DIY ?uestion

@trelja said: "True transmission lines (TL) handily outperform both bass reflex / ported and sealed and aperiodic bass loading configurations."

I agree that good transmission lines sound wonderful, and have greater potential than either reflex or sealed or aperiodic boxes. As a manufacturer, with enclosure cost being the dominant cost, I believe that I can offer more bang for the buck with a good reflex box.

Going back to my DIY days, one issue that I ran into with transmission lines is the one-wavelength cancellation dip. At the frequency where the path length is equal to one wavelength, the backwave energy emerging from the open terminus is 180 degrees out-of-phase with the energy coming from the front of the cone. The two are physically close enough that significant cancellation occurs. I could hear the dip before I figured out what was causing it. You can see it in anechoic measurements of transmission line speakers:

https://www.soundstage.com/index.php?option=com_content&view=article&id=775:nrc-measurements...

https://www.soundstage.com/measurements/pmc_gb1/

The dip is probably not as bad in the power response (summed omnidirectional response) but it will still be present. Fortunately dips look much worse on paper than they actually sound - the ear is pretty good at ignoring them.

There are techniques for mitigating the dip (positioning the woofer partway down the line, using very dense stuffing, building a Helmholtz absorber into the line), but like practically everything else in speaker design, they involve tradeoffs.

From your experience, do you have any comments on the one-wavelength dip? Did Bud Fried do anything in particular to mitigate it?

Trelja again: "As great as TL is for bass, it’s all the more advantageous for midrange loading."

Again, I agree. I think the transmission line acts like a "trap" for the backwave midrange energy, so that very little of it reflects back into the cone.

I had an article published in SpeakerBuilder magazine back in 1986 that used transmission line loading for the two 7" midwoofers and for the 30" tall ribbon they were mated with. Over the years several people have told me they tried various enclosure designs for the same ribbon and my "W-shaped" transmission line geometry worked the best.

I still think about doing a transmission line "satellite" speaker that would be augmented south of 80 Hz by a distributed multisub system, since I believe the room effects are the biggest issue in the bottom two octaves. This would let me use a smaller transmission line box (less costly) and would take advantage of its superior midrange potential (which is imo its biggest benefit). Maybe one of these days.

What I have found with reflex boxes, and this is somewhat counter-intuitive, is that placing the ports right smack behind the woofer cone results in better midrange than from a sealed box. I think this is because the (flared) ports act like traps and remove at least some of the midrange energy that would otherwise reflect back into the cone. So this is perhaps a crude first approximation of one of the benefits of a good transmission line. On these speakers I include a reduced-level rear-firing tweeter to combine with the port-escape midrange energy, to correct the spectral balance of the reflections.

Duke

Thank you very much, @trelja. I made a lot of sawdust trying out different transmission line ideas in the late 70’s and 80’s.

You wrote:

"Bud Fried considered it critical importance to select a driver with as low a Qts specification as possible. Over the years, I’ve learned that’s exactly opposite of most opinions I find in the internet discussions, which recommend the opposite."

Must admit that I had better results with medium to high Qts woofers in transmission lines, but then I’m no Bud Fried!

Can you reveal anything else about what parameters Bud looked for? I used KEF B-139’s in most of my sixty-something transmission lines, but I’m sure the over-the-counter version’s parameters were not ideal. It’s x-lim was rather modest, as I found out the hard way.

I would love to know how to use a low Qts woofer in a transmission line. Low Qts = high motor strength = good midrange articulation, in general.

If you don’t think it would violate any confidentialities, I would be very interested in the plans for the D that you mentioned. I’d love to peek under the hood and see what the master did.

Very best wishes,

Duke

@trelja ,  Thank you very much!!

I would love to see plans for the D and C, and the C's crossover!!

Yeah the series crossover is a strange beast at first glance!  Instead of blocking the lows from reaching the tweeter and the highs from reaching the woofer, the lows go AROUND the tweeter and the highs go AROUND the woofer.  So the specific filtering for each driver is actually in PARALLEL with that driver, but the components are all inter-dependent so it's really all one big filter instead of a highpass filter + a lowpass filter.

My recollection from experimenting with series crossovers many years ago is that you need to start out with drivers that have a smooth response with equal overlap on both sides of the crossover frequency.  I used a Hiquphon tweeter, probably tried several midwoofers but don't remember which ones. I got better results (to my ears) with parallel crossovers using filters that had some electrical asymmetry, so I don't think my midwoofers were a good match for the Hiquphons. I do remember adjusting the L & C ratio but thought that was just varying the amount of ovelap in the crossover region and didn't realize it was varying the slope as well. I didn't have decent measuring equipment back then (in my amateur days).

Some pretty serious EQ is needed for my constant-directivity horns, as without a crossover their response trends roughly -6 dB per octave from about 1.5 kHz on up.  (This is true both on-axis and off-axis, unlike with direct radiators, so when we fix the on-axis response we have also fixed the off-axis response - which imo is a very good thing).  So in this case an electrically asymmetrical crossover is needed to achieve approximate acoustic symmetry, and I don't think that can be done with a series crossover because the filter is one inter-dependent thing instead of being separate lowpass and highpass sections, but I'll take another look.

Did Bud have designs where he went for time coherence (first order crossovers + lining up the arrival times?) 

Can you tell me anything about the Qts range Bud preferred? 

By any chance, will you be at RMAF this year?

Thanks again!

Duke