Jump to content

Recommended Posts

Dear all,

 

Stage 2 crossover upgrade V0.23:  Tweaking the 2.5-way design with measurements.

 

So I have built up my second 2.5 way crossover. I made a few minor changes to the layout this time:

  • added back a series resistor to shape the tweeter response a little;
  • I kept the impedance smoothing section made up of C4 and R4 on the main board;
  • added a positive binding post; and
  • relented and unwound a second 3.5 mH inductor to make up my missing 2mH.

 

Here's the tweeter and upper woofer board....

 

2_5way.jpg.6fb3ad8a441ee0692f48ffe2994c6dc4.jpg

 

For my 0.5 section or lower woofer board I tried using one of the deconstructed crossover PCB's that I now have many of after harvesting all my LCR parts. This worked out great and simplified the wiring to the lower woofer which now takes a negative lead from the main board negative binding post. I have used two 5.6mH inductors in series to get my 6db roll off for the lower KG woofer....

 

330142874_0.5waysection.jpg.9c9b65a391b2e1a7c8e3ea82cfe2c6b5.jpg

 

 

So the KG pair are installed back into my main listening room and I've had a few weeks to enjoy them and I am happy to report that they do indeed sound AMAZING!

 

The first listening impression I got was a sense of a big step up in the clarity of the midrange. Vocals really shine through where before these seemed more recessed. With this clarity comes a deeper level of richness in many kinds of music that I don't think was there previously. I really wish I had another pair that I could A/B with, original crossovers to new 2.5ways, but I don't so I'm relying on memory, and all of that sense's frailty.

 

After some time I began to realise just how revealing of the recording quality this design is, and after a while I began to think I wanted to round off the treble just a smidge because I really do like just a bit of warmth, which I think makes medium to loud listening more enjoyable. 

 

So I went back to XSim and came up with a very minor tweak I'm calling my 2.5way V0.23. This adds a single series resistor R2 to the tweeter (S1m) circuit. I modelled and installed a 5 ohms unit....

871494287_2.5waycrossoverschematicV0_23.jpg.d94a874b84fbbe7b8472c08e946a0cba.jpg

This additional resistor has the effect of just slightly tilting the top end downward, softening the very high frequency treble....

 

1164415205_2.5wayCrossoverFreqincsuperimposedinline5ohmnV0_23.jpg.021b0240f0fdd22fe4ff2607037bdf71.jpg

 

R2 is highly tunable, here is the effect of R2 @ 10 ohms.....

 

783583748_2.5wayCrossoverFreqincsuperimposedinline10ohmnV0_23.jpg.f988dc95a984574ad8335b612342c447.jpg

 

Here is R2 @ 3.3 ohms....

 

964452972_2.5wayCrossoverFreqincsuperimposedinline3.3ohmnV0_23.jpg.e54c86b6bfbb52108f130652eb35dfb6.jpg

 

Feel free to tune R2 to taste, there is no right or wrong here, as Troels Gravesen says:

 

"Some call the voicing of speakers an art. I don't think so. Voicing a speaker is a matter of taste like adding spices to a stew. Some like it hot, creamy or crunchy - some don't." 

 

For completeness I wanted to show some measurements. These were all done in my lounge this time. The only change to the position of my speakers where I normally have them when listening is that I moved the right hand speaker out into the room a little so that the front driver baffle just clears my nice cabinet and DIY 15" Alpine sub.

 

Despite being far from an ideal measuring environment I think the results are great. Biased? Yes, probably! At least the data allows for some objectivity.  Here's how I measured....

 

1937636674_nroommeasurementsclose.thumb.jpg.e0cbd82fb1fec13d87eb45f0c4dd6b63.jpg

 

I didn't carefully measure 1 meter away from the cabinet front this time. I just set the mic height by putting it right up to the centre of the tweeter and then pulled the mic stand back roughly 1m and measured.....

 

927282777_nroommeasurements.jpg.3e597369ce6ba69a74ebfd107a2ce1aa.jpg

 

 

Results -  +/- 2db

Here's the left and right comparison, pretty well dead on match for frequency response. I level matched these by knocking 2db off the right enclosure. The right speaker measured a bit louder due in part to the imprecise distance I measured each speaker from and because the left speaker is placed out in what is almost 2 Pi (free standing) space adjacent to the intersection of my lounge/kitchen/dining room compared to the right enclosure being in 0.5 Pi (corner) loading....

 

1587970091_Inroom.png.86ce8037f1a28af660d95f1086b2017e.png

 

 

Zooming in I think that we might safely call this a 2db +/- result? That sharp dip between 800 and 900 hz is room related, it isn't a feature in any of my outdoor measurements ....

 

 606556472_2.5way-2db.thumb.jpg.e891e9d3671ec746009a8c8ad453ae60.jpg

 

 

Distortion measurement - a comparison

Wayne Parham at Pi Speakers has published various system and horn/driver combo frequency measurements that make for an interesting comparison.

 

His Pi H290C waveguide with a B&C DE250 have measured distortion just over 30db down.....

  • Black = measured frequency response (also known as the fundamental).
  • Blue= 2nd harmonic;
  • Pink = 3rd harmonic.

 

H290C_2.83v_unsmoothed.thumb.jpg.f02eb9cf39194af0a4f924f3d3b12992.jpg

 

 

For comparison here is the distortion sweep of the KG's. Note that the output level was running at 90db which is quite loud and indicative of where I'd listen when the family are out.

  • Black = measured frequency response.
  • Red = 2nd harmonic;
  • Purple = 3rd harmonic.

 

100341122_Distortion1_24th.thumb.png.4536d49aba67992bd5cd2cfdb0b7688f.png

 

I'm prepared to say the KG's are a very clean, low distortion system with the fundamental better than 40db above measured distortion.  Way to go Klipsch!

This graph also shows the KG's are solid down to 30hz measured like this. 

 

Based on this data I'd like to venture to anyone that says the KG's are just speakers for teenagers and rockers should rethink everything they thought they knew about them.  I believe this shows conclusively the 5.5's can be extremely fine speakers, by any measure.  :emotion-14:

 

So, I'm going to keep listening some more with this crossover design in place, I'm really enjoying it and feel that for me, the 5 ohms R2 on the tweeter hits a smooth, rich, balanced listen. 

 

Hooked up to my new, secondhand, cleaned up AX-570 (which just quietly kills my brand new Onkyo TX-RZ3100 (Pioneer SC-LX901 in a different skin!) hooked up to my KG's in my room) the bass is super articulate and authoritative. Midrange so clean and clear, treble sweeeeeet mmmmm!

 

At least to me, of course!  :P

 

 

395584391_OnkYam.thumb.jpg.c8cc7e43a50bf8e81cacb45d5e5f5e18.jpg

 

 

 

I think we have a winner here. The KG's continue to reward.  The journey continues!

 

Cheers.  :emotion-22:

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

  • 2 weeks later...

Maz4bz,

 

Just wanted to say thanks again for the time and effort you put in to improving your speakers and the benefit it will give anyone who looks over what you've done and decides to apply a similar tactic. I am still having those "I can't believe the sound I am hearing from my speakers!" moments! It's amazing what 2 10-ohm resistors did.

 

Another thing I noticed: I use the chromecast audio streamer to stream Amazon music over wi-fi. The bass and treble controls on the chromecast operate on the digital level so I believe they can be considered DSP level controls. Well, before adding the resistors, I was less than happy with the changes made by adjusting those controls. The music always sounded worse. After adding the resistors, the changes I make to bass and treble through the chromecast audio dongle make meaningful changes to music lacking a bit of high-or-low-end fidelity. The music quality does not suffer, which is supposed to be what DSP is all about. Some talk of DSP as the cure to all woes, but now I think it can work either way; improvements in passive circuits can make all the difference, even to DSP.

  • Thanks 2
Link to comment
Share on other sites

  • 5 months later...
On 9/3/2019 at 7:50 AM, Maz4bz said:

 

Maz4bz 

It's it possible to replace the 5 ohm(ish) resistors with 1 ton10 ohm 5A pots to be able to tune the tweeter?

 

Stage 2 crossover upgrade V0.23:  Tweaking the 2.5-way design with measurements.

 

So I have built up my second 2.5 way crossover. I made a few minor changes to the layout this time:

  • added back a series resistor to shape the tweeter response a little;
  • I kept the impedance smoothing section made up of C4 and R4 on the main board;
  • added a positive binding post; and
  • relented and unwound a second 3.5 mH inductor to make up my missing 2mH.

 

Here's the tweeter and upper woofer board....

 

2_5way.jpg.6fb3ad8a441ee0692f48ffe2994c6dc4.jpg

 

For my 0.5 section or lower woofer board I tried using one of the deconstructed crossover PCB's that I now have many of after harvesting all my LCR parts. This worked out great and simplified the wiring to the lower woofer which now takes a negative lead from the main board negative binding post. I have used two 5.6mH inductors in series to get my 6db roll off for the lower KG woofer....

 

330142874_0.5waysection.jpg.9c9b65a391b2e1a7c8e3ea82cfe2c6b5.jpg

 

 

So the KG pair are installed back into my main listening room and I've had a few weeks to enjoy them and I am happy to report that they do indeed sound AMAZING!

 

The first listening impression I got was a sense of a big step up in the clarity of the midrange. Vocals really shine through where before these seemed more recessed. With this clarity comes a deeper level of richness in many kinds of music that I don't think was there previously. I really wish I had another pair that I could A/B with, original crossovers to new 2.5ways, but I don't so I'm relying on memory, and all of that sense's frailty.

 

After some time I began to realise just how revealing of the recording quality this design is, and after a while I began to think I wanted to round off the treble just a smidge because I really do like just a bit of warmth, which I think makes medium to loud listening more enjoyable. 

 

So I went back to XSim and came up with a very minor tweak I'm calling my 2.5way V0.23. This adds a single series resistor R2 to the tweeter (S1m) circuit. I modelled and installed a 5 ohms unit....

871494287_2.5waycrossoverschematicV0_23.jpg.d94a874b84fbbe7b8472c08e946a0cba.jpg

This additional resistor has the effect of just slightly tilting the top end downward, softening the very high frequency treble....

 

1164415205_2.5wayCrossoverFreqincsuperimposedinline5ohmnV0_23.jpg.021b0240f0fdd22fe4ff2607037bdf71.jpg

 

R2 is highly tunable, here is the effect of R2 @ 10 ohms.....

 

783583748_2.5wayCrossoverFreqincsuperimposedinline10ohmnV0_23.jpg.f988dc95a984574ad8335b612342c447.jpg

 

Here is R2 @ 3.3 ohms....

 

964452972_2.5wayCrossoverFreqincsuperimposedinline3.3ohmnV0_23.jpg.e54c86b6bfbb52108f130652eb35dfb6.jpg

 

Feel free to tune R2 to taste, there is no right or wrong here, as Troels Gravesen says:

 

"Some call the voicing of speakers an art. I don't think so. Voicing a speaker is a matter of taste like adding spices to a stew. Some like it hot, creamy or crunchy - some don't." 

 

For completeness I wanted to show some measurements. These were all done in my lounge this time. The only change to the position of my speakers where I normally have them when listening is that I moved the right hand speaker out into the room a little so that the front driver baffle just clears my nice cabinet and DIY 15" Alpine sub.

 

Despite being far from an ideal measuring environment I think the results are great. Biased? Yes, probably! At least the data allows for some objectivity.  Here's how I measured....

 

1937636674_nroommeasurementsclose.thumb.jpg.e0cbd82fb1fec13d87eb45f0c4dd6b63.jpg

 

I didn't carefully measure 1 meter away from the cabinet front this time. I just set the mic height by putting it right up to the centre of the tweeter and then pulled the mic stand back roughly 1m and measured.....

 

927282777_nroommeasurements.jpg.3e597369ce6ba69a74ebfd107a2ce1aa.jpg

 

 

Results -  +/- 2db

Here's the left and right comparison, pretty well dead on match for frequency response. I level matched these by knocking 2db off the right enclosure. The right speaker measured a bit louder due in part to the imprecise distance I measured each speaker from and because the left speaker is placed out in what is almost 2 Pi (free standing) space adjacent to the intersection of my lounge/kitchen/dining room compared to the right enclosure being in 0.5 Pi (corner) loading....

 

1587970091_Inroom.png.86ce8037f1a28af660d95f1086b2017e.png

 

 

Zooming in I think that we might safely call this a 2db +/- result? That sharp dip between 800 and 900 hz is room related, it isn't a feature in any of my outdoor measurements ....

 

 606556472_2.5way-2db.thumb.jpg.e891e9d3671ec746009a8c8ad453ae60.jpg

 

 

Distortion measurement - a comparison

Wayne Parham at Pi Speakers has published various system and horn/driver combo frequency measurements that make for an interesting comparison.

 

His Pi H290C waveguide with a B&C DE250 have measured distortion just over 30db down.....

  • Black = measured frequency response (also known as the fundamental).
  • Blue= 2nd harmonic;
  • Pink = 3rd harmonic.

 

H290C_2.83v_unsmoothed.thumb.jpg.f02eb9cf39194af0a4f924f3d3b12992.jpg

 

 

For comparison here is the distortion sweep of the KG's. Note that the output level was running at 90db which is quite loud and indicative of where I'd listen when the family are out.

  • Black = measured frequency response.
  • Red = 2nd harmonic;
  • Purple = 3rd harmonic.

 

100341122_Distortion1_24th.thumb.png.4536d49aba67992bd5cd2cfdb0b7688f.png

 

I'm prepared to say the KG's are a very clean, low distortion system with the fundamental better than 40db above measured distortion.  Way to go Klipsch!

This graph also shows the KG's are solid down to 30hz measured like this. 

 

Based on this data I'd like to venture to anyone that says the KG's are just speakers for teenagers and rockers should rethink everything they thought they knew about them.  I believe this shows conclusively the 5.5's can be extremely fine speakers, by any measure.  :emotion-14:

 

So, I'm going to keep listening some more with this crossover design in place, I'm really enjoying it and feel that for me, the 5 ohms R2 on the tweeter hits a smooth, rich, balanced listen. 

 

Hooked up to my new, secondhand, cleaned up AX-570 (which just quietly kills my brand new Onkyo TX-RZ3100 (Pioneer SC-LX901 in a different skin!) hooked up to my KG's in my room) the bass is super articulate and authoritative. Midrange so clean and clear, treble sweeeeeet mmmmm!  :P

 

395584391_OnkYam.thumb.jpg.c8cc7e43a50bf8e81cacb45d5e5f5e18.jpg

 

 

 

I think we have a winner here. The KG's continue to reward.  The journey continues!

 

Cheers.  :emotion-22:

 

Link to comment
Share on other sites

On ‎8‎/‎14‎/‎2020 at 10:47 AM, Randall said:

It's it possible to replace the 5 ohm(ish) resistors with 1 ton10 ohm 5A pots to be able to tune the tweeter?

 

Hi Randall,

I'm no electronic engineer however I think the simple answer is yes, assuming that pot can dissipate the electrical load adequately. However I would only do this for R2 which gives the ability to tune the slope of the tweeters response. 

 

I would not replace or change R3. This would raise (or lower) the entire output of the tweeter and create a situation similar to the stock crossover where there would be a step up (or down) in the frequency response at crossover like this....

KGWoofersoriginalcrossoverInked.png.0c830eb6fa52f7fb5708e5a9e78f764f.png

This would degrade the on and off axis response, which we want to avoid. You would see this in the polar response with a sharp break in directivity something like this ....

1565666967_0-90PolarOriginalCrossover.png.bed1b8788bb01a447abc7b28bee2642a.png

Good luck and please post back what you find.

 

Cheers.   🍻

 

Link to comment
Share on other sites

  • 2 months later...
3 hours ago, trongnguyen said:

Tôi nghĩ rằng kg 5.5 có độ nhạy 98db/m . Nhưng nếu bạn nâng cấp bộ phân tần lên tời bậc 2 hay bậc 3 điều đó sẽ làm giảm độ nhạy của nó , bạn có nghĩ vậy không ?

Google Translate:

"I think kg 5.5 has a sensitivity of 98db/m . But if you upgrade the crossover to a 2nd or 3rd order winch that will reduce its sensitivity, don't you think?"

 

@trongnguyen

Welcome to this Forum! Please, try to post in English. We are international here and English is the lingua franca. By doing so You will get more reactions to your posts.

Link to comment
Share on other sites

G'day Trong,

The crossover order won't affect the efficiency, but the conversion to a 2.5 way crossover most certainly does.

 

This is because the lower woofer now rolls off early leaving the upper woofer to handle the rest of the frequency range up to crossover point with the tweeter.

 

In theory this change reduces the overall system efficiency by 3db.

 

Oh and I don't believe this system was ever 98db, more like 93 in the low end, about 96 above crossover in original spec. Mine are about 90db now which is still quite efficient by non-horn loaded speaker standards.

 

Cheers.  🍻

Link to comment
Share on other sites

  • 1 month later...

Dear all,

another minor update - more data.

 

I wanted to confirm the modelled WinISD bass response using the DATS measurements I'd taken from the woofers.

 

Here's the modelled response of the two KG woofers showing a tuning frequency in the high 40hz with a 2.5db hump above this....

 

KG5.5modelled1.png.230741ecb8f024df93b070bd8739a6c7.png

 

 

I recently had a chance to haul one of my KG's outside for some ground plain measurements. I wanted to see how the modelling compared to measured...

 

Groundplane.jpg.1d85f904a5c2725f7396b00ce6f14f38.jpg

 

This measurement approach enables a more accurate measurement of the bass response of a system.

 

These are done at 1m with Omnimic....

 

1037251161_KG5.5groundplane@1m.png.122c48c8783a30385f76bda4cc556831.png

 

 

Next is a series of distortion measurements at 5db intervals with reference to 100hz.....

 

95db:

 

1301783791_GPDistortion@95db.thumb.png.6d798803dba8d9f8b1d789429e3ee0ca.png

 

 

100db:

 

346895567_GPDistortion@100db.thumb.png.6f116abcf389e7f56aa5dfb0d1c91ebe.png

 

 

105db:

 

1333451299_GPDistortion@105db.thumb.png.04dc05aac3f494ce6b5a4596c2b044a7.png

 

 

From this I think we can deduce that the 2.5 way alignment has strong bass response down to 30hz and is very clean/low distortion up to dangerously high SPL - remembering permanent hearing damage onset is 87db.

 

In the years I've had my KG's I've had numerous speaker systems through my home, but I keep coming back to the KG's. They remain my reference speaker. 🙂

 

Cheers.

 

🍻

 

  • Like 1
Link to comment
Share on other sites

  • 3 months later...

Hi there,

Looking at the brochure for the KLF 10 I'm going to say no it probably won't. The KG uses different drivers albeit I can see the tweeters use the same diaphragm.

 

Naturally you can have a try and see if you like the result, the cost and effort is low and the modifications are completely reversible.

 

For sure the KLF 10 is a two way like the KG 5.5 so changing to 2.5 will certainly assist.  

 

The only way to know for sure would be to measure the result once built up and compare to the original unmodified measured performance. 

 

Another KLF 10 owner made the L-pad modification that I did here (but tweaked to suit the KLF 10 crossover) and reported positive results, perhaps worth a read?

 

Good luck and please do report back if you go ahead. 

 

🍻

Link to comment
Share on other sites

i was thinking of converting the klf to 2.5 as you did on your KG 5.5 i was also gonna build some stands and upgrade the tweeter to faital pro 111 , im running a class D icepower 2000as2 2x 1000 rms i was thinking of also upgrading the power handling on the speakers, i was thinking of making these a rf-7 killer 

Link to comment
Share on other sites

Sounds like an interesting plan, if you have measurement gear it might be worth it. Without measurement gear it'd be a bit like pin the tail on the donkey, so to speak. 😅

 

I would be surprised if you could hear a difference between an optimised KLF 10 with factory parts and a Frankenstein version, to be honest. 

 

In my opinion Klipsch used very good quality parts in these speakers and probably took great care to optimise the horn and driver combo.

 

If it were me, I'd be using the Faital with a matching Faital horn in a new build and keeping the KLF factory, other than the crossover.

 

Either way, good luck!   

Link to comment
Share on other sites

  • 2 weeks later...
  • 3 months later...

hello again, i have built my own crossover for a set of speakers Klipsch RF-83 with new parts I followed the schematic. but when the volume and bass is set, the amp cuts out, do you have an email I can contact you to get you to look at the crossover if something I did wrong? , looks like you know your crossovers for Klipsch, if you can 

Link to comment
Share on other sites

  • 5 months later...

Hey Maz!  I just wanted to say thank you for all the work on the KG 5.5's.  I picked up a pair locally, and heard exactly what you measured.  WAY too bright and a hole where the woofers have destructive cancellation.  I bought the Crites diaphragms, Acousti-stuffed the cabinets with additional bracing, and built a set of your 2.5 way crossovers.  I used poly film and foil caps from Solen and Dayton Audio, as well as Dayton and Jantzen air core inductors where I could.

     The sound is great.  Mainly I noticed the treble balance at first.  Then the soundstage became more 3D.  Imaging is tight, and bass is well-defined.  Super job!

  • Like 1
Link to comment
Share on other sites

Thanks for the feedback Harry. So glad you enjoyed the outcome. It's a big investment of time and effort to upgrade the KG's as we've done but the results, I feel, are warranted.

 

In the meantime, I've been enjoying looking at looking at Erin's Audio Corner where he uses a Klippel nearfield scanner (worth $100k) which gives a full suite of measurements. 

 

Comparatively I look at the polar of a JBL 4329P, for example, and see terrific likeness with the KG's performance. What I'm trying to say, I guess, is that the performance achievable with the KG 5.5's very high for the entry price.

 

Enjoy!  🍻

 

 

JBL 4329P Polar.jpg

Link to comment
Share on other sites

  • 8 months later...

Ok, I took your last posted design and put it in XSim (2.5 Way Crossover). LMK if I got anything wrong. 

 

High Pass Filter for Tweeter (S1):

  • Capacitors (C1 & C2): 3.32 μF and 47 μF
  • Inductor (L1): 0.62 mH
  • Resistors (R2 & R3): 5.1 Ohms and 5.6 Ohms

Band Pass Filter for Midrange Woofer (S2m):

  • Capacitors (C3 & C5): 13 μF and 10 μF
  • Inductors (L2 & L4): 2 mH and 1.2 mH
  • Resistor (R1): 3.3 Ohms

Low Pass Filter for Low Frequency Woofer (S3m):

  • Inductor (L3): 11 mH
  • Capacitor (C4): 47 μF
  • Resistor (R4): 20 Ohms (Zobel network)

Crossover Frequencies

  • High Pass (Tweeter): fcH: 3.6 kHz
  • Band Pass (Midrange): fc1: 313 Hz, fc2 1.23 kHz
  • Low Pass (Woofer): fcL: 69 Hz

image.png.23f5d683f0161ed499a84ff76600b357.png

 

I have no idea what I'm doing, but here's a stab (2.5AIXOOC). 

High Pass Filter for Tweeter (S1):

  • Capacitors (C1 & C2): 3.3 μF and 47 μF
  • Inductor (L1): 0.84 mH
  • Resistors (R2 & R3): 7 Ohms and 5.6 Ohms

Band Pass Filter for Midrange Woofer (S2m):

  • Capacitors (C3 & C5): 15 μF and 8.2 μF
  • Inductors (L2 & L4): 2 mH and 0.84 mH
  • Resistor (R1): 3 Ohms

Low Pass Filter for Low Frequency Woofer (S3m):

  • Inductor (L3): 3.3 mH
  • Capacitor (C4): 47 μF
  • Resistor (R4): 20 Ohms (Zobel network)

Crossover Frequencies

  • High Pass (Tweeter): fcH: 3 kHz
  • Band Pass (Midrange): fc1: 290 Hz, fc2 500 Hz
  • Low Pass (Woofer): fcL: 130 Hz

image.png.b38ab637c449987812b3eac6af0fc456.png

 

Got a pair of 5.5's that need new crossovers, ad it's either going to be your design or a tweak of it. Let me know what you think.

2.5 Way Crossover.dxo 2.5AIXOOC.dxo

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...