Jump to content

Why the 2.5mH value?


Klipschguy

Recommended Posts

Why did PWK choose a 2.5mH inductor for the Cornwall’s 1st order, 600Hz crossover point?  A 2.5mH choke used with a 4 ohm woofer (like the K33, Ze=3.6 ohm) corresponds to a crossover point of about 250Hz.
 

My understanding is the Cornwall has a hole in its frequency response from about 400 to 600Hz. Intuitively, I would think the woofer would be crossed over higher to compensate for the sagging response. I guess the hole in the frequency response could be caused by an internal standing wave where a different choke might not do much, but I am scratching my head a little bit over this one.

Link to comment
Share on other sites

On 12/28/2020 at 7:17 AM, Klipschguy said:

Why does the Cornwall use a 2.5mH inductor in its 1st order networks for a 600hz crossover point?  A 2.5mH choke used with a 4 ohm woofer (like the K33) corresponds to a crossover point of about 250Hz.
 

My understanding is the Cornwall has a hole in its frequency response from about 400 to 600Hz. Intuitively, I would think the woofer would be crossed over higher to compensate for the sagging response. I guess the hole in the frequency response could be caused by an internal standing wave where a different choke might not do much, but I am scratching my head a little bit over this one.

 

I would think that the impedance of a K33 in a Cornwall would be closer to 6 ohms (?).  A 2.5 mH inductor is 6 ohms around 380 Hz.  I assume you are referring to the Type B crossover?  The 20uF capacitor across the woofer (Type B2) will extend the -3dB point to around 500 Hz, but it drops off twice as fast after that.  If you have a B, try the capacitor.  You may have to reverse the squawker polarity.  Also be aware that the addition of the capacitor will boost the upper bass, but you may find that pleasing to 

your ears.

 

Mike

  • Like 2
Link to comment
Share on other sites

Here's a test report done in 1999 on Corn IIs (like the ones I have). The "dip" you are referring to is due to floor bounce. It shows up in measurements but room acoustics and our own ears average it out. Corn IIIs and IVs moved the woofer up the panel, so, having a longer path to the floor, that bounce will be moved down in frequency. That is, providing that Roy did not fix it by crossover design.

Narrow dips measured in speaker response normally are inaudible, since, as I said, the room acoustics average things out,. and our own ears tend to "fill in" these dips. A peak is much harder to ignore.

I would not alter the crossover without having measuring gear to see first, what problems you are trying to solve and secondly, if you solved them or not. The response problem I had with my Corns was that they tend to boom in corners. Pulling them 18 inches out from the corners fixed that.

Klipsch_Cornwall_Test_Report 2.pdf

  • Like 2
Link to comment
Share on other sites

Yes, I am talking about the B networks. To my ear, the B network gets the midrange and treble balance right, but the upper bass is not quite correct  (said with all due respect for Paul Klipsch).  I tried a B2 network; it seems to have better balance in the upper bass, but the midrange and particularly the highs do not seem properly  balanced (the B2 is too bright to my ear - and yes I listened for several months before making up my mind).  After the extended listening period, I put my original B networks back in, which sound perceptibly better, except for the upper bass (in my humble opinion).
Andy 

Link to comment
Share on other sites

On 12/28/2020 at 9:17 AM, Klipschguy said:

Why does the Cornwall use a 2.5mH inductor in its 1st order networks for a 600hz crossover point?  A 2.5mH choke used with a 4 ohm woofer (like the K33) corresponds to a crossover point of about 250Hz.
 

My understanding is the Cornwall has a hole in its frequency response from about 400 to 600Hz. Intuitively, I would think the woofer would be crossed over higher to compensate for the sagging response. I guess the hole in the frequency response could be caused by an internal standing wave where a different choke might not do much, but I am scratching my head a little bit over this one.

 

The inductor must make the speaker's response flat.  Look at the attached response curve that is typical of many woofers.  Setting the *calculated* (electrical) crossover lower will make the actual acoustic crossover higher and where you want it.  There is both art and science in speaker design and the crossover is not a separate part. 

 

Since when did a Cornwall have a "hole" in its response? 

Screenshot_20201231-012846.png

Link to comment
Share on other sites

No "holes" but a documented dip, which as I said before is an artifact of woofer to floor relationship and present in most large loudspeakers. Unless it is consistently audible with variety of music, I would disregard it.

The report attached above says:

 

"Frequency response is smooth from 25 Hz up to 20 kHz. A "deep" at 350 Hz is due to floor reflection (microphone was 1 m from front panel..."

Link to comment
Share on other sites

Here is a 1979 Cornwall in my listening room, with the only modification being the tweeter was switched from the K-77 to the CT-125 (before I bought it):

 

1859452629_Uncorrected79Cornwall(QuarterSpace)SPLandPhase.jpg.c0b5df7fc965f7a5cc3742848c709b44.jpg

 

I think the concern about passive crossover filter component values, etc. is really based on what I'd call archaic notions of "what the passive crossover values should be" based on assumptions of driver/horn electrical impedance.  The problem is, the actual input electrical impedances of the drivers with their horns attached isn't anything like "flat" or single-valued. 

 

I do recommend spending $94 on a UMIK-1 and a few bucks on a microphone stand, downloading REW, and measuring the most important part of the loudspeaker's characteristic: its acoustic output.

 

Above you will see that the real issue is not really SPL response, but rather phase response (the bottom magenta-colored trace). If you want to do something that will increase the sound quality (in this case, soundstage imaging and clarity), release the tweeters from their in-cabinet positions and move them on top of the cabinet, and backwards ~9.75 inches (for the CT-125 tweeter) until they are time-aligned with the midrange horn/driver.  That will improve the sound quality audibly.  Here is a spectrogram showing the degree of time misalignment of the tweeter to midrange:

 

1692679030_Uncorrected79Cornwall(QuarterSpace)Spectrogram.jpg.e712f8f229590ff070521cb38e384f68.jpg

 

That really sharp dogleg in the peak energy time curve at ~475 Hz is really the time misalignment of the woofer to the midrange (or alternatively, a polarity issue at the crossover between the midrange and woofer)--which is actually quite small, and probably inaudible.  What you should pay attention to is the .75 ms delay in the midrange to the tweeter above 5 kHz.

 

Chris

  • Like 1
Link to comment
Share on other sites

 

On 12/31/2020 at 12:49 AM, JohnA said:

 

The inductor must make the speaker's response flat.  Look at the attached response curve that is typical of many woofers.  Setting the *calculated* (electrical) crossover lower will make the actual acoustic crossover higher and where you want it.  There is both art and science in speaker design and the crossover is not a separate part. 

 

Since when did a Cornwall have a "hole" in its response? 

Screenshot_20201231-012846.png

 

Thank you for your response. There are a few reasons I think there is a hole in the frequency response of the Cornwall 1.  
 

1. A Dope from Hope paper discusses the design of the B2 network (as an update to the B network) addressing the frequency response hole in the crossover region (400 - 600Hz) when using the B network. There  is audible improvement in the crossover woofer/mid crossover region, but the B2 creates other compromises that make the speaker sound worse to my ears.

DFHB2.pdf

 

 

2. Here is a link to an anechoic response curve run by Trey Cannon on a Cornwall 1. The dip is pretty clear. They credit a standing wave as the culprit. My understanding is they tried different solutions, but ultimately moved the woofer up 6” to address the standing wave and hence the hole in the 400 - 600Hz region. (Interestingly, it appears they put a baffle near the top of the Cornwall IV cabinet with significant stuffing above it to address what I would guess to be more standing wave suppression, but that is a subject for a different thread...). Sorry about the link; I wish I had a better source for the information.

https://www.hometheatershack.com/threads/review-klipsch-cornwall-iii-vs-cornwall-i.744/

 

3. The manifestation of the Cornwall’s standing wave in the 400 - 600Hz is audible. I can hear the improvement in the upper bass/lower mods with the B2 network, but the balance in the upper frequencies is, to my ears, inferior to the B network.


4. The frequency dip and standing wave issue is pretty well known to Klipsch as they have made a number of attempts to correct it over the years. My guess is the Cornwall IV has the issue well figured out. Again, my desire is to keep my Cornwall 1s, which I have been quite happy with for decades except for this issue. 
 

Thank you again for your replies and a happy New Year to all!

Link to comment
Share on other sites

1 hour ago, Klipschguy said:

The manifestation of the Cornwall’s standing wave in the 400 - 600Hz is audible.

If it's a standing wave (i.e., inside the cabinet), then a crossover network isn't going to do much at all--perhaps delay the woofer even more (which is really not what is needed)--but not decrease the time delay of the woofer to the midrange (i.e., all-pass delays go up, but not down, if you increase the order of the filters).

 

I think you're looking at the time misalignment of the woofer to the midrange--including the first order crossover filter delay of 90 degrees--that is the real culprit.  This produces lobing in front and to the sides of the front baffle--not internal standing waves. (You can also experiment with adding a small amount of extra damping material in the cabinet in the upper portion of the cabinet, too. Putting absorption around the rear of the woofer would probably change its "Q" tuning fairly substantially.

 

Moving the woofer upwards shifts the internal standing waves, but that didn't happen until well after, when the "Cornwall III" came out (IIRC). With that change also came the change from the old Cornwall midrange horn to some variation of the Heresy midrange horn (again, IIRC).  So the CW III isn't the same loudspeaker. The Cornwall IV is yet another loudspeaker having different drivers and horns.

 

Instead of worrying about what someone said that happened inside an anechoic chamber 45 25 years ago, I'd strongly recommend measuring it, then it will put your mind at rest on the perceived issues, and substitute real, measurable issues that can be addressed to provide the bottom line payoff: better sound quality.

  • Like 2
Link to comment
Share on other sites

Thank you for your replies, Chris. Yes I hear you regarding the phase shift issue, particularly the 90° shift caused by the first order network inductor hooked to the woofer. Interestingly, the very early Khorn and Cornwall crossovers used a .5mH inductor in the midrange leg of the crossover which would cause 180 degree shift when coupled with the capacitor in the midrange band pass. The inductor was later deleted and just a cap was used.  I rebuilt a pair of early 1st order Khorn crossovers a few years ago (for a friend) that have the .5mH inductor in the midrange leg; they sound fantastic!

Andy

PS Yes, I will be blowing the dust off my test equipment for my endeavor; I am just fishing for ideas and perspectives here - thank you again. Also, that response curve by Trey was performed shortly before the release of the Cornwall III (2006) not 45 years ago. 

Link to comment
Share on other sites

2 hours ago, Chris A said:

Here is a 1979 Cornwall in my listening room, with the only modification being the tweeter was switched from the K-77 to the CT-125 (before I bought it):

 

1859452629_Uncorrected79Cornwall(QuarterSpace)SPLandPhase.jpg.c0b5df7fc965f7a5cc3742848c709b44.jpg

 

I think the concern about passive crossover filter component values, etc. is really based on what I'd call archaic notions of "what the passive crossover values should be" based on assumptions of driver/horn electrical impedance.  The problem is, the actual input electrical impedances of the drivers with their horns attached isn't anything like "flat" or single-valued. 

 

I do recommend spending $94 on a UMIK-1 and a few bucks on a microphone stand, downloading REW, and measuring the most important part of the loudspeaker's characteristic: its acoustic output.

 

Above you will see that the real issue is not really SPL response, but rather phase response (the bottom magenta-colored trace). If you want to do something that will increase the sound quality (in this case, soundstage imaging and clarity), release the tweeters from their in-cabinet positions and move them on top of the cabinet, and backwards ~9.75 inches (for the CT-125 tweeter) until they are time-aligned with the midrange horn/driver.  That will improve the sound quality audibly.  Here is a spectrogram showing the degree of time misalignment of the tweeter to midrange:

 

1692679030_Uncorrected79Cornwall(QuarterSpace)Spectrogram.jpg.e712f8f229590ff070521cb38e384f68.jpg

 

That really sharp dogleg in the peak energy time curve at ~475 Hz is really the time misalignment of the woofer to the midrange (or alternatively, a polarity issue at the crossover between the midrange and woofer)--which is actually quite small, and probably inaudible.  What you should pay attention to is the .75 ms delay in the midrange to the tweeter above 5 kHz.

 

Chris


Thank you, Chris, for taking time to post these graphs. They are definitely interesting and give me something to think about. Regarding moving the tweeters to the top of the cabinet, well, I would much rather keep contained inside the cabinet albeit not ideal for time coherency of the drivers. I tend to think moving the tweeters back 9+ inches would create a pretty large step baffle that would introduce diffraction problems that could outweigh the benefits of time alignment, but I may be wrong. I am a firm believer in understanding a problem thoroughly before applying a fix. Your insight is certainly appreciated. 

Link to comment
Share on other sites

5 minutes ago, Klipschguy said:

I tend to think moving the tweeters back 9+ inches would create a pretty large step baffle that would introduce diffraction problems that could outweigh the benefits of time alignment, but I may be wrong

It's easy enough to try it.  Let your ears be the final guide.

 

Phase shifts always increase the delays on the lower frequency drivers using crossover filters.  The minimum phase shift of a two-sided passive crossover filter (higher frequency and lower frequency driver filters) is zero--no filter elements at all--which is what I effectively use with the DSP crossovers (Xilica, miniDSP, Yamaha, ElectroVoice, Hypex, etc.).  Just trim off the unwanted SPL response of the drivers using controlled bandwidth notch filters at the right frequencies, no low pass, high pass, or shelving filters.  Since the drivers themselves are excellent band-passed devices by themselves well outside of their pass bands, so they do the work for you without added all-pass delays of electrical crossover filters.  The results look like this...and sound phenomenally good:

 

1824324579_RonEllisMEHcenterchannelSPLandPhaseResponse.jpg.b3ac005a9161c3986a03875355a525a3.jpg

 

Chris

Link to comment
Share on other sites

2 minutes ago, Klipschguy said:

I tend to think moving the tweeters back 9+ inches would create a pretty large step baffle that would introduce diffraction problems that could outweigh the benefits of time alignment, but I may be wrong. 

 

I tend to agree with this. Physical alignment of the drivers' phase centers is still the ideal, but it's getting to the point where amplification and digital processing is so cheap that passive crossovers and driver misalignment will soon be regarded as "archaic", IMO. Digital time alignment is still second-best, but it's very good.

Link to comment
Share on other sites

3 minutes ago, Chris A said:

Just trim off the unwanted SPL response of the drivers using controlled bandwidth notch filters at the right frequencies, no low pass, high pass, or shelving filters.  Since the drivers themselves are excellent band-passed devices by themselves well out of their pass bands, so they do the work for you without added all-pass delays of electrical crossover filters. 

 

Not quite sure what you mean here, Chris. There is quite a lot of phase shift in a notch filter ... unless you're using FIR notches, which tend to be very long at low frequencies.

Link to comment
Share on other sites

2 minutes ago, Chris A said:

You need to expand on this...

 

If drivers are physically misaligned, as occurs on a flat baffle with drivers of differing depths (more correctly, their phase centers are not all the same distance from the listener), then correction by delay is only correct on-axis. Off axis there is still a misalignment in phase centers.

Link to comment
Share on other sites

8 minutes ago, Edgar said:

Physical alignment of the drivers' phase centers is still the ideal

Only if the horns are co-axial, i.e., the same length and in the same physical space. 

 

Off-axis, physical alignment sucks badly if using separate horns of differing lengths, and is a function of radial orientation.

 

This is not true for digital delays using separate horns of differing lengths mounted with horn mouths in the same vertical plane.

 

Chris

Link to comment
Share on other sites

1 minute ago, Chris A said:

Only if the horns are co-axial, i.e., the same length and in the same physical space. 

 

Off-axis, physical alignment sucks badly if using separate horns of differing lengths, and is a function of radial orientation.

 

Also true. So the issue of time alignment is far more complicated than just lining-up the voice coils.

  • Like 1
Link to comment
Share on other sites

On 12/29/2020 at 5:36 PM, boom3 said:

Here's a test report done in 1999 on Corn IIs (like the ones I have). The "dip" you are referring to is due to floor bounce. It shows up in measurements but room acoustics and our own ears average it out. Corn IIIs and IVs moved the woofer up the panel, so, having a longer path to the floor, that bounce will be moved down in frequency. That is, providing that Roy did not fix it by crossover design.

Narrow dips measured in speaker response normally are inaudible, since, as I said, the room acoustics average things out,. and our own ears tend to "fill in" these dips. A peak is much harder to ignore.

I would not alter the crossover without having measuring gear to see first, what problems you are trying to solve and secondly, if you solved them or not. The response problem I had with my Corns was that they tend to boom in corners. Pulling them 18 inches out from the corners fixed that.

Klipsch_Cornwall_Test_Report 2.pdf 532.55 kB · 4 downloads

I like his general comment about women where he says: "women always prefer music to be only 0.1 db louder than noise within the living room." The 80/20 rule also applies to women, the 20 being reserved for the exceptional to this comment.

 

But the scope of this text goes way beyond the OP. So do women perceive phase anomalies differently than men?

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...