TD15M + BMS 4550 + QSC waveguide = az's "Spendy-Wave" build

audiozaz

rtfm!? - tl;dr
Design goals:

To build a speaker that can play beautifully and kick some ass when asked to.

To build a sensitive speaker that can accomodate high power SS amps as well as low power tube amps with higher output impedance and sound good in both small and large rooms.

To build a speaker I won't be likely to want to improve upon for many years.

Background:

I've been doing some DIY speaker stuff for a few years now and really enjoy the results of my own labor. It's much more rewarding to me than buying commercial speakers, as good as they can be. Reading up a lot on diverse aspects of speaker building as well as doing a lot of, well, doing is a good way to learn about it and is quite fun. I've had the Econo-Wave project floating somewhere in the back of my brain ever since I joined AK.

I really enjoy pro drivers, as I've been playing with some in the last year or so. Instead of taking baby steps up the quality ladder, I've decided to go with an all-out build and get something really good, now. This is definitely not econo (from my perspective), which is why I dubbed these "Spendy-Waves".

The drivers:

The BMS 4550, I already own and was bought in advance for such a build. I've toyed with it in on some small Smith horns I built, as well as the Transylvania Power Company "The Tube". It's a very good compression driver IMO. A definite step up from D220Ti, which is quite good itself.

As for a woofer, I hesitated for a while between GPA 416-8C and Acoustic Elegance Lambda TD15M. They are similarly priced and both fit most other requirements for this project. I ended-up picking TD15M, mostly based on AK members who have glowing opinions of it. Don't worry, I won't pin it on you guys if I end-up hating it. ;)

For the waveguide, I never really considered the PT waveguide as it might a bit small to go with a 15" woofer. TD15M is supposed to have good dispersion up to 2kHz though. Something larger will load the compression driver better as well. There was not a ton of options there: the Dayton 12" round WG, the 12" SEOS and the QSC were what I was able to find. Reading up on these and considering other criteria of the project, the QSC was picked, although it seems a bit hard to get one's hands on them lately.

The TD15M's have been ordered and I assume the ~4 weeks lead time stands for now, so I have some time for planning before I get these. If all goes well, the QSC waveguides will be in my hands next week, or so I hope. :)

az
 
I have the same project with the DE250 going, we should be able to collaborate on this, although lord knows when you will get your woofers ;).
 
cont'd

The basic build:

I've toyed with a few ideas and configurations, received a bit of advice and am not decided on anything too hard for now at least.

From simulations using WinISD and a bit of Hornresp, it seems to me that a 5.5 cubic feet box tuned to 35Hz is what I want. Bass response is well-damped and has decent extension.

WinISD_sim.jpg

f3 is 50Hz
f6 is 38Hz
f10 is 31Hz

I don't want an overly strong bass response that will turn out to be boomy in my small-ish listening room. Room gain should kick-in somewhere where the response is slightly falling and should result in a flatter in-room response than a true "max-flat" alignement.

If I want to run a tube amplifier with high output impedance, I still have some wiggle room up to ~DF=2, meaning output impedance of ~4ohm. Mid-bass hump is ~2dB, still acceptable, if somewhat warm-sounding. I know this, as my Richard Allan CG10T in a small reflex box had a similar transfer function and it was good, but again, if a only bit warm.

WinISD_sim_hi_Z.jpg

az
 
I have the same project with the DE250 going, we should be able to collaborate on this, although lord knows when you will get your woofers ;).

Sounds good!

I do like to get my stuff ASAP, I can be impatient, but I understand AE is a small company and John looks like a busy guy, so I can understand if that takes longer.

In the meantime, I'll be able to work on the HF block of the build. I'm not planing on building exactly Zilch's E-Wave crossover from the get go, but have a feeling I'll end-up with something close enough. CD-compensation will be needed, I'll need a swamping resistor in front of that and will likely try a 2nd order network initially, around 1.5kHz, and see what that does. I have some Celestion TF1020 10" pro woofers in 49 liters reflex tuned to 50Hz that I'll use as bass-bins in the meantime.

az
 
I'm doing ~5.5 at 40Hz. Will you be measuring? There's not much point in messing with the HF crossover without the woofer, as I see it. The woofer wants to be crossed over around 800Hz, so you'll want to take your HF down as far as it will sound okay.

Optimal spacing between this woofer and the QSC is just slightly closer than possible, so plan your cabs to jam them right up against each other.
 
The box:

As I want to have 5.5cu.f of actual reflex volume, I'll have to build slightly larger to accomodate space taken by the woofer, the waveguide, bracking and the port(s). I'm planning to buld an actual 6cu.f enclosure, which should give me ample room to fill all the above.

As for dimensions, I want to keep a somewhat wide front baffle, in order to push baffle-step down enough so that room-gain will kick-in close enough and little or no compensation will be required. Total height should bring the WG close to ear level, though I don't mind a bit lower. I might have the boxes on casters, which could elevate it by ~5". I like to move them from room to room once in a while and swap gear around. With a 28lb woofer from the start, these won't be featherweights. This remains to be decided.

Using Godlen Ratio dimensions does seem like a good idea: 0.618:1:1.618 for depth, width and height respectively. Actual dimensions for ~6cu.f would be 13.5" by 21.75" by 35.25, again, respectively. This is not so aesthetically pleasing to my eye, but once I add panel thickness all around and possibly even double-up fron and back baffles, it'll be fine. This would give a total height of just over 36", maybe ~41" with casters, not a bad height for the WG assuming I place it high enough on the baffle (which I will).

az
 
I'm doing ~5.5 at 40Hz. Will you be measuring? There's not much point in messing with the HF crossover without the woofer, as I see it. The woofer wants to be crossed over around 800Hz, so you'll want to take your HF down as far as it will sound okay.

Optimal spacing between this woofer and the QSC is just slightly closer than possible, so plan your cabs to jam them right up against each other.

800Hz, really? I though they go higher form what I could gather. The BMS 4550 suggested Xover is 800Hz 2nd order, so this should not be a problem IMO. The QSC loads down to ~1kHz IIRC, as stated at least.

I'll be measuring, for what my non-ideal setup and room is worth at least. :) I've had some success doing so to date.

az
 
Indeed, I'm talking 5.5 after subtractions. Be aware that your listening axis will be unknown until the crossover design is complete, so you'll have to plan for possible stands/tilting unless you do a test cabinet.
 
800Hz, really? I though they go higher form what I could gather. The BMS 4550 suggested Xover is 800Hz 2nd order, so this should not be a problem IMO. The QSC loads down to ~1kHz IIRC, as stated at least.
Yes, there is a resonance that results in a response ripple above ~900Hz. Getting down near 1k crossover, it's pretty much knocked out.. It may not be much of an issue in the real world, but I would definitely go a little lower than the 1.6k standard ewave, as that's not good for this driver spacing either. I will have much more to report on the crossover point soon.
 
Yes, there is a resonance that results in a response ripple above ~900Hz. Getting down near 1k crossover, it's pretty much knocked out.. It may not be much of an issue in the real world, but I would definitely go a little lower than the 1.6k standard ewave, as that's not good for this driver spacing either. I will have much more to report on the crossover point soon.

Good point, the QSC is quite "high". I'll stick it pretty close to the TD15M. ~1kHz seems doable for both drivers.

I'm hoping to keep the part count minimal on the TD15M Xover. Maybe it won't require a Zobel due to the low Le and shorting rings?

az
 
I'm in the same boat with Dumptruck...DE250 on QSC waveguide. Just waiting on a pair of TD15M to be built. Currently the cabinets are about 5 ft^3 and tuned to about 42Hz if I remember correctly. I used a pair of 2.5" (?) friction-fit ports so I can pull them and change tuning as needed.
 
It definitely doesn't need a zobel. I hope to restart my crossover work at any moment here. Work has just been a disaster lately. Shouldn't I be dealing with that instead of posting on AK? ... yes.

FWIW, 2 x 5" diameter ports are the best option I came up with in sims for this woofer. Keeps airspeed looking mighty fine with completely manageable length (~4-8").
 
I'm in the same boat with Dumptruck...DE250 on QSC waveguide. Just waiting on a pair of TD15M to be built. Currently the cabinets are about 5 ft^3 and tuned to about 42Hz if I remember correctly. I used a pair of 2.5" (?) friction-fit ports so I can pull them and change tuning as needed.

I see a few AK'ers will be churning out TD15M "Lambda-Waves" this summer! :thmbsp:

It definitely doesn't need a zobel. I hope to restart my crossover work at any moment here. Work has just been a disaster lately. Shouldn't I be dealing with that instead of posting on AK? ... yes.

That's good new to me! :)

FWIW, 2 x 5" diameter ports are the best option I came up with in sims for this woofer. Keeps airspeed looking mighty fine with completely manageable length (~4-8").

I was considering two 4" ABS pipes, 7.125" long for my 35Hz. It would hit 5% mach at 125W, where output will be close to 120dB.

I was also looking at a large slot on the bottom. Sublte changes in cabinet elevation could be used to fine tune the output a bit. It's harder to adjust the actual port dimensions versus friction-held pipes though.

Is there any advantage to an Altec Model 19 style slot/duct vent, on the side of the woofer? It kinda looks cool.

az
 
I was considering two 4" ABS pipes, 7.125" long for my 35Hz. It would hit 5% mach at 125W, where output will be close to 120dB.
Yeah, that looks pretty reasonable. You'll probably end up around 5-6dB lower than the published sensitivity, the way my measurements have been looking (rising response).

I was also looking at a large slot on the bottom....Is there any advantage to an Altec Model 19 style slot/duct vent, on the side of the woofer? It kinda looks cool.
Yeah, a slot around 12-18"x2"x? would be great, but it's tricky to hit the tuning on slots even when the sims are accurate, because of the end correction. I'm using round rear ports for now. I'll think about other options later. Here's end correction for slots (which winISD does not let you put in, so you have to do the math yourself). These are only approximate - you can end up between them when you are sort of close to a wall but not touching it.
 

Attachments

  • PortCorrection.jpg
    PortCorrection.jpg
    89.6 KB · Views: 115
Yeah, that looks pretty reasonable. You'll probably end up around 5-6dB lower than the published sensitivity, the way my measurements have been looking (rising response).


Yeah, a slot around 12-18"x2"x? would be great, but it's tricky to hit the tuning on slots even when the sims are accurate, because of the end correction. I'm using round rear ports for now. I'll think about other options later. Here's end correction for slots (which winISD does not let you put in, so you have to do the math yourself). These are only approximate - you can end up between them when you are sort of close to a wall but not touching it.

Thanks, that's some good info!

5.5cu.f @ 40Hz looks good too. I'll have to try a few tunings and get a feel of what works best for me. 35Hz to ~43Hz looks like a reasonable range for our box volume.

How much extra space are you allowing for "displaced air"?

az
 
For the waveguide, I never really considered the PT waveguide as it might a bit small to go with a 15" woofer. TD15M is supposed to have good dispersion up to 2kHz though. Something larger will load the compression driver better as well. There was not a ton of options there: the Dayton 12" round WG, the 12" SEOS and the QSC were what I was able to find. Reading up on these and considering other criteria of the project, the QSC was picked, although it seems a bit hard to get one's hands on them lately.

Did you look at 18sound? Newman had a set at a Fest a few years ago and they looked really nice. I just heard them by themselves and can't remember what driver was on them but it's somewhere else to look if you haven't already.
 
You mean my cabinet volume when it's empty? They are 5.75 ft³. I estimated the subtraction from the ports and the drivers to be around 0.4ft, I think... The difference between 5 and 6 ft³ with these drivers is almost nothing, so I'm not too concerned about it.
 
RayW - I think I saw a few 18sound WG's a while ago, but had forgotten about them. The QSC's are on their way in any case.

Dumptruck - close enough to the 0.5cu.f I roughly figured.

az
 
Indeed, Be aware that your listening axis will be unknown until the crossover design is complete, so you'll have to plan for possible stands/tilting unless you do a test cabinet.

If I understand correctly, by "listening axis", you mean the actual location of the main lobe, vertically? I guess this can shift with different acoustic center alignement and any delay in the Xover. Good point.

I knew I'd get some extra enlightenment by posting a thread way before I get my hands dirty. The time to ask questions is not when the glue is dry OTOH. :)

I guess figuring out the correct amount of CD-compensation for my driver/WG combo is about all I can do in advance then.

az
 
Last edited:
Back
Top Bottom