AH, ok. It's at 26" right now, I can scoot it out a bit if I need to. Going to run up and get some more gator clips, I seem to have run out!
Testing will resume shortly.
Thanks for all the help guys, I need to name these things after all of you, right now I've only been referring to them as "the grumpy cats".
Testing will resume shortly.
Thanks for all the help guys, I need to name these things after all of you, right now I've only been referring to them as "the grumpy cats".
Really only one driver is attached by it's basket ... the 12", the 15" is resting on a concrete slab with a simple magnet clamp, the 5" is suspended off the heil, heil is suspended off the frame.. ... but as you said I'd like to try suspending all of them a bit differently 🙂
What values do you guys use for tweeter protection caps, if you use any at all? Do you think 12uF is good enough?
The torque on the frame will be immense if he's only holding by the lateral holes.
EDIT: I see he is daisy-chaining holes. He'll have to build compliance into the structure due to tolerance stack-up.
Last edited:
We used 10 µF, 0.12 mH and 22 µF as a third order HP for the Neo3. 12 µ is ok as a 1st order cap.
Last edited:
Well, this is certainly an interesting effect. Currently with a very terrible passive crossover (I haven't obtained my second miniDSP yet, I'm on the fence about ordering another one or ordering a DCX2496) they have a really neat ability to defy their position. I flopped down on a chair that's pretty much in line with the left channel, yet I didn't really get any image leaning that I'd normally expect..and walking around in the room doesn't really "warp" the FR like I'm used to with box speakers.
They do seem a little thin sounding, but that's probably my hodgepodge passive crossover at work.
I'll attach what I was able to achieve with my single MiniDSP working on just one channel.
Cost aside, what's the general opinion in DCX2496 vs 2x 2x4 MiniDSP? I understand the Behringer unit probably offers a little more output, but the flexibility seems similar.
They do seem a little thin sounding, but that's probably my hodgepodge passive crossover at work.
I'll attach what I was able to achieve with my single MiniDSP working on just one channel.
Cost aside, what's the general opinion in DCX2496 vs 2x 2x4 MiniDSP? I understand the Behringer unit probably offers a little more output, but the flexibility seems similar.
Attachments
The experiment has evolved into a hybrid of sorts, using the MiniDSP for the mid and tweeter, with an 8.5mh / 250mfd on the woofer. Measured, that rolls it off very gracefully around 12db/oct starting around 400 hz.
Is it just me or does the MiniDSP sound terrible with more than about a quarter of a volt of input?
The purchase decision is definitely swinging toward the Ultradrive.
Is it just me or does the MiniDSP sound terrible with more than about a quarter of a volt of input?
The purchase decision is definitely swinging toward the Ultradrive.
It's just you, or how you are using it. Check all levels. Internal level should never go above 0dB, so if you are applying any combination of boost or EQ that totals more than about 12dB at any frequency, then you are likely exceeding the maximum internal levels. This is why I was arguing for 10dB-20dB gain stages after the MiniDSP - so you can reduce the gain inside the MiniDSP (e.g. at the input in this case) and then get back gain in the analog domain.
-Charlie
I've been using minidsp for 2 moths now. It starts buzzing sometimes after I save my settings (during measuring session). And you must be careful not to push it over 0dB total. It is teasingly easy to add gain in PEQ settings, but after xo settings you must be on negative values. A dipole needs lots of shelving eq which lead to lower efficiency in bot passive and digital-sp systems. Adjustable gain at analog stage is nice. I use B&O ICEpower modules 50 and 125 - this is my way of compensating this (difference perhaps 6dB)
Another issue mentioned earlier was to not use double mids in a dipole system. I don't agree! I measured my AINO today out with and without second serial high-mid. Seems like they have similar dipole behaviour in my system, but a single mid has shaper reflections and sharper reverse null - more precise phase behaviour measured with a microphone at a spesific point (I didn't try to adjust it in single version) I haven't yet listened WMT inside vs. WMTM, but outside I couln't hear a difference. We must remember that a muldtidriver dipole alwasy has deteriorating reflections on the backside from orther drivrs' bodies. It can't be avoided.
Here are vertical polar simulations of a MT an a MTM with 100mm distance M-T that I made for AINO (MTM looks better to me)
Another issue mentioned earlier was to not use double mids in a dipole system. I don't agree! I measured my AINO today out with and without second serial high-mid. Seems like they have similar dipole behaviour in my system, but a single mid has shaper reflections and sharper reverse null - more precise phase behaviour measured with a microphone at a spesific point (I didn't try to adjust it in single version) I haven't yet listened WMT inside vs. WMTM, but outside I couln't hear a difference. We must remember that a muldtidriver dipole alwasy has deteriorating reflections on the backside from orther drivrs' bodies. It can't be avoided.
Here are vertical polar simulations of a MT an a MTM with 100mm distance M-T that I made for AINO (MTM looks better to me)
Last edited:
Thanks for posting about your experiences. It seems to me that many people experience this problem with OB systems or LT boosted subwoofers and don't know that internal (digital) levels that are pushed too high cause obvious distortion. They just think that the product sounds crappy!
I wanted to comment on what you said about dual midranges: anytime you have two drivers separated by more than about 1/4 to 1/2 wavelength the off axis response will have nulls. The transition to where this starts to happen is at a frequency that corresponds to about 1/4 wavelength, and as you go up in frequency the number of nulls increases and the "pattern" narrows. If you only measure directly on axis from a point in between the two drivers you will not see this behavior. This is because the nulls occur when the pathlengths driver-A --> listener and driver-B --> listener happen to equal 1/2 wavelength and the phase is 180 degrees different. But on axis the pathlength to each driver is the same, so interference is not observed. Besides being "box less", the open baffle should radiate power into the room evenly because you are hearing more sound reflected by the room in an OB speaker, so you don't want off-axis nulls. In general this is always a bad thing IMHO, but if you only listen on axis and the speakers sound OK in your particular listening environment the "problem" may sufficiently benign to be ignored. But in general it's a good practice to avoid it.
-Charlie
You posted these after I started my reply, so let me address these separately.Here are vertical polar simulations of a MT an a MTM with 100mm distance M-T that I made for AINO (MTM looks better to me)
You need to simulate just the two woofers. To do that in Xdir, set driver placement to woofer-tweeter-woofer, enter the center-to-center distance between the woofers in the "Driver distance" box, and set the tweeter level to -99 to remove it. Set the woofer level to 0 dB. The woofer phase lag doesn't matter in this case, any value will produce the same result.
Now you will see the pattern from only the woofers, including their nulls. Below the crossover point to the tweeter, this is the response that this geometry will generate. Also, this is a simplistic model, I think it models the drivers as point sources so the nulls will not be as deep as shown. You can do more realistic modeling (but no a polar plot) using TheEdge from Tolvan.
To avoid the nulls you need to cross over to the tweeter before they become significant. Remember there are two separate issues here: the nulls you will get just between the two woofers at higher frequencies, and other nulls you will get between the two woofers and the tweeter in the crossover region.
Finally, keep in mind that this is only a model of drivers. It's assumed they have infinitely wide bandwidth from 0Hz to infinity and that the phase angle is zero for all frequencies. In the real world, drivers are like bandpass filters, the phase changes with frequency, and the crossover filters also introduce phase rotation differently to the woofer and tweeter. So, while you can make some observations with this kind of model, you need to take into account all these other things when designing a loudspeaker in order to know where the phase response will cause problems (like interference) and how to control that.
-Charlie
Last edited:
I'm actually being very careful with the input and output levels. I had the MiniDSP software open and was watching output levels (all my EQ settings are done with subtraction only, no boost, input jumper set to 2.0v) and it would seem that every time the level meter started peaking near the center of the scale (still very green) it starts to sound a bit like a 128k mp3..and if pushed any harder (into the yellow) female voices start to take on a harshness that makes me wince and grit my teeth.
Maybe there's some interference, maybe it's the wall wart I'm using to power the MiniDSP...I don't know...which is why I'm thinking about the DCX2496....it seems like it might be a little more...robust? (read: I could still use my system after having a few drinks and not let the smoke out)
Think twice about that. The DCX has just the opposite problems 🙄. You are immediately in trouble with the DCX if your analog sources and power amp stages don't have PA levels. Feeding the DCX with too low level is as bad as feeding the miniDSP with too high level.
Rudolf
Charlie said " because you are hearing more sound reflected by the room in an OB speaker, so you don't want off-axis nulls."
I disagree again! a Ob/dipole radiates less sideways and more backwards than a boxed speaker - ob is less prone to comb filtering! Double-drivers create nulls, so do all separate xo'd drivers because of overlapping frequencies. Comb-filtering nulls can't be avoided in any multidriver system (unless one uses brickwall filters). This is exactly where only a point-source full range driver masters. All others are deficient.
I can easily hear nulls when I play sine waves, but I can't hear any difference with music. From this graph you can see that I have lots of overlapping frequencies and potential comb-filtering in my 4-way AINO.
All speaker systems have flaws and speaker building is like cooking - we mix raw materials, process them, add spices etc. to get the combination that makes one/us happy!
I disagree again! a Ob/dipole radiates less sideways and more backwards than a boxed speaker - ob is less prone to comb filtering! Double-drivers create nulls, so do all separate xo'd drivers because of overlapping frequencies. Comb-filtering nulls can't be avoided in any multidriver system (unless one uses brickwall filters). This is exactly where only a point-source full range driver masters. All others are deficient.
I can easily hear nulls when I play sine waves, but I can't hear any difference with music. From this graph you can see that I have lots of overlapping frequencies and potential comb-filtering in my 4-way AINO.
All speaker systems have flaws and speaker building is like cooking - we mix raw materials, process them, add spices etc. to get the combination that makes one/us happy!
Charlie said " because you are hearing more sound reflected by the room in an OB speaker, so you don't want off-axis nulls."
I disagree again! a Ob/dipole radiates less sideways and more backwards than a boxed speaker - ob is less prone to comb filtering! Double-drivers create nulls, so do all separate xo'd drivers because of overlapping frequencies. Comb-filtering nulls can't be avoided in any multidriver system (unless one uses brickwall filters). This is exactly where only a point-source full range driver masters. All others are deficient.
I can easily hear nulls when I play sine waves, but I can't hear any difference with music. From this graph you can see that I have lots of overlapping frequencies and potential comb-filtering in my 4-way AINO.
All speaker systems have flaws and speaker building is like cooking - we mix raw materials, process them, add spices etc. to get the combination that makes one/us happy!
I strongly disagree:
ob is less prone to comb filtering WRONG!
Comb-filtering nulls can't be avoided in any multidriver system (unless one uses brickwall filters) - WRONG!
This is exactly where only a point-source full range driver masters. All others are deficient. -WRONG!
First off, any time you have "multiple sources" reproducing sound at any given frequency and those sources are more than about 1/4 of a wavelength apart you will begin to get nulls off axis. This is pretty well established. Did you ever stop to think that a single "point source" (no such thing really) full-range driver has multiple sources when the frequency is high enough - different areas of the cone are different distances away from an off axis point, and thus create nulls. Not just as bad, but worse, because the cone is also going through breakup at high frequencies. It's rubbish.
What about a multi-driver system? This allows you to use a different driver for each frequency band. As the frequency becomes higher, a smaller driver is used. Crossover between one driver and the next has to happen at an appropriate frequency to avoid off-axis interference. I try to use a robust dome tweeter and cross over "low" (e.g. 1.5kHz max) to a small midrange - the two drivers are mounted with the edge of their frames touching, to be as close together as possible. The other option I have seen used is to cross over so high (5kHz or higher) that the sound from the two drivers is uncorrelated and just adds, but you need to use a very small midrange for this (like 5cm!). I recently made an MTM system using a small "full range" driver having a 1.5" diameter radiating surface as a tweeter, so I could cross over at 700Hz second order. Works and sounds great and no comb filtering.
Why don't you show some off-axis frequency response measurements for your AINO project and then we can talk about it some more.
-Charlie
What does the graph show? On axis or off? Distance? Direct sound only or including reverberant field?
Thin sounding and bad female voices - what if you simply equalized for too much highs 😕
Hi,
I have both DCX2496 and Minidsp 2*4, my opinion is that the advantage of getting the Ultradrive, you are working in digital mode throughout whereas the Minidsp you are actually converting digital to analog using your Dac and that from analog/digital/analog via the Minidsp. Having said that Minidsp do sound a little better than Ultradrive, also a few owners/reviewers have reported this.
If you want to keep the Ultra in the long term then you need to modify it there are a few DCX mod sites and also owners on this forum in regard to DCX modifications.
Another alternative is to buy a second hand Ultradrive(quite old already) for testing and then buy a new model of Minidsp because there will be new models coming out. This area is changing rapidly.
Cheers.
Charlie, I don't want to hack this thread. I have a link to my AINO project in every post when I refer to it. Here again. You can see that I have done lots of off axis measurements. There you can see 0¤ 60¤ horizontal off-axis of just a pair and single Vifa NE95 mid drivers.
Case closed for my part here. We can go on in my thread.
Case closed for my part here. We can go on in my thread.
I let REW generate the EQ curve, but I realized today that I left Audyssey engaged when I did the measure and EQ, so I'm going to redo it today and see what I get.
Don't think this presents a real problem:
Gain Structure for Home Theater: Getting the Most from Pro Audio Equipment in Your System - Home Theater Forum and Systems - HomeTheaterShack.com
There's definitely no adequate light show from the LEDs but I wouldn't be concerned about SNR.
I would be more concerned with the occasional freeze of the DCX and SRC problems when fed digitally.
- Status
- Not open for further replies.