Jump to content

Settings for tri-amping an '82 Belle with an active digital crossover


Chris A

Recommended Posts

Interesting, That's a new concept on me.

Perhaps I could demonstrate what I'm describing if you were here. Alas.

Chris

Link to comment
Share on other sites

In order to tri-amp the Belle's drivers, I bypassed the Belle's "AB" passive crossover and connected to each driver using the AB crossover terminal blocks to each amplifier channel. By the way--the difference in sound quality is pretty significant, IMO.

Thanks for posting your specs on the Belle at the start of this thread. I realize the specs will be different for K-horns or La Scalas, etc. but yours are a good guideline.

I have read the pros and cons about going active tri-amp, especially the extra complexity, so as a first time tri-amp planner / user I just want to know the proper way to install. My loudspeakers have matching drivers and diaphragms, all three (bass, mid and treble) amplifiers are identical models, my electronic x-over has auto-time alignment, all the wiring and connections are good quality.

I understand that squawkers need less and that tweeters even less power, but I already have the amps and the ideal set of amps should have matching sensitivity, gain and frequency response characteristics. I just want to get this right the first time and not have a short cut become a weak link in the system. When I read about bi- or tri-amp dis-satisfaction, I wonder if there was a component mismatch or a calibration problem.

If upgrading a three-way speaker system from single or bi-amp configuration to a tri-amp configuration, is it recommended to add squawker and tweeter protection?

When performing auto-time alignment of drivers, should I place the calibrated measurement microphone at the primary listening position?

Any other installation advice is welcome.

Link to comment
Share on other sites

It's not that difficult to tri-amp. In fact, it is a lot easier than trying to get passive crossover networks finally setup properly for your speakers/room, since you have easy and immediate control over the EQ and crossover filters in real time.

Remember that the Belle has a shorter midrange horn than either the La Scala or Khorn, so the relative delays with the tweeter, the midrange EQ filter settings, and relative channel gains will be a bit different. If you have Khorns or La Scalas, Greg Oshiro's thread on Khorn tri-amping is probably more applicable:

https://community.klipsch.com/index.php?/topic/132275-i-finally-got-around-to-triamping-my-k-horns/

He used a Yamaha SP2060, which is a 2-in, 6-out active crossover. I have one of those and a pair of Khorn clones (i.e., Shinall KKS-1s) that I'm planning to set up properly in a smaller room, so if I get that set up for that room, I'll post the results on a thread here.

Once you get the speaker wires connected (three zip-strip speaker wires per speaker) and your three stereo amplifiers connected to the front left/right stereo speakers and to the active crossover, the rest is actually fairly easy. It's not critical that all the amplifiers are exactly the same - only the left-right stereo channels must be matched in each band (just like using passive crossovers and a single stereo amplifier). You can put your best quality stereo amplifiers on the tweeter and midrange channels, and use a lower quality bass bin amplifier with little to no audible loss of fidelity.

It's nice to have something like a SPL meter to get the PEQs and the relative channel gains set: I used 1/3 octave pink noise tracks with the SPL meter to quickly dial in the PEQs and get the channel gains set initially, and that lasted me for a while until I got some time to do a more detailed calibration.

If you have access to a "calibrated microphone" (such as the lower cost Behringer ECM8000) and a PC or laptop, you can use Room EQ Wizard to do a better job of getting everything dialed in. But I found that the SPL meter and 1/3 octave band tracks will get you very close.

Chris

Edited by Chris A
Link to comment
Share on other sites

  • 2 weeks later...

Curious about the use of the word "timbre" in this thread. I don't see how a timbre match can be obtained since the Belles are still using phenolic diaphragms. Phenolics have their own character, they don't sound anything like titanium or aluminum -- so I'm guessing they wouldn't sound much like Beryllium either. I associate "timbre" with the character or sonic signature of the driver.

See Tom Danley's response on this subject...

"All horns and drivers sound very different BUT if you go to the trouble to make them all have the same frequency response, on axis, they ALL sound vastly more similar to each other."

http://www.diyaudio.com/forums/multi-way/114481-best-horn-compression-drivers-music-3.html#post2959945

Edited by Chris A
Link to comment
Share on other sites

If upgrading a three-way speaker system from single or bi-amp configuration to a tri-amp configuration, is it recommended to add squawker and tweeter protection?

When performing auto-time alignment of drivers, should I place the calibrated measurement microphone at the primary listening position?

It is a good idea to add protection on the tweeters since tweeters don't have a lot of thermal dissipation capability, and also suffer from overdrive damage more easily than woofers and midrange drivers. If you are using high power amplifiers or lower-reliability power amplifiers (e.g., tube type amplifiers), you'd want to add protection. However, I haven't had any problems not using protection for my lower power/high reliability SS amplifiers, and I don't expect any problems in the future with this setup.

I assume that you are talking about auto-alignment using a tri-amped AV receiver or preamp/processor. It doesn't matter much if you are at the listening position or one metre in front of the loudspeaker since most of the auto-alignment functions that I'm aware of are using swept-frequency (chirped) alignment pulses. I'd make sure that you are on-axis with the center between the tweeter and midrange horn mouths wherever you put your microphone, as there can be errors introduced by vertical misalignment of the microphone to the centroid axis between the drivers/horns.

The concern I have is in the resolution of the AVR/AV auto-alignment feature to get within the corresponding time alignment that corresponds to a physical alignment precision of 1/4 inch (6 mm), assuming the midrange-tweeter crossover occurs in the 4-6 KHz band.

My rule of thumb is that the time misalignment between any two drivers (tweeter to midrange, or midrange to woofer) is less than 1/10 of the wavelength at the crossover frequency.

Edited by Chris A
Link to comment
Share on other sites

  • 2 weeks later...

My Klipschorns are now tri-amped with a digital electronic crossover set to Linkwitz-Riley 48 db / octave slopes. Plugged an ECM 8000 microphone into unit, ran automatic time delay set-up, stored settings then removed mic. Using 3 matching stereo, complementary Single Ended Push Pull SS amplifiers with specs of 0.008% THD @ 10 watts. 0.01% THD from 1/4 watt to full rated power. 0.008% Intermodulation distortion at 60 Hz : 7 kHz = 4:1 at rated output. Damping factor 100 @ 1kHz.

Connected and adjusted the system last week, then went out and listened to some live music on Friday, Saturday and Sunday nights. Three different bands, three venues, three different brands of high price pro sound speakers. My seats near center aisle at 45, 20 and 30 feet from the stage. Some people may not care for amplified live music at 100 db, but live performances are fun and I enjoy them on a regular basis, so that is my reference for validating my home stereo sound quality. Back home yesterday and today, I played some vinyl albums, CD's and Blu-ray music discs, quietly and loudly.

My tri-amped speakers sound really good. Individual music instruments are now more tonally distinct and singers voices are very natural and clear. The speakers sound precise, lively and powerful. The soundstage is bigger and deeper, and I want to listen to my entire music collection all over again.

post-58241-0-66520000-1405557351_thumb.j

Link to comment
Share on other sites

Did you mention the type of crossover that you are using? I might have missed that part. Was it a Behringer?

I certainly understand about wanting to listen to the music collection again. This happens to me whenever I do a significant change (for the better) in my active crossover settings, either for the Jubs, TH subs, or Belle center. It's actually a lot of fun to make changes and then listen. Was this exercise enjoyable for you?

Chris

Edited by Chris A
Link to comment
Share on other sites

That is an incredible amount of EQ being used on page 1 in this thread. A ton............WAYYYYYY too much IMHO.

What's easy to notice using digital active speaker processors is that the more EQ and/or processing you use on the unit, the more you hear it. Pretty easy to hear the EQs if you turn them on and off.

I have reduced Roy's K402 EQ settings to just a couple of PEQs and get almost exactly the same response. It sounds MUCH better using less EQ.

Now that I have 510s I do have a couple of PEQs for those, but maybe 2 at the most. I don't even remember.

Whenever I EQ my setup though the goal is always to use the least possible..........to reduce the number of EQ settings as much as possible.

The manual on my Ashly processor even states that the more digial EQ you use, the more you will hear it.

Link to comment
Share on other sites

That is an incredible amount of EQ being used on page 1 in this thread. A ton............WAYYYYYY too much IMHO.

Nice that someone actually looked at the settings... :) It only took 3 years for someone to say something about it (except Dan-Speakmeister).

I haven't optimized the PEQs, ramp filters, or relative gains - only spent time getting the 1/3 octave pink noise bands to EQ flat. I will spend more time with the ramp filters and relative channel gains in order to reduce the PEQs used to a minimum.

I will say that the results aren't in fact bad (as you seem to indicate above). In fact, it's pretty good as compared to the prior active settings or the stock passive balancing network (with new caps). All but one of the 1/3 pink noise octave bands are within +/-1 dBC from 80 Hz-16 KHz. - with one band at -2 dBC.

I am amazed that it worked that well, and it took me about one 20th of the time that it did to get the Jub bass bins/TH subs flat (albeit with a rising "house curve" response down to 17 Hz).

Chris

Edited by Chris A
Link to comment
Share on other sites

Three years ago I had a lot of EQ in my setups too. :)

I figured you EQ'd it flat. That's always my guess when I see so many PEQs.

I was trying to say that the more processing added "the more the life gets sucked out of the music". You probably do have a flat FR.

Link to comment
Share on other sites

Did you mention the type of crossover that you are using? I might have missed that part. Was it a Behringer?

I certainly understand about wanting to listen to the music collection again. This happens to me whenever I do a significant change (for the better) in my active crossover settings, either for the Jubs, TH subs, or Belle center. It's actually a lot of fun to make changes and then listen. Was this exercise enjoyable for you?

Chris

Installing and setting up the digital electronic crossover with adjustable X-O frequencies, selectable filter slopes, driver time alignment and a built in PEQ was an enjoyable exercise for me. Granted there was time and effort involved, plus tri-amps and several pairs of wires, but I already had some of the hardware, and it was fun work. I am using minimal PEQ trim around 200Hz and 4kHz. Yes I have the Behringer DCX-2496, and I know there are better units like the EV-DC-1 on my new wish list, but for now this trial unit was easy to program and it performs very well in my system.

Link to comment
Share on other sites

Updated the Dx38 crossover settings using REW and an ECM8000 calibrated microphone in-room about 2 metres away, on-axis, using the HDMI output from my laptop to drive my preamp, which is much, much better than using my Tascam mixer's analog output.

I also attached a 1/12 octave-smoothed FR plot.

Things sound even better now... :)

Edited by Chris A
Link to comment
Share on other sites

Three years ago I had a lot of EQ in my setups too. :)

I figured you EQ'd it flat. That's always my guess when I see so many PEQs.

I was trying to say that the more processing added "the more the life gets sucked out of the music". You probably do have a flat FR.

In order to put this into perspective, I've included a plot of the center Belle frequency response on-axis vs. the left Jubilee FR. The Belle with the settings found at the top of this thread is the red trace, while the Jubilee plus the TH sub behind the Jub is the green trace. The smoothing used for both traces is 1/48th of an octave as shown in the legend of the graph.

The Jubilee EQ and crossover settings are those that Roy generated in the anechoic chamber in Hope, with no bass boost, and only a couple of changes in the bass bin PEQs to remove response peaks below 250 Hz.

The TH sub settings are my own, and are dialed down from "flat" FR for movies and general purpose, since I find that many older recordings forgot to use infrasonic filters, having a great deal of "rumble" that detracts for the listening experience, and the LFE content of movies is about 10dB too high (which I have dealt with using my preamp settings to reduce LFE as mixed into the sub and other channels by 10 dB: my subs reproduce those LFE frequencies without the compression that direct-radiating subs typically experience). I find that the setting shown here is a good compromise for frequencies below 40 Hz (my crossover point from Jubs to subs).

Edited by Chris A
Link to comment
Share on other sites

Just some notes on this exercise--so I won't forget:

  1. I initially had the center Belle EQed using 1/3 octave pink noise bands from 80 Hz to 16 KHz, using a hand-held Radio Shack SPL meter to measure. The results of this were actually very good, but resulted in too many positive boost PEQ filters due to the methodology that I used going up and down the frequency bands.
  2. I installed REW on my laptop (once again from about four years ago), and connected a Behringer ECM8000 calibrated microphone via a Tascam mixer providing phantom power to the microphone and ADC of the microphone recordings back to the laptop via USB, and used my HDMI port from my laptop to my preamp to provide output driving signal for the REW sweeps. I spent about three hours on Friday working through the PEQs and getting the boost levels down. I intended to not use all the PEQs, but wound up using most of the PEQs available on the bass bin and midrange channels anyway - it sounded better. I stopped EQing for the evening and listened for a while with music Blu-Ray discs, until I determined that I had more work to do on the center Belle EQ - which sounded a bit "tubby". I then adjourned for the day.
  3. On Saturday, I started again with REW and ECM8000 mike to work through the channel gains, crossover points (I moved both center frequencies), dramatically reducing the gain of the midrange channel and increasing the gain of the tweeter channel. This allowed me to use PEQs in "pull down" mode instead of boost gain. There are reasons why I did this, but I'll reserve that rationale discussion for later. Having a bit of luck and playing with the "Q" of the remaining filters a bit more than the gains, I found the happy medium that you see here.
Edited by Chris A
Link to comment
Share on other sites

Observations on the sound (timbre) and sound stage effects for 5.2 operation from re-balancing the center Belle:

There were obvious effects and not-so-obvious effects that resulted from dialing-in the center Belle between two Jubilees.

  • The most obvious was the further elimination of timbre differences between the Jubs and the Belle. The entire front of the room is now one integrated sound image without disruptions in the continuity of the image. In order to hear only the center Belle sound information-it's now necessary to walk up to the Belle and put your ear within two feet (about half a metre) of the loudspeaker's front plane, This was my original aim of updating the center channel EQ.
  • There was a big improvement in the clarity and forwardness of the center channel for movies and surround sound music discs - perhaps a 50-90% improvement--a really big change. I attribute this to being able to turn up the center channel gain because its FR curve now more closely matches the FR of the Jubilees on each side. Before this exercise, Audyssey seemed to move the center channel gain downward from the Jubs, I would guess that it was keying on mid-bass and midrange SPL matching before. Now that the Belle's FR much more closely matches the Jubs, when I do level balance using either pink noise or Audyssey, the center channel is much more obvious.
  • The biggest effect was not immediately obvious: stabilization of the sound-stage image when moving left-right across the entire room. Before this exercise, I actually had two "center images"--one from the phantom center formed only by the Jubilees (which is a big effect if you acoustically treat the front and side walls of your listening room), and one from the center Belle. Now the two effects are locked together and stable if you move from left wall to right wall. It's really incredible and I wouldn't have believed it unless hearing it. I attribute most of this effect to being able to increase the gain of the center channel after re-EQing. This is a big deal since now you can listen to multi-channel movies and music anywhere in the room without perceiving changes in the position of the center of the soundstage image.
  • Unfortunately, I found that the sensitivity to getting the Belle FR to match the Jubilees to be much more sensitive than I would have imagined. I would guess that continuing to adjust the Belle's FR to match that of the Jubs to within 1/2 of a dB is clearly audible. This is bad news for those that want to "plug and play" without spending some time to get the front three loudspeaker array FRs to match.
  • I can still detect a narrower horn dispersion of the Belle's K-500 midrange horn relative to the Jubilees K-402 horns on some music and movies - and it seems to be variable disc-to-disc, with most discs being completely unnoticeable (only the synthetically mixed 5-channel surround sound discs seem to be a problem - with many of these discs sounding too hot on the center channel). Acoustic recordings made with 5-channel microphone arrays don't have this problem.
  • But note that to totally fix the center channel midrange horn dispersion issue I believe that a center K-402 AND a TAD TD-4002 driver will be required to match the horizontal dispersion of the Jubilees. I just don't have the real estate for that in my listening room--which certainly isn't a small listening room by most standards.
  • I can hear badly mixed surround sound mixes quite clearly now. It's so obvious that it draws your attention. I've found that turning down the center channel gain by one to two dB will decrease the perception issue dramatically. I've got a rotary knob gain control on my center-channel Dx38 that allows me to do this quite quickly and easily, and to return it to normal after the disc is played.
Link to comment
Share on other sites

Out1: CHANNEL EQ1,Type: PEQ, f: 180Hz, Q: 6.0, Gain: -8dB Out1: CHANNEL EQ2,Type: PEQ, f: 230Hz, Q: 2.2, Gain: -4dB Out1: CHANNEL EQ3,Type: PEQ, f: 188Hz, Q: 1.0, Gain: -10dB

Had the same "big peak" midbass issue with my LaScalas and Peavey FH-1. But LaScala lovers just call it "fast bass." LOL. I just build a flatter bass horn.

  • Like 1
Link to comment
Share on other sites

I've attached a one-third octave smoothed plot with phase information of the same Belle response as above for grins only, showing you what you typically see from most all manufacturers in order to conceal the roughness of response and phase of their loudspeakers.

post-26262-0-50920000-1405879548_thumb.j

Edited by Chris A
Link to comment
Share on other sites

The only way I can use parametric Eq is to do it in real time with pink noise (ear plugs) and view it on the PC screen as I do it, so I can instantaneously see what's happening. Doing it that way allows me be sure of what i have and use the minimum amount of filters.

Link to comment
Share on other sites

Mark,

The issue that I have with my particular setup is that the Belle must integrate between two Jubs, and that sets the boundaries of what and how I can do it.

Using REW has gotten much more friendly and useful since I last tried it, and it has several added display screen types since I last visited it. The advantage of REW is that it uses up-sweeps that separate room acoustics from "minimum phase" loudspeaker characteristics. It also produces constructed phase, harmonic distortion, step/impulse, waterfall, RT60, decay and group delay plots from the up-sweeps. This is fairly powerful, and I now can see the differences between the Jubs and the Belle much more clearly than using only band-passed 1/3 octave pink noise.

As I indicated earlier, I initially used pink noise bands to get the Belle a lot closer to flat than the stock passive crossover could. You probably would not recognize the Belle output now that it's been time aligned, EQed, channel gains adjusted, and crossover filters chosen. It sounds a lot more like a smaller Jubilee (without the bottom octave) - which was my aim. I didn't wish to change the EQ settings for the Jubilees that Roy generated since I'm very happy with those as is, and I didn't wish to change all three loudspeakers to make them sound alike but all three different than what I started with. The day may come when I do that--when I become dissatisfied with Roy's settings--but that might be a long while.

What I did notice was that the major difference in resulting measurements between the Jubs and the Belle is found in the group delay plot, which I will explore more using different PEQ settings to see their effect on the group delay curves (see below):

post-26262-0-94040000-1405955807_thumb.j

post-26262-0-14880000-1405955824_thumb.j

Edited by Chris A
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...