Here some more graphs about the acoustics behaviour:
The very wide dispersion brings some edge diffraction, but overall balanced and quite linear with good phase behaviour. Dip of 1.5 dB @1.5-3kHz in the room & power response and in vertical polar map is according to the xdir simulation and inherit to multiway systems with common high/mid xover frequencies and driver spacing:
Best regards
Peter
The very wide dispersion brings some edge diffraction, but overall balanced and quite linear with good phase behaviour. Dip of 1.5 dB @1.5-3kHz in the room & power response and in vertical polar map is according to the xdir simulation and inherit to multiway systems with common high/mid xover frequencies and driver spacing:
Best regards
Peter
Last edited:
Well, once the genral concept, drivers and the housing geometry was chosen there is narrow gap for a matching crossover frequency. T25A is specified at 2,2kHz lowest xo frequency, which is confimed by hificompass.com distortion measurements. As shown in #16 the maximum crossover frequency for the midrange is 2,5kHz due to membrane behaviour (this Kartesian sounds magic, but is a bitch...). So it had to be between 2,2-2,5kHz. Some optimization runs with preference ratrings in VirtuixCAD tended to 2,5kHz, so shall it be. Driver spacing? As narrow as possible. That's one of the advantages of "small" drivers...
Other manufacturers deal this compromise, too:
Source: https://de-de.neumann.com/kh-420#technical-data
Nice measurements Peter,
From your graphs it looks like you've optimized for the listening window, whilst trying to keep the horizontal directivity as smooth as practically possible. What was your target slopes?
I think the tradeoffs chosen are very reasonable.
When I prototype using miniDSP, I can switch between 4 different presets with about 1 second delay between them. I'm using very low distortion drivers so no much more than be done there.
But in terms of tonal balance I've tried a few different approaches.
For example, giving more weighting/importance to the Predicted In Room response instead of the On-Axis
Whilst I try to keep the phase coherency with the 3 drivers fairly close (inverted midrange results in notch ~15+dB)
I tried optimizing for PIR, VS On-axis response, vs Listening window
When optimizing for PIR gives a balanced tonality, whatever is special about the speaker kind of disappears.
It seems to lose it strengths. It's hard to describe the effect, but it reminds of Dirac Live. With a excellent speaker (WMTMW, 8/4"/ribbon, 20Hz to 20Khz, 100+dB capable), the benefits of Dirac Live are marginal, and sometimes not not helpful.
I go back to optimizing for reference axis (on-axis)- which I think makes well recorded music great and bad recordings rubbish, but I preference that. Rather than all vanilla ice-cream, all the time.
What have you experienced when optimizing your crossover?
From your graphs it looks like you've optimized for the listening window, whilst trying to keep the horizontal directivity as smooth as practically possible. What was your target slopes?
I think the tradeoffs chosen are very reasonable.
When I prototype using miniDSP, I can switch between 4 different presets with about 1 second delay between them. I'm using very low distortion drivers so no much more than be done there.
But in terms of tonal balance I've tried a few different approaches.
For example, giving more weighting/importance to the Predicted In Room response instead of the On-Axis
Whilst I try to keep the phase coherency with the 3 drivers fairly close (inverted midrange results in notch ~15+dB)
I tried optimizing for PIR, VS On-axis response, vs Listening window
When optimizing for PIR gives a balanced tonality, whatever is special about the speaker kind of disappears.
It seems to lose it strengths. It's hard to describe the effect, but it reminds of Dirac Live. With a excellent speaker (WMTMW, 8/4"/ribbon, 20Hz to 20Khz, 100+dB capable), the benefits of Dirac Live are marginal, and sometimes not not helpful.
I go back to optimizing for reference axis (on-axis)- which I think makes well recorded music great and bad recordings rubbish, but I preference that. Rather than all vanilla ice-cream, all the time.
What have you experienced when optimizing your crossover?
Your designs are always impressive and really detailed. good work. For my part, I'm always thinking adapted your "Neo" design to my DIY speakers (with a 15" instead of 18")
Hi Peter,
It is not that often one can see a kartesian driver and its dedicated mid. Thank you for the nice measurements. Seems perhaps hard for a passive filter at least for my non trained eyes. How do you like it since active smoothed, please, related to the experience you had with others brands?
It is not that often one can see a kartesian driver and its dedicated mid. Thank you for the nice measurements. Seems perhaps hard for a passive filter at least for my non trained eyes. How do you like it since active smoothed, please, related to the experience you had with others brands?
Just a little update:
a) I reworked the enclosure optical appearance:
b) and added a passive network to the midrange and tweeter:
Series resistors for local current feedback (there is much headroom in voltage swing of the FA503 available) and serial notches for the main diaphragm resonances. This should help to decrease the already low current induced distortions and thermal modulation even more.
Was to lazy tilll now to make new measurements with the slightly front baffle shape changes, so this is the actual tuning based on the legacy measurements:
(Power response is from +/-60deg Hor and +/-40deg Ver thas I use as target curve for voicing)
The "sound" of the mid-high unit with the passive network on FA503 is very very neutral, flawless and can reproduce very high dynamics without compression. Last year the the speakers were running very loud in the large, damped home cinema room of german "Klang & Ton" DIY magazine after a 12"/1" Horn tweeter / BL-Horn combo was playing; first comment was "okay that's how it sounds without distortion and resonances"....
c) During the time I am convinced that the woofer cannot keep up with the mid-high unit. It obviously limits the max output, and is a sounding a bit to "soft" regardless how I try to tune it.
So I have something new in my mind where I will reuse the mid + tweeter, passive network and FA503 but replace the WO24P with two fast, low-mass Scan Speak 22W/8534G00 with extra magnets in a new cabin; Project "FastTrack":
Simulations, calculations & construction almost ready, but need some time and money until the build can launch.... (maybe during next year)
Best regards
Peter
a) I reworked the enclosure optical appearance:
b) and added a passive network to the midrange and tweeter:
Series resistors for local current feedback (there is much headroom in voltage swing of the FA503 available) and serial notches for the main diaphragm resonances. This should help to decrease the already low current induced distortions and thermal modulation even more.
Was to lazy tilll now to make new measurements with the slightly front baffle shape changes, so this is the actual tuning based on the legacy measurements:
(Power response is from +/-60deg Hor and +/-40deg Ver thas I use as target curve for voicing)
The "sound" of the mid-high unit with the passive network on FA503 is very very neutral, flawless and can reproduce very high dynamics without compression. Last year the the speakers were running very loud in the large, damped home cinema room of german "Klang & Ton" DIY magazine after a 12"/1" Horn tweeter / BL-Horn combo was playing; first comment was "okay that's how it sounds without distortion and resonances"....
c) During the time I am convinced that the woofer cannot keep up with the mid-high unit. It obviously limits the max output, and is a sounding a bit to "soft" regardless how I try to tune it.
So I have something new in my mind where I will reuse the mid + tweeter, passive network and FA503 but replace the WO24P with two fast, low-mass Scan Speak 22W/8534G00 with extra magnets in a new cabin; Project "FastTrack":
Simulations, calculations & construction almost ready, but need some time and money until the build can launch.... (maybe during next year)
Best regards
Peter
Last edited:
Hard to describe and separate the sound of a driver without the system context and overall voicing. The mid-hi unit is very flawless and neutral, inconspicuous in the best sense of the word, not warm, not cold and capable of reproducing very high dynamic steps and volume. Like it very much and cannot imagine a better driver combination >300Hz and feel no desire to buy something "better".
Had a nice weeked with two very experienced DIYers some time ago, these were their comments after listening:
https://www-diy--hifi--forum-eu.tra...r_tl=de&_x_tr_hl=de&_x_tr_pto=wapp#post334328
have been reading it. Nice review.
What about a single 12" (12RS430) vs the two 22W/8534G00 ? You like more the neutrality of glass fiber from SS and it blends better with the Kartesian cone material ?
What about a single 12" (12RS430) vs the two 22W/8534G00 ? You like more the neutrality of glass fiber from SS and it blends better with the Kartesian cone material ?
The 12RS430 is a very nice and also affordable SOTA 12" woofer; I had actually already considered replacing the WO24P with an 10RS430 in the existing enclosures - pretty tight match, but should just barely fit.
But...
The concept "Half-Space-Radiator with Zwicker-Characteristics as compromise for finite baffle sizes" as target curve for the dispersion is proven for me (see https://www.diyaudio.com/community/...-mid120_vhe-wo24p-4-fa503.385504/post-7003114 for further explanation), so I wanted to keep that by the baffle geometry (not possible with a 12" on the front):
I also want to narrow a bit the dispersion in the vertical domain by the use of 2 woofers - that helps below the usual baffle step to keep the power response more linear and so the sound more clear. I also think the "soft" signature I complained about does not really come from the SB woofer, but from the overall small baffle that makes it hard to compromise the direct vs. the indirect sound.
I have good experience with this alignment, and also with the larger 26W/8534G00 which I really love and have used now a few times. Best woofers I had so far (especially with two per side), absolutely nothing to complain about them:
So I have already good experience with the "bigger brothers". The 8" type 22W/8534G00 have also good reputation, the GF membran material should pretty good match the neutral signature of the other drivers and - I really like the very low Mms of just 22g for pairing them to the also very, very low moving mass of mid and tweeter membranes. (Yes I know there is no scientific evidence, I just like when things fit together 😉)
In the speakers above (by the way, they are called "Scandinavian Connection MkVII"), the woofers got extra magnets Y35 115(50)x20 that whopped the BL by >9%:
So, with that experience I aim for a BL boost in 6-7% with Y35 90(35)x20 extra magnets that may give a resulting Qts = 0,75 in ~34l CB for 2 woofers. But will test that before finalizing the drawings and ordering the wood cut.
Best regards
Peter
But...
- I wanna have floorstanders again. The sideboard the speakers are standing on shows resonances, and the nearby located TV screen introduces reflections of the very widely dispersed soundfield.
- I wanna have "slim" floorstanders and no monkey coffins now for looks.
The concept "Half-Space-Radiator with Zwicker-Characteristics as compromise for finite baffle sizes" as target curve for the dispersion is proven for me (see https://www.diyaudio.com/community/...-mid120_vhe-wo24p-4-fa503.385504/post-7003114 for further explanation), so I wanted to keep that by the baffle geometry (not possible with a 12" on the front):
I also want to narrow a bit the dispersion in the vertical domain by the use of 2 woofers - that helps below the usual baffle step to keep the power response more linear and so the sound more clear. I also think the "soft" signature I complained about does not really come from the SB woofer, but from the overall small baffle that makes it hard to compromise the direct vs. the indirect sound.
I have good experience with this alignment, and also with the larger 26W/8534G00 which I really love and have used now a few times. Best woofers I had so far (especially with two per side), absolutely nothing to complain about them:
So I have already good experience with the "bigger brothers". The 8" type 22W/8534G00 have also good reputation, the GF membran material should pretty good match the neutral signature of the other drivers and - I really like the very low Mms of just 22g for pairing them to the also very, very low moving mass of mid and tweeter membranes. (Yes I know there is no scientific evidence, I just like when things fit together 😉)
In the speakers above (by the way, they are called "Scandinavian Connection MkVII"), the woofers got extra magnets Y35 115(50)x20 that whopped the BL by >9%:
So, with that experience I aim for a BL boost in 6-7% with Y35 90(35)x20 extra magnets that may give a resulting Qts = 0,75 in ~34l CB for 2 woofers. But will test that before finalizing the drawings and ordering the wood cut.
Best regards
Peter
34 L for 2 x 8" ! Nice indeed. Two 10RS430 would need 77L if I remember the thread about the new Faitals (12RS430, etc).
Peter, have you ever had the chance to compare the Mid120 to the Wom120 - either simulated or in real use? I've considered both for use between my T25Bs and Purifi PTT6.5s, though at the moment, I'm leaning toward either nothing between them (perhaps possible, but definitely a stretch, I suspect) or the Satori MW13TX. Any other mids you'd consider? Would appreciate your thoughts.
Hi Paul,
regarding the Mid120 vs. Wom120:
Clement from Cartesian did some comparison of the construction properties in this video:
(eventually, use automatic subtitles+translation to english)
Simple max. peak SPL simulation in Basta! with manufacurers data shows that Mid120 gives ~5dB more SPL, as Mid120 is more efficient due to stronger magnetic field and lighter moving mass, and can withstand more el. power due to larger voicecoil:
.:
Mid has fabric surround and very flat cone, Wom has rubber surround and a bit deeper cone.
For me - according to the construction properties, manufacturers data, measurements - Wom120 is a good midwoofer, which may be also used as a nice midrange for its money. But Mid120 is a dedicated beast of a small midrange and in my opinion one of the best 4" cone mids you can buy (together with the Scan Speak 12MU Illuminator which is more expensive but forseeable easier and flatter to XO to the tweeter).
Regarding the Satori MW13TX:
It is a size larger and a midwoofer, which may have the advantage to be crossed much lower. The larger diameter and deeper cone leads to more beaming in the midrange in comparison to the very wide radiating Mid120.
The wide radiation of the Mid120 matches well with the also very wide radiating Bliesma T25x tweeters. I like that.
Your Bliesma T25B paired with a MW13TX may result in a step in the dispersion pattern; a waveguide may help there.
So, my subjective advice:
But you say you use a Purifi PTT6.5 at the moment, and want to add a midrange between them? With that Purifi available, I don't see the MW13TX will give you any significant advantage here, maybe a little bit changed sonic signature and very few decreased intermodulation distortion in the fundamentals. Use a dedicated small 4" mid to change the overall system radiation pattern to "very wide", if you want that. Wom120 may be sufficient regarding max. SPL, as it is anyway limited by the "small" 6,5" woofer. Or put the Bliesma in a 5"-6" Somasonus waveguide to achieve a narrower, but matched constant directivity combined with just the Purifi.
Best regards
Peter
regarding the Mid120 vs. Wom120:
Clement from Cartesian did some comparison of the construction properties in this video:
Simple max. peak SPL simulation in Basta! with manufacurers data shows that Mid120 gives ~5dB more SPL, as Mid120 is more efficient due to stronger magnetic field and lighter moving mass, and can withstand more el. power due to larger voicecoil:
.:
Mid has fabric surround and very flat cone, Wom has rubber surround and a bit deeper cone.
For me - according to the construction properties, manufacturers data, measurements - Wom120 is a good midwoofer, which may be also used as a nice midrange for its money. But Mid120 is a dedicated beast of a small midrange and in my opinion one of the best 4" cone mids you can buy (together with the Scan Speak 12MU Illuminator which is more expensive but forseeable easier and flatter to XO to the tweeter).
Regarding the Satori MW13TX:
It is a size larger and a midwoofer, which may have the advantage to be crossed much lower. The larger diameter and deeper cone leads to more beaming in the midrange in comparison to the very wide radiating Mid120.
The wide radiation of the Mid120 matches well with the also very wide radiating Bliesma T25x tweeters. I like that.
Your Bliesma T25B paired with a MW13TX may result in a step in the dispersion pattern; a waveguide may help there.
So, my subjective advice:
- T25B + Mid120 will match well, will give a very wide half-space radiation and give you a SOTA MidHi-Unit without much compromises. Mid120 is hard to filter on the upper end.
- T25B + Wom120 will match well, will give a wide half-space radiation and give you a very good MidHi-Unit with ~ 3-4dB less sensitivity and ~ -5dB less max SPL than the combination with Mid120, with eventually neglectable loss in resolution/transparency. Wom120 seems easier to filter and is half the price.
- T25B + MW13TX may have a step in the radiation pattern, so I may suggest to use a 5" Somasonus waveguide for that combination. With that measure, it will give you also a SOTA MidHi-Unit without much compromises that will match very well to a narrower, but constant directivity radiation pattern, with the same or even a bit more max. SPL than the Mid120 combination.
But you say you use a Purifi PTT6.5 at the moment, and want to add a midrange between them? With that Purifi available, I don't see the MW13TX will give you any significant advantage here, maybe a little bit changed sonic signature and very few decreased intermodulation distortion in the fundamentals. Use a dedicated small 4" mid to change the overall system radiation pattern to "very wide", if you want that. Wom120 may be sufficient regarding max. SPL, as it is anyway limited by the "small" 6,5" woofer. Or put the Bliesma in a 5"-6" Somasonus waveguide to achieve a narrower, but matched constant directivity combined with just the Purifi.
Best regards
Peter
Last edited:
The WOM144 VHE looks also impressive. while their best as a pure mid seems the one you use.
Last edited:
Thank you, Peter, for the detailed response. I'd really like to avoid putting the T25Bs into any sort of waveguide, so I think I may pick up a pair of Mid120s.
Hi Paul, just some hints from experience:
Alternative 4" mids to the Mid120_vHE might also be:
Best regards
Peter
- be careful with with driver placing on the baffle and baffle/edge shape; due to the wide dispersion the tweeter/mid combo is quiet sensitive to this. Use e.g. Vituixcad baffle diffraction simulation to find a good matching geometry.
- For the XO frequency I would aim for 2,2 - 2,5kHz. Tweeter cannot go much lower, and mid not much higher. Absolute maximum range is between 2 to 3 kHz.
- If you have an active/DSP crossover the choice is fine. If you plan to filter passively, I would rethink the choice - personally, i did not want to handle the Mid120 with a passive filter (and i've build some passive 3-ways).
Alternative 4" mids to the Mid120_vHE might also be:
- Scan Speak 12MU (more expensive, but easier to lowpass/equalize. I would prefers this for a passive filter and when the budget is no issue)
- the new Purifi PTT4.0Mxx midranges (might match your existing 6.5 Purifi quiet well. Also more expensive. Did not have a deeper look on them till now)
- Tymphany NE123W (Very high reputation. Copper cap, large titanium voicecoil. Getting hardly available. Cheaper than Mid120. Personally, I do not like the surround resonance at ~1kHz and overall breakup behavior)
- Already mentioned WOM120 as a budget recommendation.
- My personally favored budget recommendation is a SB12MNRX2-25-4, that seems to be nearly flawless and models very well.
Best regards
Peter
Hi peter,
Ma I ask whu you chose 4" over 5" or 6" ?
My limited experience is I enjoyed a lot 5" crossed between 100 and 150 hz (domestic level loudspeaker), but I wonder if a 12"' cerossed below 300 hz and two 5.5" or a 6" for polart response blend will a good target ?
Of courser a 12" is asking a 36 cm widtj cabinet, too much large perhaps according your philsopjy ?
Ma I ask whu you chose 4" over 5" or 6" ?
My limited experience is I enjoyed a lot 5" crossed between 100 and 150 hz (domestic level loudspeaker), but I wonder if a 12"' cerossed below 300 hz and two 5.5" or a 6" for polart response blend will a good target ?
Of courser a 12" is asking a 36 cm widtj cabinet, too much large perhaps according your philsopjy ?
Thanks for the advice, Peter. I do plan to learn and use Vituixcad. I'd prefer a crossover at 3k, given the HD of the T25B, but 2.5k will probably be fine. I will be using a MiniDSP Flex8 for the crossovers. I might add a passive series notch to the mid as per this: Low Distortion Filter for PTT6.5X04-NAA.
The PTT4.0 midranges do look very good - much better break-up behavior than the Mid120, it seems to me. Both a higher initial peak and only one rather than multiple. Maybe they are what I should be going with. Decisions, decisions.
A very inexpensive option I've considered is the Tectonic TEBM46C20N-4B. The dispersion looks great up to 3k or beyond, but I'd need a higher crossover to the PTT6.5 (600 hz?). That might be perfectly fine though. At its price, I might give it a try.
Hi, I wanted to have a very wide radiating system, and this is primary definded by the driver sizes. And a tiny 3-way, so everything got a bit smaller.
But I like 5" mids and midwoofers a lot, also my favorite size in general.
For a 12" playing up to 300Hz a single 5" as mid is totally sufficient, if it can stand some electrical power. Best example is the JBL LSR 6332. A bit higher xover frequency keeps excursion - and so intermodulation distortion - away from the midrange and the sound in the fundamentals is pushed with authory by the driver with the most membrane area. Also group delay gets more uncritical with the higher xover frequency. But when crossed over too high, lobing in the vertical domain gives a hole in the power response and I also don't like it when the midrange bandwidth gets too narrow. My rule of thumb for the xover-frequencies of a 3-way is 200-300Hz and 2-3kHz, so the work between the drivers is nicely shared regarding many aspects.
Some other like the "2-Way + Sub" approach to keep the sonic characteristics of a compact 2-way. That's totally okay, but you lose some headroom, cannot take a "specialized" midrange, and the woofer lowpass gets hard to be realized with a passive xover.
My "philosophy" regarding the baffle geometry is not holy for me and just one of many possible approaches, everything is a tradeoff. A generously large baffle has other advantages, e.g. keeps diffraction effects away from the mid-highs, and the baffle step very low that gives a clear and neutral sound in the fundamentals, but anyway and even more so sounds "big".
Wanna have it really "Loud And Clear" 😉 like the LSR6336 for a fair budget?
https://faitalpro.com/en/products/LF_Loudspeakers/product_details/index.php?id=151050115
https://faitalpro.com/en/products/LF_Loudspeakers/product_details/index.php?id=101010100
http://www.wavecor.com/html/tw030wa11_12.html
But if you need moderate levels and prefer the two-way-compact speaker sound from a small baffle, maybe just build two good active subwoofers as supplement.
Last edited:
Okay great you may have many options with the Flex8 to handle also beastly mids!
The Tectronics is special by it's BMR concept that may have advantages as a fullrange. The distortion behavior looks not very convincing to me, in comparison to the other options you think about. If you're just curious about it and you want to play around with some prototypes - why not, try it.
But: you've already spend some serious money on very nice drivers and DSP electronics. To be honest: Will you ever be satisfied by looking on that cheap plastic thing with the ugly shape between your SOTA Purifi and the Beryllium tweeter? Or will it give a smile on you face when you see a nice matching 4" Purify mid in that place, or the exotic Kartesian tamed with your Flex8 and know that it cannot get better, eventually only a bit different? 😉