Using a High WAF Design to Block the Back Wave

MartinLogan Audio Owners Forum

Help Support MartinLogan Audio Owners Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Bernard

Well-known member
Joined
Aug 14, 2007
Messages
4,475
Reaction score
22
Location
Ottawa, Canada
My front wall consists of all windows, possibly the worst arrangement for dipoles. Panel traps are not an option (blocks the view), so I toyed with the idea of buying ASC Tube Traps and placing them behind each speaker, to block the back wave; MOON (Greg) has had great success doing this). BUT, my wife did not want to have "two more towers" in the room.

So I decided that I would place the traps on the bass cabinets of my SL3's. I contacted ASC to ask them their opinion, and they thought it was a good idea. I then decided that I would build some myself instead, as I did not need bass traps as such there; I just needed to absorb the back wave.

A friend of mine made some wooden barbells for me. I wrapped them in plastic mesh, glued 4 layers of a thin blanket to the mesh, then covered them with a nice-looking cloth. Total cost of all the material for both speakers was about $170. I got the wood and the mesh from Home Depot, the blanket from Walmart, and the cloth from a fabric store.

I'm very pleased with the results. Bass seems to be better (a bit bloated, though, on some songs), and there is a lot more clarity and detail in instruments and vocals. In some cases I hear new instruments in the mix.

My wife likes the results and also hears the changes,

Attached are pictures of various stages of the build.

Tube Trap.jpg
 
Last edited:
When I first saw that pic I thought oh my, oh no. Seeing the construction of it with the explanation has let me relate to the theory & why. Smart thinking, good way to keep the cat at bay too.
 
Good looking, nice work. I experimented using using books to adj. height to tame bloat ness and found 3 inches did the trick
Gotta get slabs made up now.
 
Great Job, Bernard! The bass may still sound a little bloated because you are taming the higher frequencies but not the lower mids. The panel puts out from about 250 Hz. on up, and I doubt you are absorbing much from 250hz. up to maybe 750 hz. These mid range frequencies will still reflect and cause comb filtering and other audible anomalies. I would also be a little concerned that with the absorption so close to the panel, you are not allowing your panel to "breath". In other words, you potentially constrict the flow of air out the backside of the panel, making it more difficult for your mylar to push in that direction, potentially muddying your transients.
 
Bernard, glad its improved your expereince.

couple of quick questions/observations:

Is the cylinder filled with anything?

If not, then redoing with it filled with standard attic insulation batting might be a big help in taming the imbalance in the bass.

I'd be really curious to see the before/after impulse response of the solution. Do you have a measurement system?
 
In other words, you potentially constrict the flow of air out the backside of the panel, making it more difficult for your mylar to push in that direction, potentially muddying your transients.
Rich, if anything the transients are better and so is the imaging and the low-level detail. Remember though that I'm comparing this result with that obtained with a wall of glass. It would possibly be different if my wall were something other than glass.
 
Is the cylinder filled with anything?

If not, then redoing with it filled with standard attic insulation batting might be a big help in taming the imbalance in the bass.

I'd be really curious to see the before/after impulse response of the solution. Do you have a measurement system?
Jonathan, the cylinder is not filled with anything. I did consider attic insulation, but that is fibreglas which I did not want to handle (mask, gloves, etc.). Do you have any safer alternatives? Redoing it, though, would be a lot of work as I glued everything inside together. I would want to do measurements before doing that. Which brings me to your second question: I do not have a measurement system, and was wondering what you'd recommend that is reasonably good but does not break the bank. I suspect that you may have already answered the question in one of your threads, so I should do a search.

Addendum: Just saw your post on the Dayton Omnimic.
 
Last edited:
I got the Dayton Omnimic as a Christmas gift, finally got over the 'flu and am well enough to do some testing. I bought a cheap mic stand from a music store (amazed that it was under $30).

I did 2 short burst sine sweeps, with and without the traps in place. The sweep was fed to both speakers, and the mic was at ear height, pointing between the speakers, at my normal listening position.

Attached is a picture. The top (red) sweep is without the traps, while the bottom (blue) sweep is with the traps in place. Apart from the region from 160 Hz to about 700 Hz, the traps do not seem to do much, but even at that the difference is minimal. I do hear the differences I mentioned, though, which is not surprising as that is within vocal range, and I listen to a lot of female vocals.

I'll be interested to read Jonathan's comments and suggestions, as he suggested measurements.
 

Attachments

  • Omni.jpg
    Omni.jpg
    23.4 KB · Views: 1,548
Last edited:
I got the Dayton Omnimic as a Christmas gift, finally got over the 'flu and am well enough to do some testing. I bought a cheap mic stand from a music store (amazed that it was under $30).

I did 2 short burst sine sweeps, with and without the traps in place. The sweep was fed to both speakers, and the mic was at ear height, pointing between the speakers, at my normal listening position.

Attached is a picture. The top (red) sweep is without the traps, while the bottom (blue) sweep is with the traps in place. Apart from the region from 160 Hz to about 700 Hz, the traps do not seem to do much, but even at that the difference is minimal. I do hear the differences I mentioned, though, which is not surprising as that is within vocal range, and I listen to a lot of female vocals.

I'll be interested to read Jonathan's comments and suggestions, as he suggested measurements.
Good for you for buying the Omnimic, Bernard! Remember, FR is only part of the story. Assuming you have the newest Omnimic software, take ETC measurements, with and without the traps in place. Try full-frequency with timescale set to 0-50mS, to start. What you should see is that the traps have "killed" a ton of front wall reflections.

/Ken
 
That's not too bad, Bernard. I did a similar thing on the cheap by borrowing some kit.:)

I could post the results if you'd like.
 
Sure Justin, please post the results.

Ken, I have to fiddle with the ETC measurements; not sure if what I'm getting makes sense, and the software just went screwy on me; have to reboot.
 
Bear in mind I altered the x-over to roll off severely after 12KHz. This was done to prevent extreme mega-resolution that I found difficult to cope with. You'll just have to believe me that subjectively, this is what I wanted. Now the new MRT ribbons are fully broken in, I believe I will be getting some external x-overs made, with a two position jumper - flat and rolled 0ff.

Anyway, here it is. FR with Amperex 211s. Other tube type not included in this plot.

AmperexSmoothed.jpg
 
Last edited:
A few points:

1) The mic for this plot is at the listening position - literally with the mic as close to where my head usually is as possible i.e. very close within an inch or two.

2) The 30 Hz bass hump was noted by Martin Colloms when he did an FR of the Duetta Sig. But if I move the mic forward to two metres from the speakers, it goes. Rear wall enforcement? Probably. Head isn't too far away from it.

KLM2 (the ribbons that were installed after the refurb) did not need this treble roll off. KLM5 plus a resistor pack change did. But subjectively, this arrangement wallops KLM2 to my mind/ears.
 
Remember, FR is only part of the story.

Agreed. The primary purpose of the traps behind your speakers is not to alter the frequency response at the listener's position by any great amount. Their purpose is to reduce late-arriving sound reflections which confuse your ears, aka comb filtering.
 
Is there any way to apply smoothing to those ETC plots, Bernard?

The traps are definately doing something:)

BTW: have you been using them all the time since you made them? Or on and off?
 
Last edited:
Justin, there does not seem to be a way to apply smoothing in the software, unless I'm missing something. However, the data is available in a text file, and consists of timestamp/amplitude pairs. It should be possible (I did this an eternity ago on a PDP-11) to feed those values into a function and get the coefficients of a Fourier Series. Then by just removing the coefficients of the higher frequencies you are left with the lower frequencies - a smoothed curve. I found this really useful at the time, as applied to EEG's (I worked in a university hospital). I believe this is how an FFT is used.

I have been using them all the time since I made them; they made a definite positive difference, so I left them in. And no, I didn't do any double-blind testing; hell, I didn't do any single-blind testing. :)
 
Last edited:
Here's an interesting plot showing the effect of different 211 tube types on FR. This is the area of the FR that showed the biggest difference. So tube rolling does make a difference - at least measurably.

Test mic was the Behringer - pretty good mic I think. Beats an iPhone anyway!:) M-Audio external USB soundcard/preamp.

Coloured coding on the left should be obvious. Best save the pic to a file, then view in a JPG viewer to zoom.

50-90Hz.jpg
 
Last edited:

Latest posts

Back
Top