Following the clock on the wall, you are working pretty fast.
The versatile construction has room for easy changes - when needed. I like that. If the sound stays as transparent as the built, it should become a big success.
Rudolf
Thanks! Yeah, I move pretty fast when it's just a prototype. If I settle on one of these designs, I'll take a lot more time with it and make sure it's perfect with much better material / paint.
I hope it will succeed, I'm looking forward to another morning outside with the microphone, thanks to yours and everyone else's help, I'm starting to feel like I almost know what I'm doing
Edit: Versatile, exactly! That's why I made the mid mount like that, so if I want to try a different mid, it's 2 screws to remove it and then just move the bolts to the new size. Easy as pie!
So, I rather haphazardly tried to get some decent indoor measurements..which seems nearly impossible. I for the most part ignored peaks and dips in FR, instead just trying to shelf in some rolloff compensation, then (very crudely)time aligned the drivers.
First thing I've noticed, is I've lost a bit more output. Perhaps the woofers need a baffle, or maybe even a shallow U-Frame just to give them a bit more... UNF. The mid seems to be a bit more well behaved like this. The tweeter's output seems bizarre, but I'm going to attribute that to being indoors, and of course, being in this bizarre room.
The image and "space" is all here though, and it's really pleasant.
The feeling that I can't seem to get rid of though, across all these designs (and worsening as I lose more and more baffle) is they really seem to lack...authority? Grunt? I'm not sure what word I'm looking for, but they don't have that... *fist to palm* slam?
As long as I don't ask them for lots of output, they're extremely musical...but they're definitely not a techno / EDM machine.
Going to try to get outside tomorrow for another proper round of measuring and tuning, I'm sure there's a lot of mistakes. I'll post what happens.
First thing I've noticed, is I've lost a bit more output. Perhaps the woofers need a baffle, or maybe even a shallow U-Frame just to give them a bit more... UNF. The mid seems to be a bit more well behaved like this. The tweeter's output seems bizarre, but I'm going to attribute that to being indoors, and of course, being in this bizarre room.
The image and "space" is all here though, and it's really pleasant.
The feeling that I can't seem to get rid of though, across all these designs (and worsening as I lose more and more baffle) is they really seem to lack...authority? Grunt? I'm not sure what word I'm looking for, but they don't have that... *fist to palm* slam?
As long as I don't ask them for lots of output, they're extremely musical...but they're definitely not a techno / EDM machine.
Going to try to get outside tomorrow for another proper round of measuring and tuning, I'm sure there's a lot of mistakes. I'll post what happens.
Did you try to gate your measurements?
Some day you should try to stop that habit. It compromises the 90° null of the dipole 8. And it is not the proper method to get the crossovers right. You better move the crossover points or play with the order of the slopes.
People tend to think too much in visual uniformity and less in acoustical requirements.
You certainly have noticed how the MT baffles of the NaoNote and the Orion dwindled with time, while the woofer frames stubbornly stayed the same size. .Authority is something which happens below 100 Hz. You need the subwoofer to hear the complete system.
Did you try to gate your measurements?
Some day you should try to stop that habit. It compromises the 90° null of the dipole 8. And it is not the proper method to get the crossovers right. You better move the crossover points or play with the order of the slopes.
People tend to think too much in visual uniformity and less in acoustical requirements.
Authority is something which happens below 100 Hz. You need the subwoofer to hear the complete system.
I've been following the method of reversing polarity for the tweeter, increase it's delay until I get the largest null I can get, then reversing it back again..same for the woofer, similar to how gainphile describes it here. Are you saying I shouldn't do that, but rather just move the crossover points? Maybe even to different points?
I actually have been listening with both my subs. Bass authority seems great, the subs go all the way down to about 18 hz or so, and I can get enough level out of them to loosen drywall from the studs. There's just something about the region of 100hz - 500hz that seems...thin. It may be that I just need a U frame for the woofer section to get more dynamics in that range, it seems like that might be the best compromise.
I just played with my AINOs inside - it's raining here too! Phase matching is a very precise procedure. I took me many weeks to get in the ballpark, now I,m just fiddling. The acoustic measured output is what counts, and the dip should be symmetrical. Minidsp plug-in may have asymmetric or different slopes- only look at measured slopes. Just 0,01ms makes a difference when you are close. Below 1kHz matching gets difficult inside because of reflections and the need for long gating.
Acoustic phase match between drivers is waveform physics. Acoustic LR2 slopes - same phase, LR4 - inverted phase, when distance/delay in the measuring axis/distance is matched.
We must use parametric eq for individual drivers to make acoustic response to follow "the book of LR". Minidsp plugin allows several "cascaded" settings, that are needed specially with dipoles. John Kreskowsky, John Reekie and gainphile use them happily too with their latest dipoles.
The efficiency of a dipole construction is never good as a boxed speaker, but with a 4-way I think we get the best possible compromise in that respect too. Thinking of transients, a dipole/open transducer used at it's best range and without harmful box backwaves is a master! I enjoy piano music now more than ever before!
Acoustic phase match between drivers is waveform physics. Acoustic LR2 slopes - same phase, LR4 - inverted phase, when distance/delay in the measuring axis/distance is matched.
We must use parametric eq for individual drivers to make acoustic response to follow "the book of LR". Minidsp plugin allows several "cascaded" settings, that are needed specially with dipoles. John Kreskowsky, John Reekie and gainphile use them happily too with their latest dipoles.
The efficiency of a dipole construction is never good as a boxed speaker, but with a 4-way I think we get the best possible compromise in that respect too. Thinking of transients, a dipole/open transducer used at it's best range and without harmful box backwaves is a master! I enjoy piano music now more than ever before!
The method I've been using with regard to slopes:
Starting with the midrange, always.
1.) Measure the driver with no crossover or EQ.
2.) Add EQ (or let rew do it, if it's result looks acceptable) until the driver plays in the range that it should, somewhat above and below where the crossover will occur.
3.) Engage the crossover for the single driver and measure that the crossover is exactly where it should be and the slopes look how they should. Change or modify any of the EQ that was applied.
4.) Unmute the tweeter, invert the polarity. Measure and adjust it's level. Slowly add delay 0.2ms at a time until the response has a sharp null, then invert the tweeter back.
5.) Mute everything but the woofer. Measure it with no crossover or EQ.
6.) Add EQ (or let rew do it, if it's result looks acceptable) until the driver plays in the range that it should, somewhat above where the crossover will occur.
7.) Repeat 3 for the woofers.
8.) Unmute the midrange, invert the woofer's polarity. Slowly add delay 0.2ms at the time until I get a sharp null, then invert the woofer back.
9.) Unmute everything and measure the complete sytem. Use the global EQ in the MiniDSP to smooth things out, or let rew do it (it's actually pretty talented at doing it.)
If this is different from how it should be done, or if there's something I should do differently, please tell me
Well, I am don't know if there is a rule... John Reekie did almost the same as you (eq each driver before xo) I did that first too, but I've started from a fresh table many times! I am not sure what would be the easiest way, or best.
What I have ended up to is that I try to minimize PEQ. I have of course measured "raw" output without any eq ao 0¤ 30¤ and 60¤ to find out how each driver behaves. Then I make an overlay of them. Then thinking, trying, making mistakes, making new mistakes, learning something, trying new things and so on... Little by little the understanding gets better and practical settings find their places. The progress is not linear! Now I really understand why passive xo-guys warn wverybody for designing 3-ways!
I didn't save my measurements of today. I mostly tried to get woofer/lower mid integration better. It must be done in the room, and I measured at several placements (to the wall) and distances to the mic. At listening position and also at 1m distance (phase). I had to use 120-200ms IR to see 100-400Hz area, 1/3 smoothing is good to see forest from the trees. Then reflections are a problem and the change a lot when I move mic or speaker 10cm (4") Simply trial and error - I made at least 100 measurements in 2 hours and drove the family crazy!
I use room eq only under 150Hz (in the plugin input PEQ), but never when setting individual drivers. I do indoor measurements but I verify them outdoors when I can. Only after that I allow myself to use "fullrange room eq"
DrDyna, it woul be nice to get over and to these together! We can never tell everything we do and think of, here at diyaudio.
What I have ended up to is that I try to minimize PEQ. I have of course measured "raw" output without any eq ao 0¤ 30¤ and 60¤ to find out how each driver behaves. Then I make an overlay of them. Then thinking, trying, making mistakes, making new mistakes, learning something, trying new things and so on... Little by little the understanding gets better and practical settings find their places. The progress is not linear! Now I really understand why passive xo-guys warn wverybody for designing 3-ways!
I didn't save my measurements of today. I mostly tried to get woofer/lower mid integration better. It must be done in the room, and I measured at several placements (to the wall) and distances to the mic. At listening position and also at 1m distance (phase). I had to use 120-200ms IR to see 100-400Hz area, 1/3 smoothing is good to see forest from the trees. Then reflections are a problem and the change a lot when I move mic or speaker 10cm (4") Simply trial and error - I made at least 100 measurements in 2 hours and drove the family crazy!
I use room eq only under 150Hz (in the plugin input PEQ), but never when setting individual drivers. I do indoor measurements but I verify them outdoors when I can. Only after that I allow myself to use "fullrange room eq"
DrDyna, it woul be nice to get over and to these together! We can never tell everything we do and think of, here at diyaudio.
Last edited:
It's true, sometimes I wish there was a hifi club or something in the area, it might be fun.
I hadn't noticed that gainphile is doing it that way. You might get away with it for the WM crossover, but I believe that I see the negative influence in your MT crossover already in this diagram: http://www.diyaudio.com/forums/atta...ive-open-baffle-design-front-yard-0-45-90.jpg
If you could tell us your respective delay values, I would go into more detail to explain.
Yes, move the crossover points or possibly change crossover orders or polarities.
Your subwoofer is radiating as an omni at the crossover frequency, your OB woofers as a dipole. The consequence: For equal SPL at 0° the sub will radiate 4.8 dB more energy into the room. You need to keep the subwoofer level (on axis) at least 3 dB below the OB level to get any balance.
As said, it is all perfect excluding the delay. Just apply the usual methods of crossover design without resorting to the delay.The method I've been using with regard to slopes:
....
If this is different from how it should be done, or if there's something I should do differently, please tell me
Rudolf
Cool, yeah, this weekend I'm going to go outside for some measurements and try to get everything set properly.
Now that I'm getting closer though and I've been listening more and actually giving these some power, the issues with the MiniDSP only having 0.9vrms of output is starting to become an issue, especially with the amounts of boost needed for shelving filters.
I'm not sure what I'm going to do, I don't really want to have to order 4 stereo pre-amplifier kits, but I guess I might have to.
Now that I'm getting closer though and I've been listening more and actually giving these some power, the issues with the MiniDSP only having 0.9vrms of output is starting to become an issue, especially with the amounts of boost needed for shelving filters.
I'm not sure what I'm going to do, I don't really want to have to order 4 stereo pre-amplifier kits, but I guess I might have to.
I have some circuit boards that I developed that you could use to implement gain stages. They use thru-hole components, and dual op amps. The layout is general in that it allows for many types of circuits to be built, depending on how the board is populated with what components and where. I mentioned them in the MiniDSP forum thread you posted in...
Each 10cm x 10cm circuit board has 2 identical independent channels on it. A logical connection diagram for one channel is pictured below. Each labeled position can be populated with a resistor or capacitor, shorted with a jumper, or left open. Radial caps with LS up to 15mm can be supported. Drop me a line if you are interested.
-Charlie
.
Each 10cm x 10cm circuit board has 2 identical independent channels on it. A logical connection diagram for one channel is pictured below. Each labeled position can be populated with a resistor or capacitor, shorted with a jumper, or left open. Radial caps with LS up to 15mm can be supported. Drop me a line if you are interested.
-Charlie
.
Thanks Charlie. Yeah, I saw that in the other thread. It does look interesting. I'll be considering something like that once I get closer to completion.
I wonder something though, could the circuit be done with a transistor or maybe even a VT instead of an OPamp? I know it probably doesn't make much difference, I'm just curious how hard it would be.
I wonder something though, could the circuit be done with a transistor or maybe even a VT instead of an OPamp? I know it probably doesn't make much difference, I'm just curious how hard it would be.
If you needed only to buffer the signal, you could build something relatively simple based on a single transistor, but in this case you need an amplifier (e.g. gain). That's a completely different mountain to tackle. A gain stage can be constructed using an op-amp with very few additional parts, and op amp electrical properties are well documented so you can do a reliable design that has low distortion, appropriate bandwidth, etc.
Not sure where you are coming from, but as Douglas Self mentions in his recent book on active filter design (pp.410-411) regarding the NE5532:
"With tiresome inevitability, the very popularity and excellent technical performance of the 5532 has led to it being savagely criticized by Subjectivists who have contrived to convince themselves that they can tell opamps apart by listening to music played through them. This always draws a hollow laugh from me, as there is probably no music on the planet that has not passed through a hundred of more 5532s on its way to the consumer."
Knowing the limitations of various op-amps and how to best make use of them in a circuit is important for getting them to work well in the intended application.
-Charlie
You can easily just purchase a MiniDSP 4x10HD, it has internal jumpers for setting output 0,9 or 2,0mV(max) for each output channel individually
Balanced outputs (Terminal block): 4Vrms (+14 dBu) or 8Vrms (20dBu)
It costs a lot but you can sell your 2x4s or use them in some other project.
Regarding delays for trimming crossover phases, the difference needed in ms is really small. My AINOs have 0,00ms for woofer, 0,13ms for Low mid, 0,06 for high mids and 0,00 for Tweeter. I have noticed that for M to T a change 0.02 ms makes a difference! At 3000Hz a full cycle of wave is 1,13ms. 1ms means 34,0cm distance of sound sources. That means that you'll easily set delay several cycles late! (correct me if comma is at wrong place!) Here is John Reekie's advice for setting delay (scroll down to "Time alignment of drivers").
Dip switch on motherboard attenuates outputs by 6dB (RCA & Terminal blocks)
Unbalanced outputs (RCA): 2Vrms (+8 dBu) or 0.9Vrms
Unbalanced outputs (RCA): 2Vrms (+8 dBu) or 0.9Vrms
Balanced outputs (Terminal block): 4Vrms (+14 dBu) or 8Vrms (20dBu)
It costs a lot but you can sell your 2x4s or use them in some other project.
Regarding delays for trimming crossover phases, the difference needed in ms is really small. My AINOs have 0,00ms for woofer, 0,13ms for Low mid, 0,06 for high mids and 0,00 for Tweeter. I have noticed that for M to T a change 0.02 ms makes a difference! At 3000Hz a full cycle of wave is 1,13ms. 1ms means 34,0cm distance of sound sources. That means that you'll easily set delay several cycles late! (correct me if comma is at wrong place!) Here is John Reekie's advice for setting delay (scroll down to "Time alignment of drivers").
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.