Slot port enclosures: successes and failures, please post...

The answer is friction. The air in a duct has significant friction with the duct walls within perhaps a quarter inch of each wall. If that high friction zone is a significant portion of the vent cross-section then chuffing will be more apparent and box tuning will be lower than otherwise.

Chuffing is more of a problem with lower tunings, as they result in higher port velocity.

Slot ports LOVE them.
Once you have listened to a design with a slot port integrated properly over a flared port same design, the bass unloads very well behaved with the slot port.
Having built a few hundred slot port speakers the main key to the slot.
REAR wall area as well as a radius.

Make sure you have at least 1.5" from the slot to the rear wall for proper air speed.
As well on the slot port on the inside radius the slot port board.f you can add in a full radius around the board you use for a slot port.
Personally I use a sharp punch and yes dimple the radius on the inside of all my slot ports.As well I line the slot port board with a peel stick floor tile for vibration control.

LTD, I too have seen the 1:8 "rule" (think it started with a JL "whitepaper"), but haven't found any supporting evidence. My own experience says it doesn't seem to matter. I'm currently running a pair of RSS315HF-4s in ~3.5 ft^3 each. Tuning is 17Hz based on the driver dip/port peak output measured by a Cross-Spectrum calibrated EMM6. Port is 1"x16"x{something} (HxWxL). The port is full cabinet width with a 90° bend to use the top, back, and side walls to form the port. I also don't use a rumble filter and the amp is -3dB at 10hz/3Hz depending on which mode I use. All of that said, I've never gotten any port noise or other apparent misbehavior. The in room measurements also track the WinISD model well once you squint enough to filter out the room effects. My port tuning did come in a bit lower than predicted...probably a combination of net volume possibly being slightly off from plan, a bit of R19 in the cabinet, and the virtual lengthening of the port caused by using the back wall of the cabinet. IIRC, I did shorten the as-built port about an 1" from the model to try to compensate for the virtual lengthening, but that was just my guestimate as there's no ROT to predict the change in tuning that I'm aware of.

BTW, in the photo, you see the cabinets, but at the time I was comparing a SVS 12.1 with the 315HF-4. Currently, both cabinets are running HF-4s.

-Brent

I've not found any ratio to be better or worse, it's just a matter of velocity versus the narrowest part of the cross-section, and unfortunately the best way to find out that it's too narrow is when you hear it chuffing. I try to avoid less than a 1.5 inch dimension, and that would be with a woofer. I wouldn't go less than 3 inches with a sub.

A cylinder is going to give the lowest surface area (port walls) to volume, and also the largest minimum distance between opposite sides. Air flow in a port is assumed to be laminar, and it will be to a large extent, until airspeed becomes high enough that the non laminar flow becomes a significant proportion of CSA and the air stops moving as a 'slug' in the port. This will to my thinking reduce it's ability to behave like you would expect for a helmholtz resonator and begin to compress, more so for slot as the side walls are closer together.

Flares at both ends would be useful for both slot and cylindrical ports to smooth the transition. B&W have done dimpled ports, but they were flared cylinders. Others may have used them too.

Bill's ROT's for dimensions sound OK to me, though I'd be more comfortable with 3 or 4 to 1 as a max.

The only advantage I see to a slot is to ease construction or to give a CSA not easily obtained in a common tube size.

I'll try to fit a test in, but it'll probably be a few days before it fits in the schedule.

FWIW, when initially built a few years ago (maybe even 5yrs, time flies), I compared and somewhat attempted to find the limits of the 315HF-4 and 12.1 by measuring each individually using the same room placement. I even placed myself and the off cabinet in the same locations to try to limit variable changes to the driver only. Similar to Ilkka's compression testing, I ran successively higher level sweeps listening for bad noises. I quit before hearing any port noise. Somewhat surprisingly given the claimed Xmax/Xsus of the TCS db500 (rumored to be leftover 12.1 stock), the 12.1 started making suspension noise at the low end of the REW sweep (10-20Hz range) while the 315HF was still dead silent, so I stopped there before either driver actually went clank*. Pretty sure I was getting just over 100dB average at the listening position from each for that final sweep level...DUT was corner loaded, ~14' to the measurement position. My notes say MV on the Onkyo 805 was at -1dB for the final sweeps, but that's kind of meaningless without knowing what the signal level coming from REW/laptop were. When I get around to this new test, I'll use Soho's test disc so the input level of the test signal is known.

*As a side note, the 315HF and 12.1 measured virtually identically on each sweep. We're talking <0.5dB variation.

-Brent

Engineering gut response that I have learned to trust over the years. Other people can argue the point and/or build whatever they want, won't make any difference to me. I was saying what I feel comfortable building not issuing a fiat.

The cylinder is the standard and anything away from that will be a compromise to some extent for the reasons above. If a 6" cylinder gave you what you needed , then the same CSA rectangular would too, up until turbulence reduces performance. The closer the walls, the earlier in my view this will kick in limiting output near max and I don't see the point in building a compromise.
Not that I have, but it would be an interesting experiment. I have no ported designs in planning or under construction, so I won't be doing it.

Well, it took me a little longer than expected to get to it, but I finally had the house to myself and a few minutes to throw at setting up the measurement rig. I fed the pair of 315HF-4s the 17hz LFE test tone from Soho54's audio test DVD at Reference level master volume. Mic is 11 feet from the subs according to Audyssey MultEQ XT.

As shown in the attached RTA capture from REW, there was nothing but the 17Hz fundamental within at least 35dBs of the 17Hz peak. I only got smart and did the RTA cap on my last run (of several at full Reference plus a few at lower levels) and didn't think about widening the SPL range until the gear was put away and the data actually examined...sorry. REW's capture function didn't include the THD window, but it was running <2% when this was taken...though, I'm not sure if REW can accurately calculate THD when it doesn't generate the signal. As for what I heard, well there was a ticking sound that was either lead slap, suspension noise, or the interconnects inside the cabinet. There was a good breeze coming from the ports, but not anything I would interpret as "chuffing" the way I've read it described and the RTA would seem to support that. Subjectively, Reference (0dB MV) sounded the same as -10dB MV, only louder.

-Brent

BTW, if you're interested, here are the REW sweeps from that session. No smoothing applied, green=Audyssey ON, purple=Audyssey Off. Throw a little smoothing at it and the green curve nearly becomes a straight line. :-) As I said earlier, they measure pretty much the way WinISD predicted they should.

-Brent

Maybe check out this site.
http://www.subwoofer-builder.com/flare-testing.htm

The long and short is that flares help some, but can't compensate for an undersized port. Next is the audibility threshold is reached with a higher port speed as the ratio of cross sectional area to surface area increases. In layman's terms, chuffing will be more pronounced with a thin, wide slot port for the same port air speed.

That's a negative, Ghostrider. Considering my typical playback volume is only -15dB to Reference, 0dB is as hard as I'm willing to push, especially since I'm not sure what was causing the ticking I was hearing. While tuning point represents minimum excursion, it also likely means maximum heat being sunk by the coil. And P-E doesn't put the 315HF's on sale for $100 any more like they did when I bought this pair so skirting the envelope of distructive testing is not in my budget.

Without knowing how much SPL I'm losing to distance, it's hard to guestimate where on the power curve I really am. And even then we're all trusting the models to be accurate regarding air velocity with excursion. Think Cantori's anemometer can measure port air speed? If we assume I'm losing only 6dB for distance (say 11'=4m for easy math), I'm getting more like 15m/s in the WinISD model. If the chuffing threshold is as low as some tend to believe and then made even worse by a 1:16 slot ratio, one would expect even your 11 m/s estimate to be producing some type of measureable unintended noise.

When I get time for another measurement session, I'll focus on the RTA more to try to find any "noise" being created by port turbulence. As it is now, any noise being created is more than 35dB down for me and would easily be masked by real world content.

-Brent

Wisound has a pretty neat idea with the ports all along the edge of the baffle. Check out some of his posts..

Well, wait a sec, he does. As port diameter increase, cross sectional area increases more than surface area.

The noise limit of a 6" port has a higher (air velocity) noise threshold than a 4" port. The difference between them is the 6" port has a higher cross-sectional to surface area ratio. So, 2, 4" ports that have reasonably the same cross-sectional area as a 6" port, but will have the same noise threshold as a single 4" port. What's different between 2, 4" ports and 1, 6" port? Lower cross-sectional to surface area ratio. And skinny slot ports have a lower cross-sectional to surface area ratio. If A=B and B=C, A=C.

Biggest thing I have yet to do is an impedance sweep to determine cab tune. I was targeting 30hz. I have had several makes and models of driver in the bottom subwoofer cabs and they all have sounded awesome, 18s and 21s. No noticeable/audible chuffing. If sidewall friction lowers the tune, I'm all for that.

Here are a few examples of slot ports I have worked successfully: