First Look - TD-18H+

Very Very nice. Great work on that bracing.


Also, please tell your son thank you for his service.

Oh my sweet jebus! I'd love to know how this one turns out.

I think you'll be quite happy with the results. One thing to note on the Xmax issue. JBL quotes 50mm P-P before damage. That is the clearance before the coil could hit the back plate. We have even more clearance than, about 75mm total physical clearance for the coil. What they don't mention is that their surround/spider combination limits about 20mm. The cone/surround we use is the same as on the EV 18's and also limits 20mm, but our spider has has more travel and is very linear to 20mm.


I'm sorry about that. I didn't realize you had never seen the driver pics. We need to get pics edited up and on the website with full information soon. The phase plug is a large part of the heatsinking ability in this driver as with our other TD woofers. There is more heatsinking area for the coil to shed heat to the adjacent steel in this driver than in the 4" coil JBL and EV drivers due to the massive thick top plate and aluminum VC former.

I'm also eagerly awaiting to hear your results on this project.

John

looks like it may have been inspired by the 18lw1400? thicker top plate maybe? gives a little bit more bl.

how are you calculating xmax? better yet, what is top plate thickness and coil length?

interesting bit about the spider. 18sound claims a double silicone spider for "improved linearity" but compared to what...who knows.

upgrades include copper sleeve. does inductance matter up to 80hz?

uses heat sinking in place of venting for cooling strategy.

curvilinear cone vs straight wall/deep dish cone. what is the advantage of curvilinear cone up to 80hz? the 18sound cone is ultra stiff and as a result has a monster breakup around 1.8khz. any frequency sweeps?

durability needs to be proven of course, but aes is so far so good, so we can be hopeful.

this looks like a very promising driver...sufficient xmax for deep bass, sufficient motor for good damping of the higher notes.

It really doesn't increase Vd by any significant amount. In a dustcapped driver it wouldn't be an issue at all. In a driver with phase plug, the coil Sd is so small compared to the driver it really doesn't matter.


Not sure I've ever looked at the 18lw1400 other than a quick glance. This is basically the same as the other Lambda drivers that originated back 8-10 years ago now. The only difference is we went up to a 2.5" VC as the 2" just didn't give quite enough Bl for the big 18" driver.


Xmax based on the Bl curve and the point at which Bl drops to 70% of the rest value. Top plate is 18mm and coil is 38mm long 4 layer flat copper.


Linearity of inductance is more important than the absolute value. The other benefits are the reduction in flux modulation which is a huge benefit. Better heat transfer as well.


The other thing to look at is with the very tight gap and the top of the coil being open allows for the air to be quickly displaced in the gap between the coil and the pole.


Really not much of an issue in the cone shape. For sub 80hz it just needs to be strong enough to do what it has to. We have some people using them to 250-300hz for large 3way speakers though.


Elite audio replaced their 12 B&C 18TBX100's with our drivers back the day before new years. They've been playing them multiple times per week for live sound use powered off Lab Gruppen amplifiers. They have had no issues to this point and have played them quite hard for some large live shows. For any type of home environment use they should do very well then.

John

Hmm, massive 3ways using this driver, TD10m, and 2" compression driver/horn of choice?

Looking good, I like the recessed speakon connector. Once I bought them, I will never go back to any other type of connector.

thanks john j. for the extensive reply.

+1. subwoofer art. internals deserve a plexi-panel.

Curious, have you done any modeling with WinISD or a similar program with this driver. I tried using the specs on ae's website and it rejects them. It tries to make the VD like 1.6 or something. I triple checked my numbers I enetered and they are all right, so I'm curious what the deal is there.

I have had no trouble with it in Unibox. I don't use WinISD as I find it a huge PITA.

Try entering only the qms, fs, vas, mms, re, bl and PE from the website. Winisd will autofill in the rest. All the numbers it substitutes are very close to John's listed values.

The only major difference is the Vd. John calculates that from peak to peak, WinISD calculates from peak to middle so cut it in half and you are good to go.

yes.

default winisd enclosure 2.9 cu. ft. 29 hz = gray.

8 cu. ft. 25 hz = red.

both with 4th order l-r filter @ 80 hz.

winisd can be frustrating. here is what i do:
make sure all fields are clear, then
enter qes then hit tab, enter qms
hit tab a couple times and let it calculate qts
hit tab a few times to move to mms
enter mms, re, bl, le, sd, xmax, and pe
by using tab after entering each data, it will calculate what it needs to
i no longer get any conflicts
sometimes the specs calculated don't match exactly the manufacturer specs, but its only rounding type errors, so it is not material

Even doing it with that I still get the Vd as 1.71 and he states it on the website as 3.4

John lists it as Peak to Peak. WinISD does resting to Peak so it is half what John lists.

So with that then all of the numbers are right correct?

The graph I'm seeing with the specs is horrible. I'm getting the -3db dropoff at 107hz. And a fsc at almost 52hz. WinISD must have lost its mind as that is with an 8 cubic foot box.

you have to model it with a low pass filter, otherwise it looks like a high f3. It models similar to the SoundSplinter drivers.

This is my first DIY so I'm not familiar with how others work exactly. The only other I've played with is a maelstrom-x and they look completely different.

Look at the pic in LTD02's post. If he did not have a low pass filter configured, then the response would rise as it went further up the range. Adding a low pass filter at 80hz, as you would in real life, shows a realistic curve.

Using Winisd...

Input the following parameters first as they are the most concrete. RE, BL, MMS, CMS. Let it calculate. Don't freak out over small differences from the published or mfg specs they really won't change things that much. Then follow with QMS, xmax, PE, LE , etc... speaker Q's, FS and VAS are all going to vary quite a bit between drivers. Again don't freak out if it says that the FS should be 23.7hz and it's listed as 24.5hz by the mfg. This is fairly negligible in the scheme of things.

tack, i'm still p-o'd that jbl didn't help you out. at a min, they should have offered a replacement at cost. that's what i would have done. my guess is that john j. would have done the same. i have been a long time jbl'r...it's fading though...

"I just hope the port is going to be big enough "

what are the port dimensions again? 8" diameter and shooting for 23hz tuning(?), what is the length? i'm calculating 26" would be required and your port doesn't appear to be that long.

those dimensions look great. i'm getting roughly 25hz tuning, so a little plus or minus will be fine. can't wait to hear your impressions.