Hi freddi,
Post #1902: "...20 degree baffle tilt might help with aperture velocities - how do you think the present design would fare under high spl transients?"
I was hoping, that the 20° driver baffle tilt would provide a better match w/ a Karlson aperture, and consequently provide more of the Karlson characteristics. I'm still trying to find the time to enter this latest version into AkAbak, but it looks promising.
Regards,
Post #1902: "...20 degree baffle tilt might help with aperture velocities - how do you think the present design would fare under high spl transients?"
I was hoping, that the 20° driver baffle tilt would provide a better match w/ a Karlson aperture, and consequently provide more of the Karlson characteristics. I'm still trying to find the time to enter this latest version into AkAbak, but it looks promising.
Regards,
For that laid back sound
I like where TB46 is going with this
, the more the baffle is tilted back (providing more space behind the upper slot) the less concern there will need to be about extreme air particle velocities at the slot ..
With the Karlflex12 prototype i did use slotted apertures for many tests , and i only had about 3.5" inches of space between the baffle and front panel , i tried both narrow and wide slots, and of course i was also using the Freddi-mod cavity..... I was surprised to find that there was no problem at all with any undesirably high air particle velocities at the top of the slot...... Akabak predictions using my homebrew multi-part aperture/slot script suggested that there would be velocity excesses in that area but in my opinion the simulations were overstating it .. My K-slot script was probably to blame in this case .. As it turned out, everything was fine
I like where TB46 is going with this
, the more the baffle is tilted back (providing more space behind the upper slot) the less concern there will need to be about extreme air particle velocities at the slot .. With the Karlflex12 prototype i did use slotted apertures for many tests , and i only had about 3.5" inches of space between the baffle and front panel , i tried both narrow and wide slots, and of course i was also using the Freddi-mod cavity..... I was surprised to find that there was no problem at all with any undesirably high air particle velocities at the top of the slot...... Akabak predictions using my homebrew multi-part aperture/slot script suggested that there would be velocity excesses in that area but in my opinion the simulations were overstating it .. My K-slot script was probably to blame in this case .. As it turned out, everything was fine
Last edited:
fwiw I could feel the air move 6 feet away from the upper part of the slot in this 41.5" tall K coupler with cone movements probably around 1/4" peak to peak. It didn't wheeze audibly with music - blocking the top 6 inches of the slot make it sound "slower" when working as a two way. I tried it with a Unity horn on top but it sounded about as good or better with a slotted pipe K-tube.
This being a conventional squished K didn't go as low as your PA310 Karlflex
This being a conventional squished K didn't go as low as your PA310 Karlflex
Tuning and such
Sebastian ,
I like the new approach in your sketch
If 15cm is a bit too tight for comfort then perhaps 12cm?
and yes, you are correct, increasing the S1 width does reduce the length of the interchamber duct, but the amount of compensation (extra panel) required to achieve the same tuning is lessened due to the FB being shifted down from the larger S1 area ....... In other words a larger S1 shifts tuning down, so you can get away with an interchamber duct that is slightly shorter but not a lot shorter .....Not to worry, being a taller Karlflex means that it will tune low enough without difficulty, and may even tune too low , i had to trim up my interchamber port quite a bit just to get the prototype's FB up to 38hz (my prototype is a few inches taller than the final simple design) .... Anyway, Akabak can be used to help approximate the amount of compensation required ..
Once again , i am enthusiastically looking forward to your experiment reports!
....
In the meantime i managed to get my hands on a two very interesting 12" drivers! They are not mine (they belong to my friend Pete) but he wanted me to test them and break them in so i have them in my possession for the moment ....... These things used to be the older generation Eminence Definimax4012 drivers (sold before Eminence made them with glossy cones) but they were blown and are now reconed ........They are custom recone jobs (not kits) performed by a local guy named Patrick from a company called "Robot Underground" who mainly builds high performance (and competition) custom car audio subs .... The T/S parameters are different now, but in a good way, he used a dual layer voicecoil instead of the original single layer type, and he also used an incredibly stiff suspension .... Break-in should be done by dawn , then i can test it again ................... Out of curiosity i may end up trying one of these in the Karlflex12 prototype just to see what it sounds like before i have to give them up ...
Making the S1 15cm will also affect interchamber duct, thus needing an extra panel. Waveguide F1, will only be around 17 cm deep (normal 23 cm), all things equal. What effect will that have on the sound... only one way to tell, rebuild again
View attachment 520451
Sebastian ,
I like the new approach in your sketch
If 15cm is a bit too tight for comfort then perhaps 12cm?
and yes, you are correct, increasing the S1 width does reduce the length of the interchamber duct, but the amount of compensation (extra panel) required to achieve the same tuning is lessened due to the FB being shifted down from the larger S1 area ....... In other words a larger S1 shifts tuning down, so you can get away with an interchamber duct that is slightly shorter but not a lot shorter .....Not to worry, being a taller Karlflex means that it will tune low enough without difficulty, and may even tune too low
, i had to trim up my interchamber port quite a bit just to get the prototype's FB up to 38hz (my prototype is a few inches taller than the final simple design) .... Anyway, Akabak can be used to help approximate the amount of compensation required .. Once again , i am enthusiastically looking forward to your experiment reports!
.... In the meantime i managed to get my hands on a two very interesting 12" drivers! They are not mine (they belong to my friend Pete) but he wanted me to test them and break them in so i have them in my possession for the moment ....... These things used to be the older generation Eminence Definimax4012 drivers (sold before Eminence made them with glossy cones) but they were blown and are now reconed ........They are custom recone jobs (not kits) performed by a local guy named Patrick from a company called "Robot Underground" who mainly builds high performance (and competition) custom car audio subs .... The T/S parameters are different now, but in a good way, he used a dual layer voicecoil instead of the original single layer type, and he also used an incredibly stiff suspension .... Break-in should be done by dawn , then i can test it again ................... Out of curiosity i may end up trying one of these in the Karlflex12 prototype just to see what it sounds like before i have to give them up
...
, enjoy your summertime holiday season! 
Definimax12 special recones
Ok here they are, the drivers i am testing for Pete .... Definimax4012HO reconed by Patrick of Robot Underground in Mesa Arizona ... He did not use a kit, instead he hand picked individual parts (as he would when rebuilding a competition car audio sub) ... He did use an Eminence cone but he also used some custom parts like a dual layer voicecoil and a tighter suspension (mainly a tighter or doubled spider i am assuming) so the parameters are a bit different even after some break-in time ..... The Eminence cone he ordered arrived untreated so he added his own treatment , sort of a cool swirled glossy look .....
The flash on my camera really brought out the authentic Arizona dust that has settled on the cone
The original stock Definimax12 response extended up to almost exactly 2khz and then dropped off like a rock above that (according to the chart Eminence provides), but these custom recones with their different voicecoil, different cone treatment, different dustcap, and and increased moving mass were very likely to have a different contour on the top-end .... So out of curiosity (after some break-in time) i sat one of the drivers up on a small raised platform (a sort of chair) cone facing upward and arranged a bunch of fabric and a few pillows behind the driver's frame to help alleviate reflections and absorb some of the midrange backwave (reducing some upper frequency cancellations in theory) so i could get a rough measurement of the top-end .... I placed the microphone a few feet above the driver facing downward on-axis with the driver ... Looks smooth with a nice bump around 1.5k , then a reduced output plateau above that out to at least 2.5k ...
I did a few listening tests just to hear the midrange with some music playing, my impression is that it has a mellow and smooth sound to it ... Should be very easy on the ears in the right box and with the right accompaniment Hopefully it will have the warm and pleasant sound that higher QTS drivers can produce under the right circumstances ..
I will get another measurement if i actually load one of these into a cabinet before Pete picks them up .. Will be interesting to see how these compare to the PA310..
Upper graph is SPL/Frequency , and the lower graph is harmonic distortion.
I REALLY like the look of these T/S parameters!
These look like nearly ideal parameters for a 12" driver for a 60 liter Karlflex ! If i really wanted to get picky and critical i could say that moving mass could have been a little less and QMS could have been a little higher , but i'm not that picky, and what we have here actually looks awesome in simulation .....
These old recycled Emmies are now some very cool and unique items!, good job Patrick!
Interesting to note that my Dayton DATS unit had some difficulty measuring the VAS figure properly (using the added mass method) because this is now a driver with an unusually tight suspension for a 12" .... I may try measuring the VAS using a different method here in the next few days to see if the result is any different ..
Ok here they are, the drivers i am testing for Pete .... Definimax4012HO reconed by Patrick of Robot Underground in Mesa Arizona ... He did not use a kit, instead he hand picked individual parts (as he would when rebuilding a competition car audio sub) ... He did use an Eminence cone but he also used some custom parts like a dual layer voicecoil and a tighter suspension (mainly a tighter or doubled spider i am assuming) so the parameters are a bit different even after some break-in time ..... The Eminence cone he ordered arrived untreated so he added his own treatment , sort of a cool swirled glossy look .....
The flash on my camera really brought out the authentic Arizona dust that has settled on the cone
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
The original stock Definimax12 response extended up to almost exactly 2khz and then dropped off like a rock above that (according to the chart Eminence provides), but these custom recones with their different voicecoil, different cone treatment, different dustcap, and and increased moving mass were very likely to have a different contour on the top-end .... So out of curiosity (after some break-in time) i sat one of the drivers up on a small raised platform (a sort of chair) cone facing upward and arranged a bunch of fabric and a few pillows behind the driver's frame to help alleviate reflections and absorb some of the midrange backwave (reducing some upper frequency cancellations in theory) so i could get a rough measurement of the top-end .... I placed the microphone a few feet above the driver facing downward on-axis with the driver ... Looks smooth with a nice bump around 1.5k , then a reduced output plateau above that out to at least 2.5k ...
I did a few listening tests just to hear the midrange with some music playing, my impression is that it has a mellow and smooth sound to it ... Should be very easy on the ears in the right box and with the right accompaniment
Hopefully it will have the warm and pleasant sound that higher QTS drivers can produce under the right circumstances ..I will get another measurement if i actually load one of these into a cabinet before Pete picks them up .. Will be interesting to see how these compare to the PA310..
Upper graph is SPL/Frequency , and the lower graph is harmonic distortion.
An externally hosted image should be here but it was not working when we last tested it.
I REALLY like the look of these T/S parameters!
These look like nearly ideal parameters for a 12" driver for a 60 liter Karlflex ! If i really wanted to get picky and critical i could say that moving mass could have been a little less and QMS could have been a little higher , but i'm not that picky, and what we have here actually looks awesome in simulation .....
These old recycled Emmies are now some very cool and unique items!
, good job Patrick!Interesting to note that my Dayton DATS unit had some difficulty measuring the VAS figure properly (using the added mass method) because this is now a driver with an unusually tight suspension for a 12" .... I may try measuring the VAS using a different method here in the next few days to see if the result is any different ..
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
Trade?
Sebastian ,
I would gladly send you some dust and sunshine if you would send us some rain
These reconed Definimax seem good so far, and the stock Definimax 12 is certainly a fantastic driver, but yes either way they are sort of expensive for a 12" driver with only 6mm of xmax (although i did find a few lower priced alternatives that sim almost as well , in the $100 price range) ...... Besides the price these Definimax woofs are also heavy for a 12" driver, which takes away from the portability somewhat ..... So there are the compromises to squeeze out an extra 5db max output at 40hz (or so the simulations say) , i am really curious to see if the real-world performance reflects that same decibel difference .... Pete hasn't come by to pick these up yet so i still have time to load one into the Karlflex prototype and get a measurement
That 12" driver looks like a real killer and very nice FR, but I'll bet they will be expensive with custom recone and what not ?!
You fight dust, we fight rain. HATE this scandinavian wheater...
Sebastian ,
I would gladly send you some dust and sunshine if you would send us some rain
These reconed Definimax seem good so far, and the stock Definimax 12 is certainly a fantastic driver, but yes either way they are sort of expensive for a 12" driver with only 6mm of xmax (although i did find a few lower priced alternatives that sim almost as well , in the $100 price range) ...... Besides the price these Definimax woofs are also heavy for a 12" driver, which takes away from the portability somewhat ..... So there are the compromises to squeeze out an extra 5db max output at 40hz (or so the simulations say) , i am really curious to see if the real-world performance reflects that same decibel difference .... Pete hasn't come by to pick these up yet so i still have time to load one into the Karlflex prototype and get a measurement
Last edited:
Happy new year to all.
So I removed the S1 panel from the Karlflex, thus making it like a portet box. It was to test if the panel itself or S1 duct was causing these coloration (150-500hz) I get from the Karlflex. I must say that it did help, the FR is more even
low pass is 900 hz. Now it sound a lot better and natural. With a little EQ it could be very good.
Matthew came up with a good idea of placing the S1 horizontal and folding it below the driver. It will be interesting to see if that cures the problem.
So I removed the S1 panel from the Karlflex, thus making it like a portet box. It was to test if the panel itself or S1 duct was causing these coloration (150-500hz) I get from the Karlflex. I must say that it did help, the FR is more even
low pass is 900 hz. Now it sound a lot better and natural. With a little EQ it could be very good.
Matthew came up with a good idea of placing the S1 horizontal and folding it below the driver
. It will be interesting to see if that cures the problem.
This looks like more progress!
That graph does looks very smooth through the mids! Was this with the EV driver? Or 15LB100? and this was with the front panel installed so it is using the front chamber correct?
Good to hear that it sounds better, so we now know that the S1 panel or duct is somehow creating the colouration effect ...
This makes your cabinet much more like the old classic Karlson cabs and maybe even a bit like XRK's XKi design (the XKi sub was cool too with it's symmetric loading, maybe we could scale that up some time)... The lack of S1 also makes your box sort of like the Metro T-15, except that your box has a more complex front chamber than all of the above...
If it turns out that the Offset driver and TL loading features are not necessary at all then this would greatly simplify things! ..
Did you happen to get a measurement (with the S1 in place) from that same mic placement and cabinet placement? Maybe we could overlap the graphs to get a better idea of what ranges changed drastically in response ..
I think folding it under was also an idea that TB46 might have been contemplating in a few of his variants so I cannot take all of the credit hehe ...
That graph does looks very smooth through the mids!
Was this with the EV driver? Or 15LB100? and this was with the front panel installed so it is using the front chamber correct? Good to hear that it sounds better, so we now know that the S1 panel or duct is somehow creating the colouration effect ...
This makes your cabinet much more like the old classic Karlson cabs and maybe even a bit like XRK's XKi design (the XKi sub was cool too with it's symmetric loading, maybe we could scale that up some time)... The lack of S1 also makes your box sort of like the Metro T-15, except that your box has a more complex front chamber than all of the above
...If it turns out that the Offset driver and TL loading features are not necessary at all then this would greatly simplify things! ..
Did you happen to get a measurement (with the S1 in place) from that same mic placement and cabinet placement? Maybe we could overlap the graphs to get a better idea of what ranges changed drastically in response ..
Matthew came up with a good idea of placing the S1 horizontal and folding it below the driver
I think folding it under was also an idea that TB46 might have been contemplating in a few of his variants so I cannot take all of the credit hehe ...
Last edited:
Matthew, this new FR without S1, was done with full front stubs, round aperture and with and with 15LB100 driver.
Here are the FR. 1. picture without S1 compared with the one with S1 panel.
Maybe not a big difference, but the sound surely was. 1. FR are more smooth and the sound was much more natural/neutral though not as deep in extention.
This new folding will barely fit under the driver
The waveguide panel in S1 will run up against the back of the driver and maybe cause problems. But it is not a big rebuild, just a little more to be done
I would like to keep the S1 stub (transmission line) to give a more full lowend and ported speakers just don't float my boat
Here are the FR. 1. picture without S1 compared with the one with S1 panel.
Maybe not a big difference, but the sound surely was. 1. FR are more smooth and the sound was much more natural/neutral though not as deep in extention.
This new folding will barely fit under the driver
The waveguide panel in S1 will run up against the back of the driver and maybe cause problems. But it is not a big rebuild, just a little more to be done
I would like to keep the S1 stub (transmission line) to give a more full lowend and ported speakers just don't float my boat
how much grief is involved for you to model Acoustic's 115bk? - can it be fudged in hornresp?
Hi Freddi !
I worked on this a bit today .... It is some major fudging! hehe ... The smaller chamber at the top cannot be emulated in HR so it just gets added into the back chamber ... and the aperture is just a very rough approximation ....
I did manage to get the internal volume of both chambers fairly close (even compensating for a few liters taken up by the rear of the driver) , tuning of 55hz , constriction (duct area of around 175 sq cm) ....
With all of the above the simulated impedance peaks were not located in exactly the right places except for the lowest peak which was good , but the middle peak is too low by quite a margin and the 3rd peak was placed too high , in the upper 200s instead of 224hz .... So it is way off base from your measurements , all fudged up
... So far HR just doesn't seem to be the proper tool for simming Karlsons but if you would like the inputs for my attempt i can send them to you .... I might try working on it a little more tomorrow, maybe i will come up with some results that more closely match your measurements..
To S1 or not to S1 , that is the question
This is great stuff Sebastian, it looks like you have really made a lot of headway on the colouration issue and are sorting out this S1 related problem
........
It is also very good to know that your large and intricate front chamber design is working out, it is the first time i have seen a front chamber system set up that way, very unique ...
Also great to know that those 15LB100 drivers have an acceptable sound quality!
Some might say "cramped" but i prefer the word "cozy" , hehe
I feel the same way, it would be nice to be able to keep the S1 driver offset and the longer path length ...
This is great stuff Sebastian, it looks like you have really made a lot of headway on the colouration issue and are sorting out this S1 related problem
........ It is also very good to know that your large and intricate front chamber design is working out,
it is the first time i have seen a front chamber system set up that way, very unique ... Also great to know that those 15LB100 drivers have an acceptable sound quality!
This new folding will barely fit under the driver
The waveguide panel in S1 will run up against the back of the driver and maybe cause problems. But it is not a big rebuild, just a little more to be done
Some might say "cramped" but i prefer the word "cozy" , hehe
I would like to keep the S1 stub (transmission line) to give a more full lowend and ported speakers just don't float my boat
I feel the same way, it would be nice to be able to keep the S1 driver offset and the longer path length ...
Who has a snow sled?
Yes, hehe, that looks like an excellent slope for the kids to race their snow sleds down in the winter!
Seriously though, the general response curve is actually not too far off from the sim!
Interesting that the harmonically related dips and peaks above 200hz are shifted over a little bit in the measurement VS sim and also that the impedance peaks in the sim are also not quite where they are supposed to be, especially the peak in the middle .... Maybe this is related to the slotted aperture used on the real cabinet (which HR cannot emulate) , or perhaps it is related to the small series chamber at the top of the cabinet, a volume which we just combined with the rear chamber in the HR sim ...
The discrepancy you are seeing near fundamental may have to do with driver parameters ... Have you tried plugging the EVM15L parameters into the simulation? The 15LB100's parameters are significantly different.
hey MMJ - the general shape of your hornresp fudge looks pretty good - I think it (and akabak) are optimistic for LF My yard is in the middle of a hill and sliding off
Yes, hehe, that looks like an excellent slope for the kids to race their snow sleds down in the winter!
Seriously though, the general response curve is actually not too far off from the sim!
Interesting that the harmonically related dips and peaks above 200hz are shifted over a little bit in the measurement VS sim and also that the impedance peaks in the sim are also not quite where they are supposed to be, especially the peak in the middle .... Maybe this is related to the slotted aperture used on the real cabinet (which HR cannot emulate) , or perhaps it is related to the small series chamber at the top of the cabinet, a volume which we just combined with the rear chamber in the HR sim ...
The discrepancy you are seeing near fundamental may have to do with driver parameters ... Have you tried plugging the EVM15L parameters into the simulation? The 15LB100's parameters are significantly different.