I had an exercise on speaker impedance measurements today using REW, UMC204HD, and a tpa6120A2 based headphone amp. All drivers measured in their enclosures, A and B indicate different speakers/enclosures.
First the woofers, so two sb23MFCL45-4 drivers connected in series. Seems to match the VituixCAD simulation quite well.
Next, the mids, which are the woofer cones of the KEF R3 drivers in their conical enclosures. There's some difference in the resonance frequency between the drivers. I'm sure the enclosure volumes are the same, and the damping materials were carefully weighed so, perhaps just driver fabrication tolerances since the difference in Fs is just 10%?
My flares fit quite tightly around the drivers thought that they might affect the stiffness of the surrounds. So I decided to measure the mids again with and without the flares. There's a slight shift of resonance frequency for both drivers when the flare is removed. Is this indeed a change in the mechanical properties of the surround or could the flare change the loading of the driver enough to cause such change?
Finally, the tweeters. They were measured with series capacitors connected.
First the woofers, so two sb23MFCL45-4 drivers connected in series. Seems to match the VituixCAD simulation quite well.
Next, the mids, which are the woofer cones of the KEF R3 drivers in their conical enclosures. There's some difference in the resonance frequency between the drivers. I'm sure the enclosure volumes are the same, and the damping materials were carefully weighed so, perhaps just driver fabrication tolerances since the difference in Fs is just 10%?
My flares fit quite tightly around the drivers thought that they might affect the stiffness of the surrounds. So I decided to measure the mids again with and without the flares. There's a slight shift of resonance frequency for both drivers when the flare is removed. Is this indeed a change in the mechanical properties of the surround or could the flare change the loading of the driver enough to cause such change?
Finally, the tweeters. They were measured with series capacitors connected.
Attachments
Last edited:
This is pretty normal, an example of 25 TC9 drivers and their Fs peak spread. As the suspensions break in this will change anyway.
Today was a good day! I got to spend some time in a cozy room which enabled anechoic measurements down to 100 Hz. No big surprises in the results. Just a better resolution throughout the spectrum. Should be good starting point for the XO design.
Attachments
Here are the distortion plots. Unfortunately I don't know the exact SPL level. A smart phone app showed 85 dB at 1 m, but it was clearly uncomfortably loud even without room reflections. I'm thinking XOs at around 350-400 Hz and 2.7-2.9 kHz..
Attachments
I‘ve spent a little over 2 years with the 150’s as one of three nearfield monitors in my studio….it’s ideally suited to the task…..but I’ve tried three times to bring them into my listening/casual space and they‘re just not good….too polite and in the midfield the imaging is meh at best. Its not a power compression thing……they’re supported on the bottom with 4 15” sealed subs which have reigned supreme in this space for over 5 years now……….after a 1 meter triangle, well…..nothing improves.…..my Totem slim two way towers outclass the KEF on every front.
A bit of tentative XO sketching with the morning cup of coffee. This is LR4 at 350Hz and 2800Hz. There's still that widening in the vertical response at 900Hz which digs a dip in the directivity and in the axial response (latter in favor of the power response). The additional capacity, symmetry, and LF vertical directivity provided by the dual 8" woofers in the woofer-coax-woofer configuration comes with this price in my execution, but I think it's worth paying.
Remains to be seen how I will like it, but I think there are things to be gained with these coaxial drivers when moving from bookshelfs+sub(s) to true three-ways which allow crossing at higher frequencies.
My (secret) goal was to finish these by Christmas. Looks like it will more or less happen despite having a very busy end of the year. The aesthetics are pretty much final although I may paint the 3D-printed parts again next summer with a graphite gray color which matches better with rubber surrounds of the drivers and makes less contrast with the enclosure. The color tone of the enclosures is something "claystone", which is neutral enough the blend these relatively large speakers in our living room quite well. I also tried to make the surface as matte as possible.
But most importantly, I'm finally listening music through these beasts! There's still some voicing to be done and I haven't applied any room correction yet, but I like the sound already. I'm coming from a 3-way with two soft domes and a 6.5" woofer in vented enclosure, so by nature these are completely different animals. The bass sounds accurate and effortless and the sound stage is very nice. I'd say there are no major flaws that I can hear, but some finetuning is still required before the final verdict.
Sorry about the dull picture. Solstice is not a good time for natural light photography in Finland.
But most importantly, I'm finally listening music through these beasts! There's still some voicing to be done and I haven't applied any room correction yet, but I like the sound already. I'm coming from a 3-way with two soft domes and a 6.5" woofer in vented enclosure, so by nature these are completely different animals. The bass sounds accurate and effortless and the sound stage is very nice. I'd say there are no major flaws that I can hear, but some finetuning is still required before the final verdict.
Sorry about the dull picture. Solstice is not a good time for natural light photography in Finland.
You're happy man!
Speakers looks great and I wish you many hours of fine and non-disturbing music listening.
My year was also fully loaded and we completed it by moving to our new house on 16.12. I plugged in the Kef iQ7SE Custom yesterday (active filters via miniDSP and a modified speaker cabinet-inside) and I don't think I can do much more for this year.
Speakers looks great and I wish you many hours of fine and non-disturbing music listening.
My year was also fully loaded and we completed it by moving to our new house on 16.12. I plugged in the Kef iQ7SE Custom yesterday (active filters via miniDSP and a modified speaker cabinet-inside) and I don't think I can do much more for this year.
Thank you @Martin87 I wish the same for you. Nice job customizing iQ7SE.You're happy man!
Speakers looks great and I wish you many hours of fine and non-disturbing music listening.
My year was also fully loaded and we completed it by moving to our new house on 16.12. I plugged in the Kef iQ7SE Custom yesterday (active filters via miniDSP and a modified speaker cabinet-inside) and I don't think I can do much more for this year.
I'm just testing a DSP setup which includes a basic correction for bass and some alternative HF shelves for toning the balance. I think I'm now at the point which is good enough for being able to relax during the holidays. All in all pretty satisfied with the outcome, particularly since this was my first take on loudspeaker design.
In room measurements to be posted sooner or later...
I should probably do something for the gain structure, which was kind of expected. I'm feeding minidsp 4x10 either digitally from the computer (toslink) or from a tube phono stage. Both channels are set so that I have 10dB headroom in the dsp input monitors for the peak levels. With this, the digital volume control ends up being at around -40db for casual listening. I could easily attenuate by 10-20dB without ever running out of headroom. I have already set the output gain jumpers in the minidsp board to the low setting (balanced), so the options are either to
1. add passive attenuation between minidsp 4x10 (1.12kohm Zout) and the TPA3255 power amp boards (20kohm Zin) or
2. adjust the input buffer gain of the TPA3255 boards.
Option 2 would be technically better, not not very convenient to experiment with since it involves replacing smd resistors on multiple boards. Option 1 on the other hand could be easily added as a fixed-level inline attenuator stage which can be easily added and removed for testing. It could be a simply a resitor board attached to the Phoenex connector at the minidsp output. I think I could make it quite compact using a line raster board and smd resistors, and a pcb-mount Phoenix connector.
Alternatively, attenuator boards could be added individually for each amp board at their input JST connectors. This would require a bit more work for installation/removal, but closer to the amp board would be better in the sense that I wouldn't have to worry about the cable capacitance. I have fairly short cables, though, since the amps are right underneath the minidsp.
Something to think about, but for now I'll just enjoy listening music instead of frequency sweeps and periodic pink noise
1. add passive attenuation between minidsp 4x10 (1.12kohm Zout) and the TPA3255 power amp boards (20kohm Zin) or
2. adjust the input buffer gain of the TPA3255 boards.
Option 2 would be technically better, not not very convenient to experiment with since it involves replacing smd resistors on multiple boards. Option 1 on the other hand could be easily added as a fixed-level inline attenuator stage which can be easily added and removed for testing. It could be a simply a resitor board attached to the Phoenex connector at the minidsp output. I think I could make it quite compact using a line raster board and smd resistors, and a pcb-mount Phoenix connector.
Alternatively, attenuator boards could be added individually for each amp board at their input JST connectors. This would require a bit more work for installation/removal, but closer to the amp board would be better in the sense that I wouldn't have to worry about the cable capacitance. I have fairly short cables, though, since the amps are right underneath the minidsp.
Something to think about, but for now I'll just enjoy listening music instead of frequency sweeps and periodic pink noise
Hi and total respect to the wonderful work presented here.🙏.
Incredible woodworking MrSticha.
A bit off topic but are these guys (Setelec) a legit KEF supplier?
I only ask because I have a pair of SP1632's that I picked up cheap but their tweet's are blown and Setelec's prices don't look terrible. Can't find the damn tweeters anywhere.
I got the tweeters out of my coaxial units but further disassembly would likely be catastrophic.
If I could replace them it would definitely make me re-think a near future project.
Many thanks and Merry Christmas to all (or whatever floats your boat).
Matt
Incredible woodworking MrSticha.
A bit off topic but are these guys (Setelec) a legit KEF supplier?
I only ask because I have a pair of SP1632's that I picked up cheap but their tweet's are blown and Setelec's prices don't look terrible. Can't find the damn tweeters anywhere.
I got the tweeters out of my coaxial units but further disassembly would likely be catastrophic.
If I could replace them it would definitely make me re-think a near future project.
Many thanks and Merry Christmas to all (or whatever floats your boat).
Matt
Hi Matt. Thank you for kind words. Merry Christmas to you too.
Setelec was recommended to me by another forum member, but beyond that I can only talk based on my own experience. Initially I had my doubts about the webstore but in the end I have no complaints. I took some months but I knew that's likely the case for items they don't have in stock. The drivers arrived well packed and seem to work nicely.
Why not simply buy a replacement part from a local hi-fi shop that sells KEF speakers?
I didn't mean a parts shop but a hi-fi shop that is authorised to sell and support complete KEF speakers. KEF has plenty of replacement parts for current and recent models which they will supply to the shops that sell and support their speakers. This is how I have bought KEF parts in the past for both speakers I own and speakers I don't.
If I recall correctly there was an issue a few years back in the US when replacement coaxial drivers were being bought by not only the odd DIYer from the US equivalent of the French shop you cite (was that Speaker Exchange?) but at least one speaker manufacturer was starting to do likewise. Not surprising given the price and performance compared to what was available on the DIY market. They got the US end to tighten up but I am not sure this was done elsewhere. In the UK my local KEF hi-fi shop had various new parts knocking around to show customers, parts for custom installations, second hand parts from warranty work (e.g. subwoofer drivers due to failed electronics).
If I recall correctly there was an issue a few years back in the US when replacement coaxial drivers were being bought by not only the odd DIYer from the US equivalent of the French shop you cite (was that Speaker Exchange?) but at least one speaker manufacturer was starting to do likewise. Not surprising given the price and performance compared to what was available on the DIY market. They got the US end to tighten up but I am not sure this was done elsewhere. In the UK my local KEF hi-fi shop had various new parts knocking around to show customers, parts for custom installations, second hand parts from warranty work (e.g. subwoofer drivers due to failed electronics).