Jump to content

An attempt to build K-402-MEH clone in wood


StabMe

Recommended Posts

Hi Chris

 

Maybe its the language-barrier!? But I have real trouble wrapping my head around what you are pointing at. But here are my guesses:

 

-Sub-Schroeder-frequency-room-modes, standing waves? Peaks can be attenuated, corrected electronically/DSP?

 

-And then there is the positioning of the speaker in the room? Speakers can be moved around, up/down, back/forward or put in corners?

 

-And then there is room-treatment to avoid early reflections? That would be my main guess!

 

Well I hope you will reveal the answer soon 🙂.

 

Regards from Denmark

 

Steffen

Edited by Supersteff
Link to comment
Share on other sites

It's called various names, but generally it's called boundary gain. 

 

What's different about it?  It's basically the same thing as horn gain, except that care must be taken to carry the acoustic loading all the way back to the throat of the horn in the loudspeaker (like in the case of corner horn bass bins and the K-4020-MEH in quarter or eighth space loading).  Almost everyone seems to think that boundary gain is:

 

  1. something to avoid (the "direct radiating loudspeaker folks") who place their loudspeakers away from the walls, thus robbing themselves of either 6 or 12 dB of free bass response
     
  2. something that is a "fault" of the loudspeaker and have to EQ the bass and midbass response down flat again
     
  3. something that has the same issues as boundary cancellations (~1/4 wavelength bounce) and resonances (~1/2 wavelength bounce) in that many people think boundary gain creates problems that can't be EQed out

 

It's boundary gain that should be taken advantage of when using "room correction software" and when taking in-room acoustic measurements.  How?

  1. First, by placing the loudspeakers next to a corner or wall/floor intersection
     
  2. By placing sufficient midrange absorption material on the adjacent walls to control early midrange reflections (mainly 200-2000 Hz) to regain phantom center stereo imaging and the complete soundstage (it doesn't take very much, but it takes some, and it can be hidden from view using curtains and other acoustically transparent materials
     
  3. By paying attention to how close the loudspeaker (having full-range directivity down to the room's Schroeder frequency-like the K-402-MEH and the other horn-loaded bass bins) is so that the combined horn-loaded gain is not lost outside of the horn before it can be picked up again by the loudspeaker's horn.  (More on this subject to come perhaps later.)
     
  4. By taking acoustic measurements close enough to the loudspeakers to avoid floor and wall bounce (i.e., 1 m) and avoid trying to boost and cut room nulls and resonances related to non-minimum phase reflections.
     

A case in point, I have tried to use Audyssey and Dirac in my listening room, and in each case, it produces unsatisfactory results. The latest such test was using Dirac (Live Full version) after I had dialed in the entire setup by hand using the DSP crossovers.  (The Jubilees are crossed at 40 Hz to the TH subs located just behind each Jubilee bass bin, and therefore is acoustically coupled to the Jubilee output at the 40 Hz crossover frequency.)

 

One might think that Dirac would take out the remaining acoustic issues. It didn't do that.  Instead, it created more issues and unbalanced the EQ of the setup in ways that are not possible to find the reverse EQ curve within Dirac to counteract the issues measuring at the listening positions. I think Dirac uses 7 different microphone positions in a specified order.  Here is what happened when Dirac wasn't supposed to do anything in-room, but did anyway:

 

173102015_LeftJubileeWith(Red)andwithout(Green)Dirac1mMeasurements.jpg.0e1271b896f71ef1e1d7f2c3f8f1c3a3.jpg

 

The rise is response above 6 kHz is explainable (the loudspeaker wasn't pointed directly at the listening positions, which has since been corrected).  Dirac isn't supposed to boost any room modes--but it does.  It's doing this because it can't figure out the difference between a room mode and a true deficit or resonance of minimum phase SPL at 40-70, 210, 375, and 1300-1700 Hz.  This is from trying to take measurements at the listening position instead of 1 m on-axis in front of each loudspeaker. 

 

If you think about the fact that loudspeakers are supposed to have ±2 dB response flatness (via loudspeaker standards--generally assumed to be smoothed data at 1/24th octave or coarser smoothing), this is a disaster when using "room correction software".

 

This is the problem that is easily avoided, mainly by measuring at 1m instead of the listening position.

 

Chris

  • Like 1
  • Thanks 2
Link to comment
Share on other sites

Hi Chris

 

Thank you for answering 🙂

 

It seems you did not finish the sentence in your post?

Quote

3. By paying attention to how close the loudspeaker (having full-range directivity down to the room's Schroeder frequency-like the K-402-MEH and the other horn-loaded bass bins)

 

 

I´m not sure how to understand that sentence? Are you referring to 1/4 wave spacing of the Full-range MEH and a Sub-horn?

 

Steffen

Link to comment
Share on other sites

I have edited my response, above (as you were typing your response).  See if what I wrote makes sense now.

 

Chris

Link to comment
Share on other sites

Regarding the subject of "not losing too much horn gain", there is a famous article that has largely been forgotten over the years, written by Harry Olson, and titled A Horn Consisting of Manifold Exponential Sections. A copy of that article can be found here: https://community.klipsch.com/applications/core/interface/file/attachment.php?id=33561

 

Here is figure 5 from that article, annotated by another author at the top of the figure:

 

Rubber throat.gif

 

Why is this important?  Think about the idea of a full-range (hi-fi) MEH--this is how that occurs, and it also wrapped up with the measurements taken at 1 m instead of at the listening position (which is generally 3-5 m, and always contaminated with room reflections). 

 

So the idea here is that the horn is placed in-room such that the room's boundaries pick up the horn gain before it is lost (i.e., "S2" is close enough to couple acoustically to "S3" in the figure above before the real part of the acoustical impedance--otherwise called "horn loading"--is lost).

 

In the K-402-MEH, if the room dimensions and placement of the loudspeaker mouth is placed advantageously, response down to 17 Hz is easily achieved--and in half/quarter/eighth space (i.e., back against a wall, coupling to the floor, ceiling and front wall at some frequency).  This is how that occurs. 

 

Just about everyone that I've read in since PWK wrote of this has "lost the bubble" on this subject.  In my experience, this isn't something that you should lose track of.  It should be one of the first things that hi-fi enthusiasts learn in the quest for better low frequency response.

 

Chris

  • Like 1
  • Thanks 1
Link to comment
Share on other sites

Hi again Chris

 

Have you tried Acourate?

 

https://www.audiovero.de/en/acourate.php

 

I have read a lot of good things about that software, its world-famous in Germany! Mitch Barnett has written a book about it, using the fir-filters generated by Acourate in JRiver. I think there is more to read on the computer-audiophile forum.

 

Following your recommendations, it seems to make no sense to use room-correction other then correcting for room-modes, if the system is capable to detect them!?

 

Steffen

 

PS: Im also earning my right to PM again!!!!

Edited by Supersteff
Link to comment
Share on other sites

I've read about it at length.  It's so complicated (from Mitch's book) that I don't recommend it.  Instead, I'd recommend DSP crossovers and perhaps convolvers on PC player applications (using something like rePhase to determine the FIR filter coefficients) to do the job.  It's easier, cheaper (much cheaper), and much more effective, since it can be controlled.  Trying to automate this is, I find, plagued by pitfalls that I still haven't seen solved.  Even Acourate...

 

Chris

Link to comment
Share on other sites

@ Chris

 

I have read through your edited response, and the sentence:

 

Quote

3.By paying attention to how close the loudspeaker (having full-range directivity down to the room's Schroeder frequency-like the K-402-MEH and the other horn-loaded bass bins)

 

It still doesn´t make sense to me. Paying attention to what?

 

Steffen  

Link to comment
Share on other sites

On 3/27/2021 at 10:08 AM, Chris A said:

3. By paying attention to how close the loudspeaker (having full-range directivity down to the room's Schroeder frequency-like the K-402-MEH and the other horn-loaded bass bins) is so that the combined horn-loaded gain is not lost outside of the horn before it can be picked up again by the loudspeaker's horn.  (More on this subject to come perhaps later.)

Should read "by paying attention to how close the loudspeaker (having full-range directivity down to the room's Schroeder frequency-like the K-402-MEH and the other horn-loaded bass bins) is to the room boundaries..."

 

In other words, you need to keep the loudspeaker's mouth close enough to the room's boundaries such that when the first 1/4 wavelengths of acoustic energy reaches past the mouth of the loudspeaker horn itself (the MEH horn, or the bass bin horn mouths--i.e., dual "W" section bass bin mouths), that the room's boundaries pick up the support of that first 1/4 wavelength to continue the support of the real part of the acoustic impedance at the horn throat (at the woofers) still couple acoustically to the driver/horn. 

 

For "W" section bass bins, such as the Jubilee and Khorn bass bins, that frequency--corresponding to 1/4 wavelength of the internal path length of the horn, is generally about 40 Hz.  For the La Scala and Belle--it's a bit over 100 Hz.  What is forgotten is that the exponential horns they use have a cutoff frequency that curtails output below some frequency--in a manner similar in behavior to a bass reflex box.  Below a certain frequency, their exponential bass bin horns can't produce impedance at the throat of the horn to the driver--like the impedance plot shown below:

 

Exp horn impedance vs freq.PNG

___________________________________________________________________

 

In the case of the MEH, it uses a straight-sided horn whose real part of its acoustic impedance back to the driver doesn't go to zero below the axial 1/4 wavelength that most people think is the "cutoff frequency".  Instead, the horn continues to provide real impedance back to the driver--because the mouth dimensions of the horn contains the exiting wavelengths directivity.  This can been seen in the impedance plot below for a straight-sided horn:

 

Conical horn impedance vs freq.PNG

 

The real part of the acoustic impedance at the horn's throat doesn't go to zero below the same-size exponential horn's "cutoff frequency", so the mouth-located driver still feels the acoustic loading of the horn (in the same exact fashion as a manifold exponential horn that I discussed above), where the horn loading is still much higher than the loading that the drivers would see in direct radiating mode. 

 

This is key, i.e., half the loading of the straight-sided horn is only a -3 dB power loss--which isn't that much in a world dominated by DSP crossovers and powerful amplifiers--without incurring most of the effects of modulation distortion of direct radiating mode woofers, and low frequency phase growth of direct radiating bass bins--especially bass reflex types. 

 

If horns typically provide 15-20 dB of SPL gain over using the same drivers in direct radiating mode, then the -6 dB loss of using a straight-sided horn instead of exponential, and using boundary gain below 1/4 wavelength axially (i.e., the frequency associated with that wavelength), you still have 9-14 dB of horn loaded gain using a straight-sided horn in boundary gain, than using the same woofers in direct radiating mode.  That's a lot of advantage--a lot more than 3 dB of bass reflex gain, without the accompanying problems of modulation distortion or phase/group delay growth of the bass reflex box.  And if you place the MEHs in the room corners, you recover almost all of that on-axis SPL loss below the "1/4-wavelength axially" frequency.  So if you choose to use MEHs in a corner configuration, they behave just like corner horns--without needing "W" section bass bins or horn folding.

 

The full-range MEH will produce very low frequencies in boundary loading (if the boundaries aren't too far away).  You're using that bit of impedance left over below the frequency corresponding to 1/4 wavelength axially in the horn to load the drivers acoustically, and it will support the handover of the output acoustically to the room's boundaries--much better than using direct radiating mode.  It can do this because of the size of the MEH mouth--which is even larger than corner horn "W" bass bins in terms of controlling loss of directivity.  The K-402-MEH doesn't lose directivity in the horizontal direction until ~40 Hz.  At this point, the handover of the low frequency waves to the room's boundaries (...do you remember I mentioned the importance of boundary loading and to distinguish it from non-minimum phase reflections?...) permit the woofers in the K-402-MEH to take the bass extension down to 17 Hz with less than 10 dB of overall EQ flattening--just like the K-402 experiences with 2" compression drivers on the high frequency end.

 

Chris

  • Like 2
  • Thanks 1
Link to comment
Share on other sites

So to continue the thinking about MEH and bass extension that I broached above:  the K-402-MEH-sized horn can get almost three octaves of bass response extension below that which would be expected using exponential horns having the "1/4 wavelength internally" rule, but the MEH only does this in home-sized listening rooms that are not too big (i.e., the MEH can couple acoustically to the walls, floor and ceiling of the home listening room if they are not too far away from the horn mouth at 40-170 Hz and below), and not too small (such that the largest dimension of the room is below the MEH's roll-off frequency after EQ)

 

My posting of performance of the MEH down to 17 Hz is not a mirage: it's real, and it can act as its own subwoofer (subject to the understanding that FM distortion due to its increased bandwidth is a tradeoff factor).  By placing the K-402-MEH in the center location in a room sized for deep bass extension, I avoid having to use a subwoofer underneath it.  A subwoofer can be used with it--no problem.  But it's nice to know that it isn't required if the MEH is in 1/4-to-1/8th loading in room at the loss of horizontal directivity of the horn's mouth. 

 

If you had five of the K-402-MEHs in surround sound (5.1, etc.) around the listening position, my guess is that you probably wouldn't need subwoofers, if the room isn't too large (two dimensions--like width and height,  or width and length, etc.).  You would have a surround array of subwoofer-capable loudspeakers that could be used instead.

 

YMMV.

 

Chris

  • Like 1
Link to comment
Share on other sites

Thank you Chris for your repeated explanation and patience! Very much appreciated. 

 

I will read my way through that OLSON article, a lot to digest. In a way its like starting from scratch. A lot of thinking has to be turned upside down to get right. But I like that your responses are based on science. I guess things get even clearer, when I get started to build a big MEH and make my own experiences and measurements.

 

By the way, I have noticed that you have tested the new Celestion AXI2050 driver and recommend it with boost in the top-octave. Interesting to see what implications that can have for the placement of the off-axis-ports of the woofers, i.e. lowering the crossover-frequency. We will see.......

 

Steffen

Link to comment
Share on other sites

  • 2 weeks later...

Hey!

 

The project is moving, albeit somewhat slowly. The guy at the carpentry is now gluing the parts together. Brought loudspeakers to him yesterday so that he can he if they fit nicely and do amendments if necessary.

 

Some pics:

 

IMG_20210415_095049.jpg.e927250777a00edea4420b97d0b8ee6a.jpg

 

IMG_20210415_094911.jpg.64c1b8dd0b75e786519e5226ba71bf39.jpg

 

IMG_20210405_095655.jpg.029e7250a4dc246757d518be6c32ecaa.jpg

 

 

  • Like 8
  • Thanks 1
Link to comment
Share on other sites

I decided to share 3D files i used with the community. If this helps the community as a whole or any person here individually, i am very glad that i can pay back just this little.

Files are available at this link. Most of the info (even file names) are in Russian. But the text part of the instructions are somewhat redundant and nobody really going to have problems understanding which parts goes where. And even then there is Google Translate which can handle the text, i believe. Oh, and I am here and can help with translation as well. 

  • Like 6
  • Thanks 1
Link to comment
Share on other sites

BTW, can anyone comment on an idea of using JLH amp for HF? There a few versions of this 10W per channel amp available at Chinese stores that can be bought relatively cheap and have good reviews. For the LF i will be buying something separate. Shooting for about 100W per channel. Or will i need more power to be able to boost some lows as well?

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...