and SB audience rosso 15sw800 for under 250$, especially considering that the 4015lf seems to be discontinued. At least it's not available from any of the usual retailers.

I will say the 24 inch wide box is pretty much the perfect size for a sub. With the kappalites they are about as loud as a single 18 reflex and weigh a tad bit more, and they are about the size of a compact double 18. But then you get all this fun horn physics. By the time you have 4 boxes it gets funky pretty quick. Now with the prv driver it seems the t48s maximum output becomes more comparable to double 18 reflex. Bring 4 of those, a v plate and shove it into a wall or corner and now you're cooking. This would award you pretty much as much spl as one could possibly want or need for most applications, dubstep included. And you would achieve this with less than 500 lbs of sub, taking up merely a 4x4x4 cube with a piece of wood on top. It's brilliant. It's like playing with transformers but they're really fking loud