Thanks for that, good to see acurate factory datasheet with burned in drivers.
I saw your #240 post about cold and warm TSP, I assume they are close enough to choose either for modelling then according the box inside modes (your famous 250 Hz mode) and subjective bass response to tune in real life with port length. I wanted initially to use it as a midwoof sealed (vented too much complex for me)) but it will lacks too much low end with something between a Bessel and Butterworth. Hence your thread I read for inspiration for a vented with a close behavior to try to avoid bass sub and not waste the group delay.
I saw your #240 post about cold and warm TSP, I assume they are close enough to choose either for modelling then according the box inside modes (your famous 250 Hz mode) and subjective bass response to tune in real life with port length. I wanted initially to use it as a midwoof sealed (vented too much complex for me)) but it will lacks too much low end with something between a Bessel and Butterworth. Hence your thread I read for inspiration for a vented with a close behavior to try to avoid bass sub and not waste the group delay.
I can't imagine needing a sub. Room dependent I'm sure. Plenty of bass for us here—really depends on the record though. I can get waves of bass with the right recording and certain listening positions—bathing in the bass! DAC hi-res is plenty strong too. Never occurs to me it's lacking... Everyone's different though! I haven't had a moment of concern having these speaks in our life...I'd imagine that the tower version has even more extension—haven't heard them however...
It is about what music style you listen to most and your room gain. Here we are with the choice and that is not that easy.
Difficult to know before having experienced it in your listen room. Just because room gain and each people preference differ, not saying many todays digital sources are a little boomy flat in the lows to my tastes.
@motokok , my understanding is Mathias and Paul chose the most extended low which means a faster dB roll off when going to the lows. Look at some usefull numbers : - F3, F6, F10, that are usefull if you can interpret it in relation to your room. The goal being subjective tigther and "faster" bass by a better group delay behavior that copes better with the upper drivers as they are mandatories to feel that transcient, snapp, fast, whatever you call it (the sacro saint acoustical flat phase you have to deal with the passive filter and box load to say it short while it seems to matters most in the lows bass to mid than mid to treble)
My very basic understanding taking raw datasheet is this is more or less close to a Bessel acoustic high pass, which is the nearest to a sealed although different. The vented will have lower F3 and higher spl in the beginning of the roll off VS a sealed with a good group delay tunning but less spl from 100 hz to 50 hz circq, needs large box and Qtc around 0.507 (Bessel).
0.707 (50 L sealed) being the maximum flat says the book at the price of a sligthy less good group delay behavior VS the 0,.507 sealed Qtc, although some say they do not notice a difference listsning to both.
And here we stand as said Kofifarm member, choose your camp :
OSMC : near Bessel (best BR group delay, smoothest roll off with the max low end but with less SPL when going to the lows, hence -F3 is higher : se qbove your room gain and what music you listen to, jazz and rock perhaps not the best choice here, fact is Mathiass need to lift up a little the spl at the price of a steeper roll off to be happier related to room/personal taste.
OSMT (open source monkey tree, sorry but a free monkey is climbing a tree not a tower) : Koffifarm load choice. My understanding : Legendre BR load which is theorically with datasheet spec the said 107 L with Fb at 34.7 Hz. The Legendre is chosen when you need the max low - F3 but at the price of a less max low end extension which Paul's sim shown clearly. For a reason Koffifarm chose a sligthy higher Fb at 38 hz, so ot a pure Legendre, perhaps because to stay near the pqssive filter development for the ease of the novice diyers like Don in the OSMT thread which is a short and great thread thread thqt helped your servitor to u derstand the trade offs of all the brillant developpers and helpers.
If choosing the best sealed for the woof in time domain for the group delay, which I hesitate, -f3 is according the numerous TSP, cold, warm, etc, of the thread and also with the different TS of Paul's sim (factory's or Mathias's ?)around 75 hz. Yes you need a Sub with a low pass equal to the high pass sealed load you chose deterlined by the box volume. For instance, maximum flat is Qtc 0.707 butterworth = sealed 50 L, sligthy less good behavior than transcient perfect. TBL member soon in the ref thread said timing behavior being eaten by the room cause all the bounce delays is you do not care. So chose the volume according the bass extension you empirically prefers if I shortcut and sumarize what I understood.
As said Pharell just above, OSMC is what you just needs and it is cool as theorically it is the best group delay if it matters (subject not cleary solved) according the TSP : 67,5 L with Fb at 38 hz or closer to 76 L acxording the dead of the thread, the Pdf or what ypu measure cold, warm, your own unit with factory spec dispersion : Mathias is confide t with Factory specs which is ha dy to seem for noobs like I that miss the experience and brain.
My bet today but can evolves: build bigger like Koifarm 101L and then if prefer 67,7 L to 76 L, put bricks or sand bag to know according your room if reducing the volume load is needed. Length ports are more or less easy to adapt, cqbinet size not.With the back panel that is removable you can swap from BR to sealed if the ports are in the back.
From my seat and wallet, the BR is what I lurk and brillant OSMC thread helped me a lot to understand more as a basic sealed bass loudspeaker enthusiast. We all experienced awfull boomy BR and the road map of Mathias does figth this. (imho the theorical mismatch between him and Koifarm but non critical at the end).
Now what just worry mee as an unexperienced diyer is the resonant mode in the port the discuss highligthed when increasing the biggest internal length of the cabinet : the 250 hz discuss.
I would like to facet a little à la Joachim Gherard Suesskind flagship loudspeaker to reduce or distributes the spl diffraction peaks doing so (my basic understanding at least), so very off topic Mathias's road map, but near Zwu member standalone cabinet for the 12PR320 as a study, smalled width at 33,5 cm, so really not cope to the thread but thanks to the brillant posts I could understand more what I want and thanks for that Mathias, Paul, and contributors
. I do not know if I really understood it but this is how I sumarize to try to make a choice.
Difficult to know before having experienced it in your listen room. Just because room gain and each people preference differ, not saying many todays digital sources are a little boomy flat in the lows to my tastes.
@motokok , my understanding is Mathias and Paul chose the most extended low which means a faster dB roll off when going to the lows. Look at some usefull numbers : - F3, F6, F10, that are usefull if you can interpret it in relation to your room. The goal being subjective tigther and "faster" bass by a better group delay behavior that copes better with the upper drivers as they are mandatories to feel that transcient, snapp, fast, whatever you call it (the sacro saint acoustical flat phase you have to deal with the passive filter and box load to say it short while it seems to matters most in the lows bass to mid than mid to treble)
My very basic understanding taking raw datasheet is this is more or less close to a Bessel acoustic high pass, which is the nearest to a sealed although different. The vented will have lower F3 and higher spl in the beginning of the roll off VS a sealed with a good group delay tunning but less spl from 100 hz to 50 hz circq, needs large box and Qtc around 0.507 (Bessel).
0.707 (50 L sealed) being the maximum flat says the book at the price of a sligthy less good group delay behavior VS the 0,.507 sealed Qtc, although some say they do not notice a difference listsning to both.
And here we stand as said Kofifarm member, choose your camp :
OSMC : near Bessel (best BR group delay, smoothest roll off with the max low end but with less SPL when going to the lows, hence -F3 is higher : se qbove your room gain and what music you listen to, jazz and rock perhaps not the best choice here, fact is Mathiass need to lift up a little the spl at the price of a steeper roll off to be happier related to room/personal taste.
OSMT (open source monkey tree, sorry but a free monkey is climbing a tree not a tower) : Koffifarm load choice. My understanding : Legendre BR load which is theorically with datasheet spec the said 107 L with Fb at 34.7 Hz. The Legendre is chosen when you need the max low - F3 but at the price of a less max low end extension which Paul's sim shown clearly. For a reason Koffifarm chose a sligthy higher Fb at 38 hz, so ot a pure Legendre, perhaps because to stay near the pqssive filter development for the ease of the novice diyers like Don in the OSMT thread which is a short and great thread thread thqt helped your servitor to u derstand the trade offs of all the brillant developpers and helpers.
If choosing the best sealed for the woof in time domain for the group delay, which I hesitate, -f3 is according the numerous TSP, cold, warm, etc, of the thread and also with the different TS of Paul's sim (factory's or Mathias's ?)around 75 hz. Yes you need a Sub with a low pass equal to the high pass sealed load you chose deterlined by the box volume. For instance, maximum flat is Qtc 0.707 butterworth = sealed 50 L, sligthy less good behavior than transcient perfect. TBL member soon in the ref thread said timing behavior being eaten by the room cause all the bounce delays is you do not care. So chose the volume according the bass extension you empirically prefers if I shortcut and sumarize what I understood.
As said Pharell just above, OSMC is what you just needs and it is cool as theorically it is the best group delay if it matters (subject not cleary solved) according the TSP : 67,5 L with Fb at 38 hz or closer to 76 L acxording the dead of the thread, the Pdf or what ypu measure cold, warm, your own unit with factory spec dispersion : Mathias is confide t with Factory specs which is ha dy to seem for noobs like I that miss the experience and brain.
My bet today but can evolves: build bigger like Koifarm 101L and then if prefer 67,7 L to 76 L, put bricks or sand bag to know according your room if reducing the volume load is needed. Length ports are more or less easy to adapt, cqbinet size not.With the back panel that is removable you can swap from BR to sealed if the ports are in the back.
From my seat and wallet, the BR is what I lurk and brillant OSMC thread helped me a lot to understand more as a basic sealed bass loudspeaker enthusiast. We all experienced awfull boomy BR and the road map of Mathias does figth this. (imho the theorical mismatch between him and Koifarm but non critical at the end).
Now what just worry mee as an unexperienced diyer is the resonant mode in the port the discuss highligthed when increasing the biggest internal length of the cabinet : the 250 hz discuss.
I would like to facet a little à la Joachim Gherard Suesskind flagship loudspeaker to reduce or distributes the spl diffraction peaks doing so (my basic understanding at least), so very off topic Mathias's road map, but near Zwu member standalone cabinet for the 12PR320 as a study, smalled width at 33,5 cm, so really not cope to the thread but thanks to the brillant posts I could understand more what I want and thanks for that Mathias, Paul, and contributors
. I do not know if I really understood it but this is how I sumarize to try to make a choice.
Last edited:
Don't get too excited over all the theoretical bass response functions and the corresponding names of famous mathematicians. I don't think Mr. Bessel, Mr. Butterworth, or Mr. Legendre were interested in high-end HiFi loudspeakers.
What we did with the OSMC and the OSMT was to tune the bass empirically using LEAP and VituixCad to calculate the bass response curves (both SPL and impedance). At any given box volume, one can tweak the shape of the response curve by choosing different port geometries (port length). I personally like a smooth response (i.e., low group delay). But hey, it's easy to try different port lengths and find out how you like it!
Regarding subwoofers, my active subwoofer was still there from my previous speakers when I set up my OSMCs in my listening room. I gave the subs a try, and yes, it's fun to shake the room at infrasonic frequencies. However, I soon turned the subs off and don't miss them at all (I'd only have to flip a switch, but that never happened since then...).
What we did with the OSMC and the OSMT was to tune the bass empirically using LEAP and VituixCad to calculate the bass response curves (both SPL and impedance). At any given box volume, one can tweak the shape of the response curve by choosing different port geometries (port length). I personally like a smooth response (i.e., low group delay). But hey, it's easy to try different port lengths and find out how you like it!
Regarding subwoofers, my active subwoofer was still there from my previous speakers when I set up my OSMCs in my listening room. I gave the subs a try, and yes, it's fun to shake the room at infrasonic frequencies. However, I soon turned the subs off and don't miss them at all (I'd only have to flip a switch, but that never happened since then...).
Thanks! I found the earlier post a bit hard to follow.
If I may ask some more basic questions - where might I learn more about elliptic filters, e.g. how them sum together and why it was chosen vs LR4?
Finally, how was the decision taken to give a slightly falling on-axis frequency response?
If I may ask some more basic questions - where might I learn more about elliptic filters, e.g. how them sum together and why it was chosen vs LR4?
Finally, how was the decision taken to give a slightly falling on-axis frequency response?
Mathias, I am really not excited by these names and do not think either the choice you made or were simed were by luck nore the sims which are close to that names are by chance but just because it works. After all the tunning stays in their maths...no excitment just physik that work and is reflected in sim formulas tools. Pedentic names are easy for summary and short writting.
Yep, this is what I undetstood from the thread : size the port in real life... I am sorry my post didn't reflected that and I already regreat to tried to sumarize and posted.
Thanks for the testimonial about the sub and again about this thread.
What Mototoc had pain to follow is the choice of volume load related to port length I had fail to explain to him in my earlier post and you rephrase.
I can try simplier for Mototok : construct bigger whatever the port size as it is easy to reduce a volume load but not to increase it. At worst if not sure of the diameter length ratio: do a removable port panel. Even more clearly : you will drop between the OSMC and OSMT which both are classic choice, but with one enclosure try...hope that helped. At least the read helped me and Mathias said it again above with simplier words.
Yep, this is what I undetstood from the thread : size the port in real life... I am sorry my post didn't reflected that and I already regreat to tried to sumarize and posted.
Thanks for the testimonial about the sub and again about this thread.
What Mototoc had pain to follow is the choice of volume load related to port length I had fail to explain to him in my earlier post and you rephrase.
I can try simplier for Mototok : construct bigger whatever the port size as it is easy to reduce a volume load but not to increase it. At worst if not sure of the diameter length ratio: do a removable port panel. Even more clearly : you will drop between the OSMC and OSMT which both are classic choice, but with one enclosure try...hope that helped. At least the read helped me and Mathias said it again above with simplier words.
Last edited:
I recommend you take a look at the OSMC paper first. For more background regarding the decision of elliptic filters vs. filters with a "classic" topology, take a look at the corresponding discussion earlier in this thread (not LR4 / "Linkwitz Riley 4th order", which is something else entirely -- see my above comment about Bessel, Butterworth and Legendre...).
That's true for the box tuning, but it's not as easy as this if you consider the loudspeaker system as a whole. Changing the box volume will change the frequency of the upper bass-reflex peak in the impedance curve, and the impedance compensation for the x-over needs to be adjusted accordingly.
After 2 more surgeries, I am finally recovered enough to get back into my shop:
I could not find 20mm material for the front baffle. These measure 25mm. I had to "thin" the back behind the midrange to create the 20mm depth that it requires.
I found a sound absorption product (Owens Corning 703) that has almost identical absorption specs to the Basotect. Any issues using something like this?
Here is a link to the pdf for the 703 stuff: https://dcpd6wotaa0mb.cloudfront.ne...lation-Product-Data-Sheet.pdf?v=1644155734000
I could not find 20mm material for the front baffle. These measure 25mm. I had to "thin" the back behind the midrange to create the 20mm depth that it requires.
I found a sound absorption product (Owens Corning 703) that has almost identical absorption specs to the Basotect. Any issues using something like this?
Here is a link to the pdf for the 703 stuff: https://dcpd6wotaa0mb.cloudfront.ne...lation-Product-Data-Sheet.pdf?v=1644155734000
Fantastic! Recently did a little tweaking and got them off the floor for better bass refinement (Yes, great)—Trying out the J2/BabelAmazeness. Wow. Just Wow. Mad gratitude to Matthias and crew every day. 🙏
Who is this person? An amalgam of myself and Koifarm?
Good for you! Reduced room loading helps in the definition and detail of lower frequencies in a big way.
I will step back into the shadows again now. Have fun.
Surgery... I am in the same boat. Good to hear you're on the way back! Enjoy the workshop time!
I am not familiar with the Owens Corning stuff. Looks like fiberglass. I guess that should work okay, just give it a try and see how you like it. You can always experiment with the amount of damping material.
It is. Works great in lieu of the typical polyester fill, just install it flush with these nails. '50s era version of it installed.
Hi,
I'm starting to collect parts for the XO, and looking at some of the inductors from the suggested manufacturers, I might try DIY coils. Alternatively, are there any suppliers for these parts in the UK? I've located hificollective.co.uk but for example are there alternative to Mundorf?
Many thanks
Earl Josh
I'm starting to collect parts for the XO, and looking at some of the inductors from the suggested manufacturers, I might try DIY coils. Alternatively, are there any suppliers for these parts in the UK? I've located hificollective.co.uk but for example are there alternative to Mundorf?
Many thanks
Earl Josh
Just use good quality parts of the same value. The manufacturer / brand is not so important. The manufactuerers / brands listed in the OSMC paper are just for reference of what I used.
Making good quality DIY inductors is not trivial. I would only go this way if I knew what I am doing (I don't).
Making good quality DIY inductors is not trivial. I would only go this way if I knew what I am doing (I don't).
Did you try the Audiohobby guys? Links at the bottom of the first post of this thread? They had most of the Xover parts when I built my speakers—and they shipped to the US no issue—I'm pretty sure they will send parts worldwide—UK shouldn't be a problem—shipping at the time was free too—They are in Estonia I believe....great experience with them...