I already have some equipment that will allow me to experiment with actively controlling things.
I have more than enough amplification on hand. This past year I built a new pair of stereo Babelfish J2 amps specifically for this purpose.
I think I mentioned previously I have a pair if MiniDSP 2x4 that I’ve used before. I also have the UMIK-1 and REW set up on my older laptop.
However, I was not ultimately satisfied with the sound quality of the MiniDSP when I used it with my Oris horns previously.
My goal would be to use the MiniDSP as a tool for establishing crossover points and slopes that I could mimic instead with some JFET based active Bi-Amp crossover kits that I purchased from the DIYAudio store.
I’m limited with the Bi-Amp kits as far as EQ in concerned…but I’ve never been a fan of EQing a problem to fix things.
I can use some L-pads if needed to adjust levels…and convert to resistors later if desired.
I also have some Hypex amps that I have been using to drive Dipole bass.
They work really well in that application and I would have no reservations in trying them to drive the LaScala low frequency horns.
Unfortunately my experience in trying the Hypex amps for the midrange and above is that they leave me with sense that they are sterile and lifeless to my ears.
I have more than enough amplification on hand. This past year I built a new pair of stereo Babelfish J2 amps specifically for this purpose.
I think I mentioned previously I have a pair if MiniDSP 2x4 that I’ve used before. I also have the UMIK-1 and REW set up on my older laptop.
However, I was not ultimately satisfied with the sound quality of the MiniDSP when I used it with my Oris horns previously.
My goal would be to use the MiniDSP as a tool for establishing crossover points and slopes that I could mimic instead with some JFET based active Bi-Amp crossover kits that I purchased from the DIYAudio store.
I’m limited with the Bi-Amp kits as far as EQ in concerned…but I’ve never been a fan of EQing a problem to fix things.
I can use some L-pads if needed to adjust levels…and convert to resistors later if desired.
I also have some Hypex amps that I have been using to drive Dipole bass.
They work really well in that application and I would have no reservations in trying them to drive the LaScala low frequency horns.
Unfortunately my experience in trying the Hypex amps for the midrange and above is that they leave me with sense that they are sterile and lifeless to my ears.
My 4 & 8 ohm derivations of the ALK Universal crossover are specifically tailored for the integration of the PRV D2200Ph with a 4 or 8 ohm woofer in a ported or non-ported LaScala and a B&C DE120/LaVoce DF10.142 tweeter.
The ALK Universal is only a 6dB/octave on the woofer and midrange high pass. A 6dB on the woofer made the upper bass from the LaScala bass enclosure seem too "tubby" for my tastes. Going to a 12dB/octave on the woofer, plus the acoustic roll-off of the low frequency horn gave me a nice steep roll-off around 425Hz. The 6dB midrange high pass on the ALK Universal was also changed to 12dB/octave because the D2200Ph is not supposed to be crossed over lower than 500Hz if you want to be able to use its full power handling ability (which most of us will not). Crossing the D2200Ph over at 12dB/octave @ 400Hz on the 18x10 horn has not had any negatives anomalies that we have heard or measured.
The D2200Ph will go out to 8kHz, but you want to limit it to around 6kHz so you are well within the polar pattern of the 18x10 horn when you crossover the upper midrange to the tweeter for a smooth crossover region (acoustically). The 12dB/octave low pass filter for the midrange has been maintained (as per the ALK Universal) but modified specifically for the D2200Ph frequency/impedance curve. The 18dB/octave tweeter crossover has been maintained as in the original ALK Universal too.
@gigantic won't steer you wrong.
I was disappointed with the sound of my La Scalas, even after adding bracing, replacing the stock tweeters with SMAHLs, trying several different passive crossovers (AA, ALK Universal, Universal/4500, AL), and adding room treatments, After reading several positive reports, I figured I'd try tri-amping them. Since I wasn't sure that the experiment would pan out, I wanted to keep the costs down. I used inexpensive Aiyima A07 amps and a MiniDSP Flex Eight DSP.
I've been entirely satisfied with the results. In my opinion, the sound is superb. I'm a convert!
Configuring the system required a lot of effort. You'll get to know REW really well by the time you're finished. But the results will be worth it.
I documented my efforts in this thread.
I have recently switched to passive bi-amping for my LaScalas. I wasn't able to get the bass horns balanced. I thought about something like that Dayton DSP 408 but it works at 48khz. I like the idea of being able to time-align the bass horns but I don't like the idea of effectively turning my high resolution digital files and vinyl collection into CDs.
My 300b amps seem to like the easier load that the high-pass section of the squawker network presents rather than driving the woofers. I'm using a line level low-pass filter for the bass amp but I'm running the upper amps straight through. It has taken me a while to get everything balanced but I'm very happy with the results. My system still sounds fundamentally like itself but it is better balanced top to bottom than ever.
Pete
My 300b amps seem to like the easier load that the high-pass section of the squawker network presents rather than driving the woofers. I'm using a line level low-pass filter for the bass amp but I'm running the upper amps straight through. It has taken me a while to get everything balanced but I'm very happy with the results. My system still sounds fundamentally like itself but it is better balanced top to bottom than ever.
Pete
The quest for better resolution has me looking at other DSP solutions for my setup, in order to get a better sample rate. It's worth noting, however, that the DSP-408 does it's conversions at a much higher bitrate, which has less of a deleterious effect on the sound than one might expect, given the bit rate, processing at 28-56 kbps.
I can't quite make out the model, but it's probably an ESMT solution.
https://www.diyaudio.com/community/threads/t-racks-4x4-mini-dsp-undressed.394296/
https://www.diyaudio.com/community/threads/t-racks-4x4-mini-dsp-undressed.394296/
Yes I agree. I can’t seem to get over my baggage on this issue.
I will say I’ve been happy enough with using the Hypex amps for bass support. Which in the LaScala scenario would allow for time alignment.
Since the mids and tweets need to be delayed compared to the bass horns I'm not sure how that works. I've decided to ignore the misalignment of the bass but getting the tweeters aligned with the squawkers is a nice improvement in focus and coherency.
Pete
From the Dayton Audio website:
"ADAU1701 SigmaDSP 28-/56 Bit Audio Processor"
It can use up to 56 bits for the math but it's running at 48khz as far as I can tell.
Pete
It doesn’t…except in my little world where I am thinking a$$ end backwards 🤪
Putting the tweeter on top of the mid cabinet so it can be slid back and aligned makes most sense…even if it may not be as aesthetically pleasing as mounting it in the same baffle as the mid.
When I tri-amped my La Scalas using DSP, I time aligned the drivers in the La Scalas.
To determine delays, I took 4-5 REW sweeps (and threw out any where the delay wasn't close to the delays from the other sweeps) and used the average. There were very slight variations between the two speakers, The woofers had no delay; the squawkers were delayed 1.24ms (left) and 0.96ms (right); the tweeters were delayed 2.70ms (left) and 2.41ms (right).
So, it seems that both tweeters are delayed 1.45ms compared to the squawkers. When I got 9.3 inches did I do the math right? That sounds wrong, roughly half the physical distance that I would expect.
I grossly aligned the diaphragms and used REW to generate a 5khz tone. I adjusted the position of the tweeters to maximize the output. The level changes several dB as I adjust the position forward and backward an inch or so. It feels pretty easy to find the middle.
The results sound great!
Pete
Are you guys talking past one another? No one seems to be talking about the order and type of crossover filters used, and these typically induce a large change in the delay values used to achieve time alignment at center crossover.
When tri-amping a La Scala II for a guy that lives in New Zealand, here are the settings that I found for him (noting of course that first order crossover filters are used--just like PWK used for the original La Scalas). The use of first order IIR filters is to minimize the all-pass (excess) phase growth of the entire loudspeaker--a big factor in how it sounds overall in my experience.
Also note that the midrange has no low-pass filter on it--just like PWK did for the original La Scala--because the cutoff of the midrange compression driver/horn is so sharp that no electrical/electronic/DSP filter is needed on the midrange high end. This saves an additional 45 degrees of phase growth for the entire loudspeaker. First the bass bin:
(Note the polarity reversal on the bass bin, above, in order to save 180 degrees of phase growth. This also affects the midrange/tweeter channel delays.)
Then the midrange:
Finally, the tweeter:
I used input channel PEQs also. Here are the two input channel PEQs used to flatten the amplitude further through the crossover bands. Remember, I was performing this remotely via his emailing of REW measurement files to me, and Xilica configuration files (.xdat) were sent back to the owner, thus completing one "round robin" trial. This reduces the latitude of how tightly the dial-in can be (the number of PEQs, etc.), but the result is the same--flat amplitude and smooth phase response:
The above was for the left-side La Scala II. The right side La Scala II has a few tweaks but the delays, channel gains, and most of the PEQs above the room's Schroeder frequency (about 150 Hz) were basically unchanged. This was a typical dial-in exercise that took about 8 round robins, including initial individual measurement of each driver (woofer, midrange, tweeter) to flatten their response.
Chris
When tri-amping a La Scala II for a guy that lives in New Zealand, here are the settings that I found for him (noting of course that first order crossover filters are used--just like PWK used for the original La Scalas). The use of first order IIR filters is to minimize the all-pass (excess) phase growth of the entire loudspeaker--a big factor in how it sounds overall in my experience.
Also note that the midrange has no low-pass filter on it--just like PWK did for the original La Scala--because the cutoff of the midrange compression driver/horn is so sharp that no electrical/electronic/DSP filter is needed on the midrange high end. This saves an additional 45 degrees of phase growth for the entire loudspeaker. First the bass bin:
(Note the polarity reversal on the bass bin, above, in order to save 180 degrees of phase growth. This also affects the midrange/tweeter channel delays.)
Then the midrange:
Finally, the tweeter:
I used input channel PEQs also. Here are the two input channel PEQs used to flatten the amplitude further through the crossover bands. Remember, I was performing this remotely via his emailing of REW measurement files to me, and Xilica configuration files (.xdat) were sent back to the owner, thus completing one "round robin" trial. This reduces the latitude of how tightly the dial-in can be (the number of PEQs, etc.), but the result is the same--flat amplitude and smooth phase response:
The above was for the left-side La Scala II. The right side La Scala II has a few tweaks but the delays, channel gains, and most of the PEQs above the room's Schroeder frequency (about 150 Hz) were basically unchanged. This was a typical dial-in exercise that took about 8 round robins, including initial individual measurement of each driver (woofer, midrange, tweeter) to flatten their response.
Chris