I used a Technics SU-9070 preamp and SE-9060 power amp, all from the 9xxx slimline rack-mounted series.
Without any compensation, the sound was clear but nothing special. The Reference 105 I also used in the same room gave better bass.
Introducing the peaks through the parametric equalizer gave a much more present low frequency spectrum. The drivers were nowhere near the end of their excursion limits even with the amplifier close to maximum output. Listening to tracks from Jacques Loussier Bach series with 5 stringed upright bass solos, it was revealing with every single note sounding different - as far as I've read, a good indication that the bass is tight and clear rather than a muddy boom. Drum solos from the same series of albums were also impressive, with the mastering helping a lot. The drum kit was recorded extending from the left to the right edge of the stereo image. With the speakers spaced around 3meters (9ft) apart, it really sounded like I had the drummer in front of me. The kick drum wasn't harsh, but had a nice sounding feeling of listening the membrane being kicked and oscilating.
The KUBE didn't change much in the lower end, except an improvement in feeling safer with higher volumes. With some types of music (mainly pop) I would reduce the volume as occasionally the drivers would sound clicking at their limits if played very loud. That is eliminated with the KUBE, obviously because of a better suited equalization curve.
What did change with the KUBE was high frequency response. It is then when I realized that the tweeters did sound a bit bright without it. It's something I had noticed and initially thought it was ferrofluid drying out. I was wondering how a tweeter with dried out ferrofluid sounded (dull or bright). With the KUBE activated something changed in the upper frequencies and the tweeters are "tamed" (?) without sounding dull. I was able to re-create (almost) the same effect with the equalizer by bringing down the 3 upper bands by (IIRC) 1, 2 and 3dB with a very low Q.
What I have not decided yet is which settings on the KUBE I like best. It's been only 6 months since I hooked it up and because of the winter I was not in the basement often. Will listen more as spring brings up the temperature.
What is certain is that when listening to the 107s with the KUBE, he equalizer is again in bypass mode, as it was with the R105s. And the big difference is that when I listen sitting on the couch (a puffy leather couch) I can feel the vibrations of the lower frequency transients (especially kick drum and upright bass).
I attribute most of the performance to the treatment of the room. I have made angled mineral wool absorbers (10" thick) and mounted them behind the speakers on both corners, from floor to ceiling (8ft high). The room was boomy before them and bass frequencies are much more contained, tight and predictable now.
Without any compensation, the sound was clear but nothing special. The Reference 105 I also used in the same room gave better bass.
Introducing the peaks through the parametric equalizer gave a much more present low frequency spectrum. The drivers were nowhere near the end of their excursion limits even with the amplifier close to maximum output. Listening to tracks from Jacques Loussier Bach series with 5 stringed upright bass solos, it was revealing with every single note sounding different - as far as I've read, a good indication that the bass is tight and clear rather than a muddy boom. Drum solos from the same series of albums were also impressive, with the mastering helping a lot. The drum kit was recorded extending from the left to the right edge of the stereo image. With the speakers spaced around 3meters (9ft) apart, it really sounded like I had the drummer in front of me. The kick drum wasn't harsh, but had a nice sounding feeling of listening the membrane being kicked and oscilating.
The KUBE didn't change much in the lower end, except an improvement in feeling safer with higher volumes. With some types of music (mainly pop) I would reduce the volume as occasionally the drivers would sound clicking at their limits if played very loud. That is eliminated with the KUBE, obviously because of a better suited equalization curve.
What did change with the KUBE was high frequency response. It is then when I realized that the tweeters did sound a bit bright without it. It's something I had noticed and initially thought it was ferrofluid drying out. I was wondering how a tweeter with dried out ferrofluid sounded (dull or bright). With the KUBE activated something changed in the upper frequencies and the tweeters are "tamed" (?) without sounding dull. I was able to re-create (almost) the same effect with the equalizer by bringing down the 3 upper bands by (IIRC) 1, 2 and 3dB with a very low Q.
What I have not decided yet is which settings on the KUBE I like best. It's been only 6 months since I hooked it up and because of the winter I was not in the basement often. Will listen more as spring brings up the temperature.
What is certain is that when listening to the 107s with the KUBE, he equalizer is again in bypass mode, as it was with the R105s. And the big difference is that when I listen sitting on the couch (a puffy leather couch) I can feel the vibrations of the lower frequency transients (especially kick drum and upright bass).
I attribute most of the performance to the treatment of the room. I have made angled mineral wool absorbers (10" thick) and mounted them behind the speakers on both corners, from floor to ceiling (8ft high). The room was boomy before them and bass frequencies are much more contained, tight and predictable now.
I'd like to know this too. I've a "Kube Pro" from a pair of other Kef monitors (P60s? Can't remember the model) and it looks like nothing more than the same Kube circuit in a rack-mount and XLR inputs . In other words, not something to get excited about - Im sure ther are 107 owners drooling over the idea of finding a Kube Pro..
Here you go...I'm still overseas
I must have missed your reply to mine ... They look great.
Did you ever go home to listen to them? How are the Raal ribbons?
I would like to bring it back to life too.... It's about time I did something to my KEf 107s (sitting waiting for me to refurb them for years) - I at least have the foam surrounds and a Meridian 601 with problems that I got for £200 last year (working well enough to replace the blown Kube of mine).
The head unit needs the bass filtered out before it reaches the head unit cross-over - I would say "that's all".
I'd have to go back to AndrewJ's explanation of the bass unit cross-over to work out which part is the filter and which the conjugate loading.
The head unit needs the bass filtered out before it reaches the head unit cross-over - I would say "that's all".
I'd have to go back to AndrewJ's explanation of the bass unit cross-over to work out which part is the filter and which the conjugate loading.
Anybody still there after four years...?
I have a pair of 107's sans Cube and they have a pronounced hiss at upper mid/high frequenies.
Is that to be expected withoput the Cube? I understand that the Cube basically is for extention of bass response but that it also can change the complete tonal balance?
BTW Did anybody succeed in tracing out that mid/high xover in the turret?
Jan
A long time to wait for a reply, but the original Kube for the Mk1s also EQ'd the mid/tweeter by taking out a 4khz peak and increasing their efficiency. I'd say that without the Kube and no other mods, it will exhibit a 4Khz peak along with the lack of bass (by the band-pass bass enclosure loosing efficiency compared to with Kube) which may then lead the 4Khz peak to sound even more subjectively noticable.
If its not too much a task, perhaps it's just best to make a whole new x-over for the head unit with a new tweeter, taking some components from the high pass part of the bass unit x-over. I have Scanspeak 97000s which have the same mounting holes as the T33 when you remove its metal plate, but require a slightly larger hole for the Scanspeak's tiny horn bit. So a small mod and it will fit directly on the Kef plate.
As an aside, I was recommended ages ago to isolate the head unit more - by someone who praised to results of doing this with Kef 105s - and I coincidently have some granite slabs that fit exactly in place of the fabric'd top of the bass units. I was looking to remove the post the head mounts on and instead create an adjustable stand for it to sit on the granite via sorbothane .
As an aside, I was recommended ages ago to isolate the head unit more - by someone who praised to results of doing this with Kef 105s - and I coincidently have some granite slabs that fit exactly in place of the fabric'd top of the bass units. I was looking to remove the post the head mounts on and instead create an adjustable stand for it to sit on the granite via sorbothane .
I did an A/B with and without the KUBE on 107's (MK1). I was expecting the obvious difference in bass. I wasn't expecting a significant difference in the mids/highs. At lower levels without the KUBE they sounded maybe just more detailed. At more spirited levels, they were harsh to the point of wincing for me, and my dogs ears twitching. I believe this is attributed to the 4 kHz bump though haven't done a sweep yet with my new mic. I was aware of it already due mostly to this thread. While I've always been interested in DIY, I joined specifically to keep an eye on this topic.
With not finding so far that someone else has cleared the path, I was considering getting professional help for crossover mods or replacement for goals below. May not be in the spirit of DIY, but I don't have the knowledge or skills to do so. I have one set of 107's with a KUBE and one set without. I thought I had a solution for bass EQ and do without the KUBE, but didn't understand at that time that no KUBE would cause issues (for my tastes) on the mids/highs.
1.) Bi-amp splitting between bass and head units. I would like to be able to bi-amp, but this would open up more options for isolating bass EQ, if still required.
2.) Modify bass unit crossovers to flatten response and eliminate or reduce need for KUBE or other EQ. Or leave as is and use parametric EQ or DSP device to adjust response.
3.) Modify head unit crossovers to eliminate 4 kHz peak without use of KUBE. I have no idea what that does for conjugation or what that means, or concerns for crossover circuit impedance from Andrew's post. I'm not sure how big of a deal what the KUBE does is to help the circuit impedance, but I happily ran without the KUBE with bass EQ except where I did an A/B where the harshness without KUBE was obvious. Not sure if the harshness is frequency only or could be other issues in the circuit.
I also thought about tweeter replacement. I have seen owners who claim better results and those who say they sound worse with newer tweeters. That is probably without crossover changes. I would feel better if an alternative tweeter was interchangeable without head modifications and using the same crossover as the T33 in case I wanted to go back to it. I would like them to sound as good as possible, but I really enjoy the sound signature as is and would not to turn these into an over-analytical or forward/harsh sounding speaker.
Natberg - isolating the head unit is an interesting idea. I hadn't heard of anyone trying that. Would you still intend it to swivel?
As an aside, I noticed what I think is a resonance in the head unit at around 160 kHz. At first I thought it was an issue with 1 B110, but then noticed it from both units. I haven't caught it playing music yet, but it sounded pretty obvious/bad with a test tone. Has anyone else experienced this?
With not finding so far that someone else has cleared the path, I was considering getting professional help for crossover mods or replacement for goals below. May not be in the spirit of DIY, but I don't have the knowledge or skills to do so. I have one set of 107's with a KUBE and one set without. I thought I had a solution for bass EQ and do without the KUBE, but didn't understand at that time that no KUBE would cause issues (for my tastes) on the mids/highs.
1.) Bi-amp splitting between bass and head units. I would like to be able to bi-amp, but this would open up more options for isolating bass EQ, if still required.
2.) Modify bass unit crossovers to flatten response and eliminate or reduce need for KUBE or other EQ. Or leave as is and use parametric EQ or DSP device to adjust response.
3.) Modify head unit crossovers to eliminate 4 kHz peak without use of KUBE. I have no idea what that does for conjugation or what that means, or concerns for crossover circuit impedance from Andrew's post. I'm not sure how big of a deal what the KUBE does is to help the circuit impedance, but I happily ran without the KUBE with bass EQ except where I did an A/B where the harshness without KUBE was obvious. Not sure if the harshness is frequency only or could be other issues in the circuit.
I also thought about tweeter replacement. I have seen owners who claim better results and those who say they sound worse with newer tweeters. That is probably without crossover changes. I would feel better if an alternative tweeter was interchangeable without head modifications and using the same crossover as the T33 in case I wanted to go back to it. I would like them to sound as good as possible, but I really enjoy the sound signature as is and would not to turn these into an over-analytical or forward/harsh sounding speaker.
Natberg - isolating the head unit is an interesting idea. I hadn't heard of anyone trying that. Would you still intend it to swivel?
As an aside, I noticed what I think is a resonance in the head unit at around 160 kHz. At first I thought it was an issue with 1 B110, but then noticed it from both units. I haven't caught it playing music yet, but it sounded pretty obvious/bad with a test tone. Has anyone else experienced this?
To clarify some comments, the conjugate load matching crossover sections are part of the passive network and are fully independent of the KUBE equalizer. If I remember correctly there were some sections that fully encompassed the network (went from + input to -input without having a connection to drivers) and there were also some sections more "interior" to the network. All-encompassing sections can be snipped without changing system response but the interior sections can not.
Crossovers can't be modified to smooth out the bass response. This would require huge component values and a large loss of system efficiency. The benefit of the Kube was the ability to flatten and extend bass more than passive networks would allow.
If you were to attempt a tweeter crossover change or a tweeter replacement, then the conjugate sections should be removed. They would only suit the original crossover and network and might give nasty impedance consequences if other components around the conjugates were revised.
Biamping would probably just require separating the woofer from the mid/tweeter sections after removing any of the all encompassing conjugate sections (leaving any of the interior sections).
I believe the head assembly used compliant driver mountings to reduce cabinet resonances. I'm not sure what the 160Hz problem is. Note that removing the head assembly from its stock location is likely to mess up response due to altering its phase blending to the woofer, which was surely optimized in the KEF design process.
If you really want to alter things then I think that careful measurement with REW and development of custom equalization for an otherwise stock system is the way to go. In the end you would have bass EQ that exactly fit your room and seating locations. You would be able to deal with the 4k issue and also add your own favorite house curve.
Apply the results to the best DSP you can afford, open a bottle of your favorite bourbon, sit back and...
Crossovers can't be modified to smooth out the bass response. This would require huge component values and a large loss of system efficiency. The benefit of the Kube was the ability to flatten and extend bass more than passive networks would allow.
If you were to attempt a tweeter crossover change or a tweeter replacement, then the conjugate sections should be removed. They would only suit the original crossover and network and might give nasty impedance consequences if other components around the conjugates were revised.
Biamping would probably just require separating the woofer from the mid/tweeter sections after removing any of the all encompassing conjugate sections (leaving any of the interior sections).
I believe the head assembly used compliant driver mountings to reduce cabinet resonances. I'm not sure what the 160Hz problem is. Note that removing the head assembly from its stock location is likely to mess up response due to altering its phase blending to the woofer, which was surely optimized in the KEF design process.
If you really want to alter things then I think that careful measurement with REW and development of custom equalization for an otherwise stock system is the way to go. In the end you would have bass EQ that exactly fit your room and seating locations. You would be able to deal with the 4k issue and also add your own favorite house curve.
Apply the results to the best DSP you can afford, open a bottle of your favorite bourbon, sit back and...
Thank you Speaker Dave! That may talk me off the ledge of thinking of trying to modify the crossover other than maybe bi-amping. I did skip a few planned steps. I am going to take measurements with new UMIK-1 mic and get familiar using REW. I will then use REW to adjust KUBE and/or bass EQ as well as integrating with subwoofer. Hopefully by then I can determine what DSP would be a best first attempt for me. I am currently thinking of a miniDSP 2x4HD. I would start focused around adjustments for below 160 Hz and subwoofer integration, and then make attempts for up to full frequency range pending what my REW results are showing. If that went well I may stay there or upgrade to DIRAC for improved room correction.
I was jumping ahead to what I thought may be the next step would be to have "corrections" made in crossover that could eliminate the need for KUBE. Not so much for my in room response, but the known bass and 4 kHz areas. I thought some of the KEF KUBE explanation was to help users adjust for in room response and say cut out low frequency thumps for vinyl, but looks like from this thread that the KUBE is required for correct frequency response.
I was thinking I may not want any DSP as it may degrade or distort the signal. I roughly understand that everything you add will have some impact to the signal. I have heard about DSP for years, but am just really learning about it and wonder why it isn't implemented more for 2-channel. I previously wanted to get away from subs and measurements so went to 2-channel, and was certainly not thinking of wanting to use software instead when I did. Not sure if it's not as widely used because the results are poorer, people don't want to mess with it like I didn't, or don't know/see it as an option. There aren't many 2-channel pre-amps that I see that include DSP. I mostly see users commenting on their home theatre or computer/studio setups.
I'm not sure what the best DSP I could afford would be. I know there are various standalone DSP and pre-amps with them integrated. After some research I was thinking miniDSP as user-friendly for a beginner. Looking at miniDSP the 2x4HD seems to get you up to about the highest level with full range capability and ability to add DIRAC. The other levels add other non-DSP features or bass control and bass management that I think mostly apply if you are running multiple subs (which I will eventually). To get to the highest level you have to go up to say a $4,000 Monolith processor. Not sure that "bass management" or "bass control" is really needed with 107's as KEF already worked out phase and optimizing crossover to head unit (unless I decide to break their crossover design).
I know I can find more detailed DSP discussion elsewhere, but for the purpose of this thread, is DSP the way to go KUBE-less and possibly better than KUBE with capability to adjust for room issues across entire range (I think Speaker Dave is saying yes?, if I should open a favorite bottle to enjoy!). From my experience using just my ears, a bass EQ is a good replacement for the KUBE for bass response. I would expect DSP would be even better for adjusting bass. I don't know if DSP will cause new issues when trying to remove the 4 kHz bump. I guess I will have to try and see.
I'm not sure how to do this physically on the board, but would the correct concept for bi-amping be below?
I was jumping ahead to what I thought may be the next step would be to have "corrections" made in crossover that could eliminate the need for KUBE. Not so much for my in room response, but the known bass and 4 kHz areas. I thought some of the KEF KUBE explanation was to help users adjust for in room response and say cut out low frequency thumps for vinyl, but looks like from this thread that the KUBE is required for correct frequency response.
I was thinking I may not want any DSP as it may degrade or distort the signal. I roughly understand that everything you add will have some impact to the signal. I have heard about DSP for years, but am just really learning about it and wonder why it isn't implemented more for 2-channel. I previously wanted to get away from subs and measurements so went to 2-channel, and was certainly not thinking of wanting to use software instead when I did. Not sure if it's not as widely used because the results are poorer, people don't want to mess with it like I didn't, or don't know/see it as an option. There aren't many 2-channel pre-amps that I see that include DSP. I mostly see users commenting on their home theatre or computer/studio setups.
I'm not sure what the best DSP I could afford would be. I know there are various standalone DSP and pre-amps with them integrated. After some research I was thinking miniDSP as user-friendly for a beginner. Looking at miniDSP the 2x4HD seems to get you up to about the highest level with full range capability and ability to add DIRAC. The other levels add other non-DSP features or bass control and bass management that I think mostly apply if you are running multiple subs (which I will eventually). To get to the highest level you have to go up to say a $4,000 Monolith processor. Not sure that "bass management" or "bass control" is really needed with 107's as KEF already worked out phase and optimizing crossover to head unit (unless I decide to break their crossover design).
I know I can find more detailed DSP discussion elsewhere, but for the purpose of this thread, is DSP the way to go KUBE-less and possibly better than KUBE with capability to adjust for room issues across entire range (I think Speaker Dave is saying yes?, if I should open a favorite bottle to enjoy!). From my experience using just my ears, a bass EQ is a good replacement for the KUBE for bass response. I would expect DSP would be even better for adjusting bass. I don't know if DSP will cause new issues when trying to remove the 4 kHz bump. I guess I will have to try and see.
I'm not sure how to do this physically on the board, but would the correct concept for bi-amping be below?
Yes, that would be a correct point for splitting for biamping. since the two chains are driven from a low impedance amplifier output they don't interact at that point.
I see the conjugate bits, like all the middle bits of the woofer path, are following the initial crossover (the L, L, C on the left). That means that none of the conjugate goodies can be removed without seriously changing the crossover shape.
I see the conjugate bits, like all the middle bits of the woofer path, are following the initial crossover (the L, L, C on the left). That means that none of the conjugate goodies can be removed without seriously changing the crossover shape.
The most important thing to do first is to re-read post #16 https://www.diyaudio.com/community/threads/crossover-instead-of-kube-on-kef.99056/#post-1766135
AndrewJ designed the Kef 107
AndrewJ designed the Kef 107
Regarding phase and changing the head mounting, I wouldn't have thought there would be much difference between a small accidental movement from it's stock position by having it unfixed on a granite slab, and phase differences from turning the head as intended. At the cross-over point I would have thought the wavelength would still be relatively long so small differences in location would be a small phase change.
Then I was wondering if any of that mattered much anyway with the vent firing upward.. ?
A simple mark on the granite and the head should keep the head positioned over where the old swivel mounting would have been.
Then I was wondering if any of that mattered much anyway with the vent firing upward.. ?
A simple mark on the granite and the head should keep the head positioned over where the old swivel mounting would have been.
I'm pretty sure the 330 Hz notch was for a pipe resonance for the woofer port. It was probably out of band but severe enough that it couldn't be overlooked. Per the circuit shown, it is part of the woofer network. Regarding bandwidth, the coupled cavity designs (104.2 and 107) were fairly narrow, less than 2 Octaves to maintain high efficiency.
Bravo Yoder77
Finally someone gave exact answer. This is what I was looking for a while now. Thank You. I didnt want to introduce KUBE into mid/hi xover as that would negate putting Raal instead of T33 ( tbh it wasnt me who's done it but nevertheless it has been done with care and love by someone who's work I respect immensely)
Also, I wanted to go with miniDSP 2x4(HD?) instead of KUBE or with something KUBE-like with some more effort put into it( NadjaDSP or Schiit ). Still deciding which route to take and this info will give me something to think about.
Could someone tell me what is the purpose of 2x100uF caps in parallel in bass xover ( ones that go into coil and then into MF/HF head connections...NOT the two below 4 x resistors) ? All caps in my bass xover are upgraded to new bi-polars but ONLY in this position there are expensive and huge polypropylene caps instead.
Thanks guys!
Finally someone gave exact answer. This is what I was looking for a while now. Thank You. I didnt want to introduce KUBE into mid/hi xover as that would negate putting Raal instead of T33 ( tbh it wasnt me who's done it but nevertheless it has been done with care and love by someone who's work I respect immensely)
Also, I wanted to go with miniDSP 2x4(HD?) instead of KUBE or with something KUBE-like with some more effort put into it( NadjaDSP or Schiit ). Still deciding which route to take and this info will give me something to think about.
Could someone tell me what is the purpose of 2x100uF caps in parallel in bass xover ( ones that go into coil and then into MF/HF head connections...NOT the two below 4 x resistors) ? All caps in my bass xover are upgraded to new bi-polars but ONLY in this position there are expensive and huge polypropylene caps instead.
Thanks guys!
I guess my question was, if this 330Hz is near to the cross-over point, both the bass unit and the mid-range will be reproducing that frequency to some extent. So I was wondering if the notch was applied to both, or to the bass only... It would make sense to be bass only of course, but I wasn't sure.
The reason for asking is to understand what is being passed to the MF/HF head, just a regular high-pass, or a notched high-pass .. and therefore to check any x-over split would be correct.