Jump to content

ZOD Audio M.A.U.L. Test Results and Discussion


Contrasseur

Recommended Posts

I didn't run any numbers, and I can't say I've studied this design enough to know what I'm doing yet.  But I did also have thoughts along similar lines about using DSP to alter signal to each driver to make them load better.  I also thought about how the that would affect the port loading and concluded it likely wouldn't be a problem because they are in different ranges of frequencies.

 

However, I see a big problem with doing this.  Once the system starts to be pushed, it will no longer behave linearly, and the DSP will no longer work as intended.  Because of this, I would be very concerned about instability of the system.  One moment, you're rocking the house, the next minute, the drivers are machine gunning back-and-forth between their mechanical extremes.  Ouch.

 

Of course, I don't know for sure how it would misbehave.  I would expect to have to do a fair amount of modelling and/or testing to figure this out, and there is the risk that the design is inherently unstable.  On the other hand, a closed-loop feedback control system could be used to stabilize the drivers while also potentially extracting a bit more performance out of them.  Indeed, if I understand right, the IPAL system is *designed* to work this way.

 

As an aside, I'm not sure where "closed loop" came into subwoofer terminology to mean "a design with a built-in amp" because in controls theory, closed loop implies a feedback loop within the system.  Rythmik and the flagship Velodyne are designs I'm aware of that would fit this later definition, but most vendors of amp+driver integrated systems do not sell closed loop systems, in a controls sense.

 

The IPAL's super powerful and efficient motor with well-managed inductance attached to a light cone driven by a high voltage amp gives the DSP a lot of loop gain to work with and minimizes physical response latency.  If the overall latency is low enough, the system can control the driver motion very precisely, even when it's being constantly pushed to its limits.  Xmax be damned here, because the control system can force the cone to move linearly so long as there are enough volts (and thermal headroom) for it to do it's thing.

 

I do think a M.A.U.L. with IPAL drivers and integrated feedback control would have no problem topping this one in performance, but that's not an easy thing to pull off.  For example, I don't think I could get the latency low enough on my PC-based system to do it.  I could be wrong, but I would have to do a lot of study/research to get started.  I'm guessing that the latency needs to be quite low to adequately suppress higher order harmonics.  But wouldn't it be awesome?  The IPALs have ridiculous mechanical stroke, don't they?

Link to comment
Share on other sites

  • Replies 166
  • Created
  • Last Reply

There are a couple of major issues with the possibility of delaying the offset drivers to correct the cone excursion behavior seen in the 21-Ipal M.A.U.L. version, IMHO of course. I already outlined most of them in my previous post. I would've built the 4x 21 version as it can offer some improvements on paper. Higher sensitivity, flatter mid band response, more extended upper range response, lighter weight, slightly lower cost, and smaller size which allows more room in the cabinet for bracing or different port configurations. They have about 4-5dB less functional displacement though so the output in the deep bass would be less and since they would be operating closer to their excursion maximum THD would probably be a bit higher in the frequencies corresponding to high driver excursion.  The Ipal surrounds are done way before 40mm xmech. The coil leaves the gap completely at 33mm. Functional useful excursion with reasonable distortion is about 20-25mm one way. Plenty of output still but I wouldn't say it would have topped this one in all aspects of performance because of the loss of deep bass headroom. Better in some aspects yes but not all. One huge one being that the 19's do not need 2 separately processed amp channels. Plug in any amp comfortable with 4ohm nominal and run.

 

The differences in driver behavior are related to the offset positions in the short horn that is clear. My current thought is that the differences in excursion behavior are primarily related to the 19's having double the mms of the 21-Ipals. This design does not have separated back chambers for the driver pairs. It would have been extremely difficult to make that happen with the driver compliments and cab space available for this build. A folded slot port for the front driver pair would have eaten a lot of internal volume, not to mention the internal cab walls and a carry through slot port for the back 2 drivers. The port geometry would've been greatly different for the 2 sets and the performance might have been slightly different too. It could be done and the performance differences probably wouldn't amount to enough to lose sleep over but the complexity and added chance for error are not worth it for me. That could be done much easier in a different design which allows more space for the drivers and perhaps a different form factor and / or allowing for the output to radiate from multiple faces of the enclosure.

 

My #1 issue with trying to correct by delaying the driver pairs is using 2 amplifiers or at least 2 channels with differing signals to drive one cabinet. There was some talk of the IPAL or MForce DPC circuit to do this. You are immediately talking about 2 amplifiers in one cab to pull this off. Each would drive one set of drivers and each would have its own DSP settings and its own DPC circuit attempting to control the drivers. It is possible that one my be attempting to offset the effects of the other set of drivers and vice versa setting up a catastrophic run away condition. In fact in this type of case the DPC circuit will invariably be picking up the behavior of the second set of drivers almost as much as the primaries. I'm almost certain that running 2 DPC feedback circuits in one common shared cab is a recipe for failure. It would be very unstable and risky IMO. I would not attempt it. In fact I bet Powersoft would tell you absolutely not to try this and probably void the warranty on the amps if this was attempted. A different type of feedback circuit similar to using an accelerometer, laser measurement or sensing coil could work potentially. The actual IPAL system I would say no.

 

Now if we get rid of the double feedback loop circuitry and use normal amps and dsp it should work. You still have to use 2 channels and process them both slightly different and use some investigative measurements to see what the right amount of delay might improve the situation. That should be do able. I'm not entirely sure it will fix or even improve the issue much. Hell I'm not even entirely sure that the 21Ipal drivers would have had the simulated issue at all. Perhaps the real cab would've not performed as badly as suggested by the sims.  Other scary thoughts...What if you get the processed signal backwards to the drivers? What if one of the amp channels or amps fails while operating? I just didn't feel like taking these extra chances on a project this big which already had a lot of chances being taken.

 

Mike and SME,

We should get away from this specific design and just talk about delay or feedback correcting adjusting asymmetric loaded drivers in slots, throats or horn paths in general because this isn't specific to this type of cabinet. If care was taken at the outset of a ground up design to plan for it and compensate it, it might be a worthy approach. I would like to see someone investigate it. That would be really interesting data. I do have the second cab still sitting here waiting on drivers. I could do some further testing when I get it fired up. I already placed the first one so I don't want to move it again unless I absolutely have to and I'd have to rewire the drivers to drive the pairs separately. This would probably be best investigated using some small cheap drivers in a 5min "design" with scrap wood made just to test the feasibility of this specifically. Could be a TH, a FLH, slot loaded bandpass, BLH whatever. Any takers? Perhaps someone already has a smaller less involved design that could be used to investigate it. I know I've seen other horns out there with drivers loaded into different spots in the path.

 

Couple shots of one of the 21-Ipal MAUL designs. Also had variations with a top to bottom slot port and also a more square slot vent.

post-5-0-74414700-1470237747_thumb.png

post-5-0-80108100-1470237759_thumb.png

post-5-0-83842600-1470237773_thumb.png

post-5-0-69507200-1470237786_thumb.png

post-5-0-35981200-1470237799_thumb.png

Link to comment
Share on other sites

I'd like your opinion on this hypothetical, Josh...

 

Let's say someone else builds the MAUL with quad RFT3 drivers but can't or won't purchase the rather pricey (hey, I spent all my money on these damn RF drivers!!) K20 amplifier.

 

But they do have (or can afford) the ever popular Speaker Power something-or-other. Let's say the 12000. *cough totally not fishing info for Luke cough cough*

 

Would you recommend dual 2ohm hookup (~3k per driver) as a single 4ohm bridged load won't work with this amp or several others that can not be bridged?

 

Just wondering and this is more simple than requesting new designs for the MAUL. :rolleyes::lol::P

Link to comment
Share on other sites

That'd work just fine unless one of the modules quits during high output playback. Kind of doubtful it'd happen but you never know. Actually I may repeat some of the tests on the second cab once it's all put together using the 12K in just that configuration instead. I'd like to see how it does compared to the K20. I think I already know which is the 20 does better in the bursts and the 12K does better with the long term sweeps, but I'd like to see how much of a difference there is for each either way. Similar to the tests I already did at home but with a different load for another set of data points.

 

I wouldn't clone this cab even if using the RF drivers. If the dimensions or shape can be changed or enlarged or whatever the cab design could probably be improved and simplified a lot. If using normal drivers that can be back mounted I'd totally have the horn slot in the center and switch to dual vented chambers on the outside for bracing and structural purposes. That was how the final 21-Ipal design would've been.

Link to comment
Share on other sites

I did finally get the first one in place and had a chance to listen to it the other night. I have some work to do getting it blended in with the rest of the speaker system. I didn't have my measurement gear there so I just used my best guesses and my ears. Run without any processing it sounds like ass. No surprise. All of those peaks above 100Hz sound nasty. Once a suitable low pass filter was engaged things improved dramatically. So far it sounds good. I didn't get a chance to blast it yet either. I'll wait until I have it somewhat integrated.

 

For anyone wondering what type of signal processing is needed to get it sounding smooth and extended here are some clues. :D

 

post-5-0-13300600-1470251307_thumb.jpg

post-5-0-87730900-1470251317_thumb.jpg

post-5-0-84745200-1470251324_thumb.jpg

 

That's nothing but a single low pass filter in each case.

Link to comment
Share on other sites

My bad.  I think confused it with some other driver that has more like 45 mm one-way travel.

 

I admit, I didn't think *too* hard about the series horn loading because I would probably still opt to build cabs with two drivers each instead of four.  I much prefer the idea of having multiple smaller cabs both to reduce weight and increase flexibility.

 

I did do some quick investigation of a half-size / 2 driver version for my living room, but I still don't have enough volume to pull off the low tuning frequency without other unacceptable compromises, and I'd rather tune even lower if I could.  About 12 Hz would hit the spot nicely but just won't fit.  At least in my case, the extra extension < 10 Hz and > 100 Hz with sealed may be a bigger win.  But in the long run, if I got more interested in doing commercial scale audio, I would definitely consider going down this road.

Link to comment
Share on other sites

Doesn't need to be 4 drivers or even 2. Could be just 1. Think of it as a modular design. I like 2 drivers in one cab for force cancellation. 4 begins to be unmanageable as far as size, weight, cost and design complexity. You want louder then make more cabs.

 

Alternate M.A.U.L. tuning would've been around 21Hz if I recall. It improved the output above 20Hz quite a bit. I really wanted a 3 way tuning with one tuning up near 30Hz. Would've gained nearly 10dB at 30Hz. Just couldn't fit it all in with my limited options. Wish I could've. If I get time maybe I'll post some about my Skhorn design.

Link to comment
Share on other sites

Go with one driver per box, but build them in pairs with high clamping force buckles and a layer of soft material to go between each pair.  Now you have the best of both worlds: transportation as smaller single-driver cabs and force cancellation when in use.

Link to comment
Share on other sites

Funny I was looking at using a pair of 21IPALS's or go the quad 18IPAL route. The great thing about the 18 was the fact that they are 2 ohm verse 1ohm. SO I hope to see more on your dual 21 build in the future. 21hz would be low enough for my wants and the higher tuning would be fun.

Link to comment
Share on other sites

  • 2 weeks later...

We should get away from this specific design and just talk about delay or feedback correcting adjusting asymmetric loaded drivers in slots, throats or horn paths in general because this isn't specific to this type of cabinet.

 

Very true. I guess it came up here because it wasn't obvious what was happening in your simulations.

 

One other thought to consider....Roy Delgado over at Klipsch swears by having the port exit into the throat of the horn section. There's no room for that in this design given your original requirements, but it would be easy to start with this topology and extend that back wall a bit for a new design. The size of that front horn doesn't seem too far off from a slowly flared vent anyway - so maybe there's some total system volume gains to be realized? I think that starts to cross over between that gray line of tapped horn and venting into the throat....

Link to comment
Share on other sites

Very true. I guess it came up here because it wasn't obvious what was happening in your simulations.

 

One other thought to consider....Roy Delgado over at Klipsch swears by having the port exit into the throat of the horn section. There's no room for that in this design given your original requirements, but it would be easy to start with this topology and extend that back wall a bit for a new design. The size of that front horn doesn't seem too far off from a slowly flared vent anyway - so maybe there's some total system volume gains to be realized? I think that starts to cross over between that gray line of tapped horn and venting into the throat....

 

I've not actually simulated what happens in that case. Can HR even handle that type of simulation?

Link to comment
Share on other sites

Hi, 

 I wonder how this looks inside? It uses Ipal driver-amp combo on the active version. I appeared around 2008 but never heard or seen one. I wonder if it has a bit of horn in one of those vents? the numbers and mumbo jumbo marketing doesn't reveal anything, a graph would show us much more

http://www.outlinearray.com/images/box/lab21/lab21.pdf

Link to comment
Share on other sites

"30mm peak," what driver are they talking about? The SW152. The IPAL has 40mm peak although Ricci has stated it wont ever go that far. SO 25mm would be a more realistic peak. And I am guessing that the design linked must be about 30hz tune or something. All mine that tuned at 25hz are at about 19/20hz -10db with 4th order Hpass in place.

Link to comment
Share on other sites

  • 2 weeks later...

Figured I would ask hear, so what types of port speeds are ok at 140db?

 

I ask because 2800cm2 is starting to get a little too big for a 25hz tune. :D I saw in your build you had a 730cm or something like that but it was for a much lower tune. SO speeds would have been much lower.(Perhaps)

Link to comment
Share on other sites

I've got a ton of people asking about this type of cab for various apps end I just don't have time right now. Once I get some bandwidth opened back up I'll start working on finishing my 21Ipal music version and hopefully have time to work on a few others.

Link to comment
Share on other sites

I've got a ton of people asking about this type of cab for various apps end I just don't have time right now. Once I get some bandwidth opened back up I'll start working on finishing my 21Ipal music version and hopefully have time to work on a few others.

 

Good stuff, appreciated as always!

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.


×
×
  • Create New...