Maybe You don’t have one note because you’re tuned at 60hz (bp4) with a flat response below that with cabin gain
and avoiding the issue that a ported box where Fs= Fb would be really bad because it’s underdamped(high Qts ) and tight vas/Vb relationship typical of car audio junk? It’s a really bad combo and keeps getting worse as ‘SPL’ and insane electrical/amp to move those drivers is crazy
and avoiding the issue that a ported box where Fs= Fb would be really bad because it’s underdamped(high Qts ) and tight vas/Vb relationship typical of car audio junk? It’s a really bad combo and keeps getting worse as ‘SPL’ and insane electrical/amp to move those drivers is crazy
Thank you I will most likely go with this enclosure since it looks the most easy to make do you have any more diagrams on how to make this cheers Ricky this should play down to 20hz ish right if tuned to 30-35hz ?
I like using Excel since it's like using drafting paper. I can be precise going from HR to Excel. The CAD has issues with the inputs. It doesn't do ALL inches or ALL cm's.
It does.
You can easily change units from International Standard to Imperial ones and others
Attachments
Hi all, Merry Christmas! Thanks for sharing all of your knowledge on this thread. I just fired up my self built 2 x Tham15 in birch ply for the first time and they sound epic but one of the grilles is buzzing. Can anyone recommend a way to brace the grille or some thicker grille material? I have used 1.2mm thick steel mesh, sourced from ebay "UK Made Cabinet Steel Mesh Grill, Heavy duty for Speaker 600 x 521 x 1,2mm." I have a strip of 2mm EVA foam as a backing strip around the perimeter between the ply and the mesh. What have others used successfully in their Tham15 builds?
Happy New Year everyone! I've enjoyed all the detailed responses and the lived experience shared here - I'm grateful for the opportunity to learn from such a passionate collective. Thanks to Anders in particular for the original design! Have managed to knock out an OG Tham15 using 18mm marine ply and an Australian made driver - the Lorantz AC400X-B9s-8. I found time for some testing over the holidays. I'm still very much learning the ropes, but my cautious (hopeful??) assessment was that the driver performs reasonably close to the sim, especially from the DATS impedence curve (measured plot is blue, Hornresp sim is pink). Would welcome other assessments and/or critique of course!
Ground plane measurement at 1m using calibrated Umik-1 is a little untidy due I think to some unwanted reflections but not too far off the Hornresp sim. I'm wondering whether the LF response may improve once the driver breaks in? Using some EQ, and experimenting with a few high and low pass filters makes for a pretty usable response and my subjective experience of the bass output using Behringer NX6000 amplification was very positive.
Ground plane measurement at 1m using calibrated Umik-1 is a little untidy due I think to some unwanted reflections but not too far off the Hornresp sim. I'm wondering whether the LF response may improve once the driver breaks in? Using some EQ, and experimenting with a few high and low pass filters makes for a pretty usable response and my subjective experience of the bass output using Behringer NX6000 amplification was very positive.
Last edited:
Yes, "Exp" is definitely the wrong expansion setting for that build.
If you use the driver's impedance curve to derive its semi-inductance parameters, and then use those in the Hornresp sim, you should get a closer match with the predicted impedance curve and the response in the passband. See - https://www.diysubwoofers.org/misc/semi-inductance.html
Of interest is that the predicted peak below 200 Hz is entirely absent in the measured response of your TH. It's actually replaced with a DIP in the response. That's quite unexpected - the Hornresp sim is usually pretty good at estimating what the response would look like, up to a few hundred Hz. I'm tempted to suggest that perhaps one of the larger bends in the design is acting as a LP filter at that frequency, but I suggest correcting the sim first to see what the corrected sim suggests the response would look like at that frequency.
If you use the driver's impedance curve to derive its semi-inductance parameters, and then use those in the Hornresp sim, you should get a closer match with the predicted impedance curve and the response in the passband. See - https://www.diysubwoofers.org/misc/semi-inductance.html
Of interest is that the predicted peak below 200 Hz is entirely absent in the measured response of your TH. It's actually replaced with a DIP in the response. That's quite unexpected - the Hornresp sim is usually pretty good at estimating what the response would look like, up to a few hundred Hz. I'm tempted to suggest that perhaps one of the larger bends in the design is acting as a LP filter at that frequency, but I suggest correcting the sim first to see what the corrected sim suggests the response would look like at that frequency.
Oh, I forgot to mention that perhaps you can use my BOXPLAN-THAM workbook to provide more accurate parameters for your Hornresp sim, based on your as-built design. See - https://www.diysubwoofers.org/sheets/
Ok, how did I do this time? I copied the parameters straight from www.martinsson.cc originally but it appears I even managed to transpose one of those settings incorrectly. The peak is still in the same spot, and I reviewed the cutting plan and realised that since I used the little bass heads sound system (big up those guys) publicly available fusion360 model the dimensions are slightly different to the martinsson ones. Following my usual MO here of learning by charging ahead and making errors lol. I'll go have a look at your spreadsheet Brian, thanks for that tip.. and I'll also recheck my driver TS parameters.
Last edited:
Ok, so in an attempt to see get closer to the box I've actuall built, I've entered all the dimensions from Fusion360 into Brian's excellent Boxplan-THAM spreadsheet and hit optimise before exporting into hornresp. It looks fairly close in most of the response, just the measured dip vs the modelled peak persists. Did I read your hornresp sim for the POC2 project Brian you actually seemed to predict the dip at a similar spot to mine as well as measure it? Still a bit unsure if I've got my model right here as exported from the spreadsheet although as you say it may be the folding near the horn mouth not being handled. If the L34 value is the distance from driver to the horn mouth (have I got that right?) then 53.7 looks a bit longer than on my cabinet (on most THAM models it seems around 20). In any case, I'm learning a great deal here which is of course the name of the game! None of this in any way detracts I might add from the real world experience of this particular THAM15 - in the passband it really has surprising output and sounds great!
None of this in any way detracts I might add from the real world experience of this particular THAM15 - in the passband it really has surprising output and sounds great!
Last edited:
Robot, why you care above 100Hz , a sub will only take over from 100 and below on most cases , so strange things happening @200 or above.... Who cares! , it will be shaved by the low pass filter and besides, the rest of the sim matches , in the name of science you wanted that matched equally along the entire spectrum but what it matters is just the passband freq range on a subwoofer.
Imho, you did a great job , don't obsess with the spare change meanwhile you have pocketed the lottery.
Imho, you did a great job , don't obsess with the spare change meanwhile you have pocketed the lottery.
The distance from driver to mouth is L45, so your 22cm is certainly in the right ballpark. I wouldn't expect the 2cm drop to 20 to make any noticeable difference in real life, but you can certainly play with it in hornresp to see for yourself.
Looks like you enabled the semi-inductance feature for the driver as well, which will make the sim a bit more accurate. Note the difference in the shape and slope of the predicted passband curve.Ok, so in an attempt to see get closer to the box I've actuall built, I've entered all the dimensions from Fusion360 into Brian's excellent Boxplan-THAM spreadsheet and hit optimise before exporting into hornresp. It looks fairly close in most of the response, just the measured dip vs the modelled peak persists. Did I read your hornresp sim for the POC2 project Brian you actually seemed to predict the dip at a similar spot to mine as well as measure it? Still a bit unsure if I've got my model right here as exported from the spreadsheet although as you say it may be the folding near the horn mouth not being handled. If the L34 value is the distance from driver to the horn mouth (have I got that right?) then 53.7 looks a bit longer than on my cabinet (on most THAM models it seems around 20). In any case, I'm learning a great deal here which is of course the name of the game!
View attachment 1258632 View attachment 1258635
I haven't run any calculations yet, but I suspect that the absence of the peak just below 200 Hz is due to the large bends in the path just before the mouth. The effect and impact of those bends on the frequency response is not something that can be predicted by Hornresp. In any case, removal of that peak is actually a good thing.
I notice that there's a similar dip in the measured response of my POC2 which had a THAM-style fold, so it looks like a feature of that type of fold.
Comparing Stoneeh's measurements (different driver) from post #687 to yours the peak and dip locations look pretty similar, though the amount of smoothing in your measurement may be masking the the peak just below 200 Hz that can be seen in his.
The Lorantz AC400X-B9s-8 is 130mm deep, the 15LB075-UW4 ("The Box" driver Stoneeh used) is 156.5mm deep, the deeper cone does present more mouth obstruction and an increased volume in the throat area which also may account for some differences.
I doubt the LF response will improve once the driver "breaks in", but a short run of 25Hz at Xmax before testing would loosen it up and give a "for sure" answer.
Art
Robot, why you care above 100Hz , a sub will only take over from 100 and below on most cases , so strange things happening @200 or above.... Who cares! , it will be shaved by the low pass filter and besides, the rest of the sim matches , in the name of science you wanted that matched equally along the entire spectrum but what it matters is just the passband freq range on a subwoofer.
Imho, you did a great job , don't obsess with the spare change meanwhile you have pocketed the lottery.
If your distortion levels are bad above 100hz, stuff the 1st 3rd of horn path.
Over on avsforum.com, there is thread where a guy did extensive testing on a Devastator (BP6P) enclosure with 4 of the $29 JBL 12's from Best Buy. He put 2 pillows in the correct location (after my suggestion) and the pillows killed the distortion peaks and smoothed out the frequency response. No $500 subwoofer can compete with that build.