Jump to content

I finally got around to triamping my K-horns


Greg Oshiro

Recommended Posts

No cigar. See the attachment. Note the 544 Hz suck out and the excess phase with the extra 2.4 mSec delay. I stand by my original settings. Sort of. I just tweaked the woofers for smoother response and better left-to-right matching (measurements anyway....).

I just saw this. How did you measure phase? Upsweep/downsweep chirps, impulse...?

Chris

EDIT: Actually, if the Khorn 6 ms tweeter-bass bin time misalignment is correct, then it would explain discrepancies I found between horn-length calculations and Heyser's article, which I've wanted to check out myself (...not enough Round Tuits...).

If the Heyser measurement errors also apply to the Khorn midrange-tweeter delay (Greg's 1.6 ms vs. Heyser's 1.8 ms) it would also clear up some issues in horn-length calculations that I independently had on midrange/tweeter time misalignment that I dealt with when I was setting up the tri-amped Belle.

Link to comment
Share on other sites

<snip>I just saw this. How did you measure phase? Upsweep/downsweep chirps, impulse...?

Chris

<snip>

TEF uses Time Delay Spectrometry with a linearly swept sine wave. REW (I think) can do the equivalent. I downloaded the REW Rev 5 help pdf and skimmed through it. See the "Systems and Transfer Functions" section starting on page 13 and continue through "Windowing the Impulse Response". TEF does the equivalent of an FFT of the time-windowed impulse response. However, the window that TEF uses may not be the same as REW. The window shown in the REW manual doesn't look like the window that TEF uses. My measurements were taken with a ~20 mSec window. There are a number of different transfer function measurement methods, each with its' own advantages and disadvantages.

What I attempted to do with my measurements is remove room reflections from the measurement to obtain the anechoic transfer function. In my world of large venue work, it's known that starting with a"flat" anechoic transfer function works. I want to see how that works in a domestic environment. However, I can tell that there are reflections corrupting the "anechoic" data below about 400 Hz, hence my lack of confidence in the woofer/squawker crossover filters and levels. I figure if I can get a reasonably flat transfer function at 3 feet on axis of the squawker/tweeter axis, I'm off to a good start. I'm happy with the squawker/tweeter crossover, given the raw speaker response, but I don't trust the woofer/squawker raw speaker data in the crossover region.

Other news: I bypassed the woofer high-Q PEQ boosts. It sounds better. Usual caveats about judging the quality of my own work...

<snip> horn-length calculations<snip>

The transfer function phase includes both the driver/horn phase response and the phase delay due to the length. At crossover, the phase response of the adjacent range also comes into play. Knowledge of the phase delay due to path-length difference alone is not sufficient to determine the optimum crossover filters.

Link to comment
Share on other sites

This looks exactly like what my intentions are for the LaScalas I recently aquired, if/when I get them refinished sufficiently that my wife lets them in the house. http://forums.klipsch.com/forums/t/155314.aspx

The transfer function phase includes both the driver/horn phase response and the phase delay due to the length. At crossover, the phase response of the adjacent range also comes into play. Knowledge of the phase delay due to path-length difference alone is not sufficient to determine the optimum crossover filters.

Exactly. Crossovers take time, whether analog, digital, or acoustical. The time offset between raw drivers is not the same as the excess delay that needs to be added to any particular crossover alignment. People like Ed Long were aligning Altec 604's by juggling the passive crossover long before anything digital was viable.

Link to comment
Share on other sites

  • 2 years later...

This one is Greg's - I just watched and encouraged him to type in the results...

I haven't talked to Roy about active settings for a Khorn - but if he's got anechoic results to share... :)

Edited by Chris A
  • Like 1
Link to comment
Share on other sites

  • 1 year later...

Opening this old thread because I'm just starting to venture into triamping my 2 channel set up. I've got the Yamaha SP- 2060 but only one amp so far. I have Khorns and La Scalas with various horn and driver modifications and ALK ES crossovers. Triamping seems to be the next step into this insanity(hobby). But I have a question that is causing me confusion. So Heyser measured the bass horn of the KHorn to be 8.4 mS behind (delayed) the tweeter? So to remedy the tweeter is delayed 6 mS more by Greg? And Chris suggests trying the full 8.4 mS. Am I thinking backwards?

 And while I'm here a BIG THANK YOU to Greg for "publishing" his settings for the KHorn.

Link to comment
Share on other sites

No, no... Greg's delays are the correct ones - 6 ms for the tweeter channel and 4.4 ms for the midrange channel.  I asked a question and Greg did the tests to show me that Greg got it right and Heyser didn't.

 

The relative gains are also important (-10 dB for the midrange channel, and -9 dB for the tweeter).  That's how much power is being attenuated by the passives into the midrange and tweeter channels, either via reactance  increase or pure resistance.  Either way, what you're going to find out is that the effective output power of your loudspeakers/amplifiers is going to be significantly increased.

 

Once you get everything set up, let me know, then we can set up the limiter to match your amplifier and your K-77s to make sure that you can never blow the diaphragms, like Cantilope finally accomplished with his active setup, thus eliminating that problem forever. 

 

Chris

Edited by Chris A
Link to comment
Share on other sites

Okay, that's good. 

 

The K-77 tweeter, while good for purpose on Klipsch Heritage, can't take more than 5W music power--with some models limited to 3W.  That's why you see a lot of people buying replacement diaphragms for them--the passive crossovers simply cannot protect them from all transients and overloads, while the digital crossover can protect them--and without side effects.  You can also use the limiter function on your SP2060 if you absolutely want to protect your drivers from "devious fingers".  ;)

 

Chris

Link to comment
Share on other sites

I was very surprised to see recent activity on this thread.

 

babadono, please keep in mind that my settings were based on measuring the frequency response and tweaking the crossover settings. Your settings will be different because of individual driver variation and the B&C drivers instead of stock Klipsch. I used a very old-fashioned system called TEF-25. There is free software called Room Eq Wizard (REW) that IIRC Chris A is familiar with that  appears to be able to measure what needs to be measured. The difficult part is interpreting the measurements and deciding what crossover settings to use.

 

Strangely enough, I spent the day in the woodshop making temporary baffles to try some BMS 4540 drivers on Electro-Voice HPT60 horns in place of my K77's.

 

--Greg

  • Like 1
Link to comment
Share on other sites

Hi Greg!  How are things going?

 

I think that your active crossover settings are probably better than you might credit them--at least as a really starting point.  Dialing in the settings after getting close is fairly straightforward--one that can be done much more easily than starting from scratch for those not accustomed to using the PEQs and shelf filters., as well as relative channel gains and delays.

 

Chris

Link to comment
Share on other sites

  • 3 years later...
On 12/11/2011 at 7:40 AM, Greg Oshiro said:

I've made a few changes:

  1. Original setup was CD player analog output to DSP box analog input. It's now CD player SPDIF to cheap HOSA SPDIF-to-AES3 converter to DSP box AES3 input. It's louder. Post-DSP-box attenuators changed from 18 dB to 30 dB attenuation. It sounds *much* better. I think it's because the CD player D/A converter and DSP A/D input converter are no longer in the chain. Also, the Yamaha DSP box analog input can take something like 20 volts before it clips. I think the CD player puts out 2 volts at 0dBFS. I think the new connection uses more bits in the DSP box.
  2. The low frequencies around 63 Hz pile up when listening in stereo. there is now a gentle (-2 dB, 63 Hz, Q=2) cut in the overall EQ of both channels that sounds better to my less-than-golden ears.

Other impressions:

  1. Walking around the room, I can plainly hear the polar pattern of the tweeters. In the old passively crossed-over days, this was not so obvious.
  2. The passive crossover had a very"splashy" high-hat sound. It now sounds much more like the real thing.
  3. Reverb tail decay can now be heard in much greater detail. I use Steely Dan's "Babylon Sisters" track for this evaluation.
  4. Drum attack, especially toms, sound better.
  5. The "air" around saxophones is clearer.
  6. The difference between recordings with/without low-frequency extension is much clearer.
  7. It is easier to listen to individual instruments in a mix.
  8. Pianos sound much more realistic.
  9. I'm listening to Lyle Lovett's "Live in Texas" right now. I can hear him inhaling between phrases.

Over all, I'd say this doesn't suck. ;)

What differences between passive and tri-amped setups have you folks heard?

 

I've spent the last few days dusting off the Khorn clones (Gary Shinall KKSs), getting rid of the accumulated spider webs and various ornaments and accessories that have adorned the top of them for the past decade or so while they sat in the computer room unused.  I also cobbled together a laptop-based REW measurement setup with an old Windows 7 laptop and digital-to-analog converters like Greg mentioned above.

 

KKS small.jpg

 

I did a REW measurement on the mono-amped passive crossover version of these beautiful loudspeakers before fully tri-amping using the Yamaha SP2060 out, which has been in storage, and gathered together three stereo amplifiers also in storage: two Crown D75As and an ICEpower 50ASX2 module amplifier.  I'm still awaiting delivery of a USB--AES/EBU DAC and digital signal converter from China to connect with the digital input port on the Yamaha SP2060. In the mean time, I'm using an HDMI--RCA analog converter to connect the laptop to the Yamaha SP2060, and it seems to be performing well. 

 

The results are at least what you read about above from Greg Oshiro's experiences in 2011--just over 8 years ago.  I would say that it has increased the lightness of sound in higher frequencies fairly dramatically, and gave me a much more tonally balanced presentation as I have dialed-in the EQ to flatten its SPL response, including bass and lower midrange that were lacking in the passive version.  The sound now grabs you and makes you want to sit and listen to whole albums instead of just partial music tracks that I experienced with the mono-amped passive version.  The sound of string orchestras is now compelling.  I'm using first order crossover filters, which introduce only 90 degrees of phase shift and which are contributing to a very flat phase/group delay curve--all but the transition from the midrange to bass bin mouths. 

 

When I get a chance, I'll transfer some of the REW measurement plots here for comparisons with the passive version. 

 

Chris

  • Like 1
Link to comment
Share on other sites

Some plots from the above activity.  Some observations:

  • Passive crossover version is in red trace, while active/tri-amped crossover is in yellow trace. 
  • The bump in the bass bin--midrange phase and group delay can be seen better in the spectrogram plot, and is a function of the loudspeaker's design. 
  • The passive crossover's phase growth is so great that it cannot be fully plotted within REW (i.e., more than 4000 degrees of phase growth from high frequency to low frequency). 
  • The group delay cleanup at higher frequencies using active tri-amping is probably responsible for the greatest part of the subjective listening improvement.  Note the vertical axis of the group delay scale--this plot is zoomed out in order to see most of the departures from zero group delay (measured in milliseconds).  The passive version exhibits very large local departures from flat group delay above 100 Hz, while the active version responds to the jump from the midrange horn to the dual-mouth bass bins.  The tri-amped version uses first order crossover filters (just like the passive version) but crosses at 250 Hz rather than 400 Hz in the passive crossover version.
  • The jump in the phase curve for the active tri-amped version is actually inaccurate--the actual phase growth can be seen more clearly in the spectrogram peak energy curves.
  • The step response and the spectrogram plots both show the jump in time of arrivals at the crossover frequencies for the passive crossover, and the alignment of the bass bin-to-midrange and midrange-to-tweeter crossover frequency bands (i.e., 400 Hz and 5 kHz for the passive version, and 250 Hz and 5 kHz for the active tri-amped version)

 

1615590539_ShinallKKSPassivevsActiveSPLResponse.thumb.jpg.e5386401dd80442639c31e60f86ecb54.jpg

2133480465_ShinallKKSPassivevs.ActivePhaseResponse.thumb.jpg.d4ba82ab6832c458de3e0b40c63124fc.jpg

1918865973_ShinallKKSPassivevs.ActiveGroupDelayResponse.thumb.jpg.2ab0ae54bb26d75e7b35ff4313ac02ce.jpg

749780871_ShinallKKSPassivevs.ActiveStepResponse.thumb.jpg.b9bd3e76731b260861c7eaecc6c59c05.jpg

692228481_ShinallKKSPassiveXoversSpectrogram.thumb.jpg.4bffec7e8c73d921790902dd3b128195.jpg

1097105934_ShinallKKSTri-AmpedSpectrogram.thumb.jpg.1eb08d8411a0b6f0438cd9d4634e619f.jpg

 

The results from subjective listening are (yet again) best judged from the standpoint of the listener's subconscious willingness to listen to full albums being played with the actively crossover version vs. only parts of music tracks in the passive case.

 

Chris

  • Like 1
Link to comment
Share on other sites

I'm taking a little time early this morning as the sun rises to listen to some favorite recordings (notably most of Vivaldi's concertos).  I'm moved by the sound that I'm listening to--to write a little on the sound quality that I'm hearing.  Note that the phase response of these loudspeakers is now within 45 degrees of flat phase from below 100 Hz to 15+ kHz with a notable rise and fall in phase growth back to nominal a little above where the crossover occurs from the bass bin to midrange (355 Hz).

 

As I listen, I'm thinking back on my impressions of the K-77 tweeter harshness and the horn-sound midrange issues that I remember from years passed with this and other Klipsch Heritage loudspeakers (Belles, Cornwalls, Heresies, and to a lesser extent, La Scalas owned by friends and family), I can now authoritatively say that most of those harshness and horn-sound issues are curtailed simply by time aligning the drivers using a DSP crossover (tri-amping) and by flattening the SPL response to ±2 dB or less.  Had I known that there would be this much difference in sound quality, I would be taken the time many years ago to make the changeover.  The satisfaction level in listening to these loudspeakers is like nothing that I've heard from other Klipsch Heritage loudspeakers.  The enjoyment level now pulls me into the room to have a listen whenever I can. 

 

Highly recommended.

 

Chris

Link to comment
Share on other sites

It is interesting that John Atkinson of Stereophile ended his written up measurements of the Klipschorn by wondering what a fully DSP-corrected, tri-amplified version of the Klipschorn would sound like:  https://www.stereophile.com/content/klipsch-klipschorn-ak6-loudspeaker-measurements

 

Apparently it can sound pretty good.

 

Mark

 

 

Edited by Marks
  • Like 1
Link to comment
Share on other sites

1 hour ago, Marks said:

"I can't help wondering what a fully DSP-corrected, tri-amplified version of this loudspeaker, with the high sensitivity coupled with optimized step and frequency responses, would sound like."John Atkinson

 

Apparently it can sound pretty good.

 

Yes, it does.  Welcome (assuming you're new to the forum).

 

Note again that I'm listening to Khorn clones (for the record)--but the sound quality improvement is substantial and mostly related to subconscious improvement in listening performance.  The highs sound much cleaner and the characteristic EV T-35 tweeter (i.e., the Klipsch K-77 tweeter) harshness is gone, and bass sounds much more clear, effortless, and balanced  (although I did equalize the SPL response overall which brought up the bass response by almost 5 dB above 60 Hz--the comparison SPL response plot posted again below):

 

Shinall KKS Passive vs Active SPL Response.jpg

 

Chris

Link to comment
Share on other sites

1 hour ago, Khornukopia said:

I have read somewhere in the past that the build quality of the Shinall cabinets were very good, and that he had some sort of agreement with Klipsch.

Yes...https://community.klipsch.com/index.php?/topic/40203-klipsch-knockoffs-shinallhave-you-seen-these/&do=findComment&comment=367500

 

Gary died of ALS in July 2008.  When I bought the pair of KKSs from him in the fall of 2007, he enlisted the aid of his wife to translate for him because his speech was heavily affected by his disorder at that time.  Even then, he was fiercely protective of his loudspeakers as I loaded them into a full-sized van I rented from Enterprise with a load of furniture pads to drive back north 200 miles home.  He made sure they were going to get home without a ding or a nick, and they made the trip without incident.  I still look at these loudspeakers as a legacy to the man that made them one at a time by hand with pride in his workmanship.  It shows.

 

Chris

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...