One problem I have with this worksheet is that the tuning frequency is independent of the driver's resonant frequency and is user defined. The key assumption in the alignment tables is that the TL tuning frequency is equal to driver's resonant frequency. If you change the worksheet's default tuning frequency to equal the driver fs, the recommended TL geometry changes significantly. If you were to go ahead and then use the MathCad worksheets for further optimization this is not big problem. But for an inexperienced user depending entirely on the worksheet this could lead to a bad design.
Thanks for pointing that out Martin, I will add this clearly to the next version of the sheet.
maybe I've been lucky, or your tables allow for unintended extrapolation, but even for designs that are tuned rather far away from fd the geometry from the allignment tables proved an excelent starting point which needed little alteration.
I have inputted the values into the new sheet provided and came up with the values. Something seems off to me though. The X and Y on the second page are not making sense to me. Never mind thinking about it I figured it out. Now the thing I still am confused about is the taper and the effects it has. What is better a large taper (this is verses straight up and down) or a small taper. In the sheet I left the default 1/11. I have attached the sheet filled out and the start of a very simple pic to illustrate the design. Note I simply opened paint and did it.
1st thing: you have a lower tuning than fd. might work, but as Martin explained the tables were made for use with tuning=fd. like I said, this might still work, but luck must be on yr side.
on the tapering: strong taper will lead to a shorter line, just choose 1:3 in the sheet and watch the change.
other thing: I prefer to put the driver at approx 1/3 of the line, this helps in having a flatter response.
on the tapering: strong taper will lead to a shorter line, just choose 1:3 in the sheet and watch the change.
other thing: I prefer to put the driver at approx 1/3 of the line, this helps in having a flatter response.
There is a table in the alignment documents that suggests the optimum driver offset (Zd, Martin uses the greek tsi)
dave
Hmm, have you simmed this in MathCad? Tuning a 70 Hz driver this far below Fs with a TL isn't normally a good plan and as MJK noted, this SS doesn't accurately take all his parameters into account when a tuning is different from Fs, so if you really want it to be 50 Hz, then best to use the tables/formula in MJK's pdf.
Also, not sure what the adjustable 'Correction on the line cross section' is for since it's not part of MJK's design routine, so until we know it's probably best to leave it at 100%.
Also, there's the driver's location along the line to consider. Placed near/at the top yields the greatest LF gains, but the trade-off is the highest gain 'ripple' higher up in its BW, so requires the most stuffing to quell them which can in turn audibly 'suck the life out' of a wide BW driver, ergo such alignments are best used only for (sub)woofer apps that will be XO'd no higher than at the 3rd harmonic (first) dip. MJK has calc'd positions for each taper ratio, so I recommend you use it at least for a starting point.
There's been much debate about what taper is 'best' for the driver's specs, but you get the most bang/buck with high reverse taper ratios such as 1/11, but without simming them using the actual driver's measured specs there's no way I know of for choosing the optimum taper. Frankly, IME there's a point of rapidly diminishing returns for taper ratios >10:1 (1/10) and why it's been my default ratio for decades.
GM
Also, not sure what the adjustable 'Correction on the line cross section' is for since it's not part of MJK's design routine, so until we know it's probably best to leave it at 100%.
Also, there's the driver's location along the line to consider. Placed near/at the top yields the greatest LF gains, but the trade-off is the highest gain 'ripple' higher up in its BW, so requires the most stuffing to quell them which can in turn audibly 'suck the life out' of a wide BW driver, ergo such alignments are best used only for (sub)woofer apps that will be XO'd no higher than at the 3rd harmonic (first) dip. MJK has calc'd positions for each taper ratio, so I recommend you use it at least for a starting point.
There's been much debate about what taper is 'best' for the driver's specs, but you get the most bang/buck with high reverse taper ratios such as 1/11, but without simming them using the actual driver's measured specs there's no way I know of for choosing the optimum taper. Frankly, IME there's a point of rapidly diminishing returns for taper ratios >10:1 (1/10) and why it's been my default ratio for decades.
GM
I added that to be able to use the sketch on the 2nd page also when the final design would differ from the dimensions from the allignment tables. so leave that to 100% until sure you'll want to change the cross section. (again something on the todo list for the update of the sheet: move all the not standard stuff to a separate location)
A lot of info.
So what I get from this is I need to pick a speaker that has a lower Fs and use its Fs in all calculations. With that I should go back to my original idea of the Jordans' which have a FO of 45. Which is plenty low for a 2.1 or even a 5.1 system. I will rework the sheet for the Jordan speakers and see what I get there.
Thanks guys for the help.
So what I get from this is I need to pick a speaker that has a lower Fs and use its Fs in all calculations. With that I should go back to my original idea of the Jordans' which have a FO of 45. Which is plenty low for a 2.1 or even a 5.1 system. I will rework the sheet for the Jordan speakers and see what I get there.
Thanks guys for the help.
Hey I have found a site where a guy measured the TS params of the Jordan. Question is should I use his measurements or the manufactures? Has anyone done the same and can verify the findings?
http://diyaudioprojects.com/Drivers/JX92S/
http://diyaudioprojects.com/Drivers/JX92S/
You're welcome!
Hmm, if there's a 'sub' system, then the FX120 should do fine in a TL as long as the XO point/slope is at least a typical HT's 80 Hz/2nd order and if a sim is to be believed, the Jordan's lower Fs, much higher Xmax doesn't really 'buy' you anything audibly better in this scenario.
Note too that if might ever be interested in listening at DD/DTS/THX reference, i.e. typical piano bar SPLs, neither can do them at low distortion except possibly in a HTPC app where you're typically < 1m away.
GM
Hmm, if there's a 'sub' system, then the FX120 should do fine in a TL as long as the XO point/slope is at least a typical HT's 80 Hz/2nd order and if a sim is to be believed, the Jordan's lower Fs, much higher Xmax doesn't really 'buy' you anything audibly better in this scenario.
Note too that if might ever be interested in listening at DD/DTS/THX reference, i.e. typical piano bar SPLs, neither can do them at low distortion except possibly in a HTPC app where you're typically < 1m away.
GM
'Measured' is always better if done properly (which these have) and best is to use the measured specs of your drivers, though if the manufacturer keeps a tighter 'rein' on his product's specs than it does to its published specs, then the difference will be minor enough not to matter in the scheme of things.
GM
GM
Well, I am more trying this an test/hobby type thing. I like audio stuff. Just last night I was driving home listening to KWS 10 Days Out CD when BB King came on and I just jammed it out. I mean jammmmmed it out. It was good. I enjoy good music. So, this project is me learning and trying something for fun.
Aside from that the reason I was wanting lower then 70Hz is because I kinda like to cross the sub around 60Hz, but can easily do 80Hz. I will use the Fs value for the Fostex and see what I get. Also, size is a factor. It must be kinda small cause soon to be wife will not like something that dominates the living room.
Here is a link to the sub w/specs that I may use for home sub as all it is doing now is collecting dust.
http://www.imagedynamicsusa.com/manuals/iDMax/iDMax12D4V3.pdf
May do a TL with it as well. We will see. That is another project.
Ok which MJK worksheet do I need to use for this design. Ref the 4th post of this thread to see the design. The design originaly came from SEAS W15LY001 + for ref. I like it and want to come up with my own version called Scoperta SD, but all the sheets that I have been using has soemthing wrong with it. Which means go to the source. If I was to guess I would say I need to use the "Open End TL Worksheet" or "Sections TL Worksheet"
Dave. I must agree with you in that I have reservations about the Jordans, but only after I built the G Chang with Fostex 207E. Just no contest in side by side comparisons. Is it the Yorkshire puddings that make Mr. Moose so clever?
if only it where that easy - we're lucky that every generation has a few minds like these that are willing so selflessly
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.